Cornell Muiversity Library

BOUGHT WITH THE INCOME
FROM THE

SAGE ENDOWMENT FUND

THE GIFT OF

Henry W. Sage

1891

ALLS aaj]

 

 

 

ENGINEERING LIBRARY
ii

3
( VONVENVHONVH JO UNdS NUALSVH) LSGM HIMON FHL WOU -7100e%2-OAWIS

-gegt Amp eMoTeg ‘seowo erpuy jo Aeamng a “UURULyOY  UojsUYOP ,,sseyy 4G ydesSoio4g 2 WO Surmaye-cx904

 

 
A MANUAL

GEOLOGY OF INDIA.

CHIEFLY COMPILED FROM THE OBSERVATIONS
OF THE GEOLOGICAL SURVEY.

STRATIGRAPHICAL AND STRUCTURAL GEOLOGY.

SECOND EDITION REVISED AND LARGELY REWRITTEN BY

R. D. OLDHAM, A.R.S.M.,

SUPERINTENDENT, GEOLOGICAL SURVEY OF INDIA.

 

Dublished by order of the Gobernment of Edit.

CALCUTTA:

SOLD AT
OFFICE OF THE SUPERINTENDENT OF GOVERNMENT PRINTING;
GEOLOGICAL SURVEY OFFICE; AND BY ALL BOOKSELLERS ;
LONDON: TRUBNER & CO. F

 

MDCCCXCIII.

Price Rs, 8 (x6s.)
CALCUTIA:
PRINTED BY TUE SUPERINTENDENT OF GOVERNMENT PRINTING,
8, HASTINGS STREET.
PREFATORY NOTICE.

 

N the beginning of 1887 my predecessor, Mr. Medlicott, wrote as fol-
lows in his Annual Report of the Geological Survey :—“ The two first
parts of the Manual of the Geology of India, issued in 1879, have been
out of print for some time, and the question of re-writing it has been
much upon my mind, Parts of it would require abridgment, leaving local
information to be sought for in the special Memoirs; and parts of it
would need alteration and addition in view of extended information.
The greater part of the two volumes was written by Mr. Blanford,
who was for the time relieved of other work. To re-write the whole
while carrying on the manifold current duties of the Survey has been more
than I could attempt in India with any justice to either.”

The directing of the Survey since Mr. Medlicott’s retirement is even
fuller of current duties, not the least of which has been a considerably
increased system of frequent tours over the length and breadth of the
land ; so that, however pressing it may also have been on my mind,
Ihave been unable even to venture on the elaboration of a revised form
of Messrs. Medlicott and Blanford’s most excellent work’; and I therefore
gladly accepted Mr. R. D. Oldham's offer to prepare a fresh issue accord-
ant with our progressive survey of the Empire.

Mr. Oldham had had a varied experience of survey work over widely
separated tracts in India where he had opportunities of studying most of
our representative formations in their peninsular and extra-peninsular
development : while of his own motion he devoted his first period of
well-earned leave to a comparative study of our Gondwd4na representatives
in Australia. His close acquaintance with the literature, as evidenced
in the careful Bréliography of Indian Geology, compiled by him in 1888,
had already indeed predisposed me in favour of a possible ultimate
placing of a second issue of the Manual in his hands; and in now au-
thorising that issue I would fain hope that my choice may be justified.

WILLIAM KING,
Director, Geological Survey of India.
PREFACE TO THE FIRST EDITION.

 

‘HE want of a general account of Indian Geology has been felt for
some years. The regular Geological Survey of India may be
considered to have commenced in 1851; and but few of those who took
part in the work during the earlier years now remain in the service. It is
desirable, before all the older members of the Survey pass away, that
some record of the early observations, many of which are unpublished,
should be rescued from oblivion, for the benefit of future explorers. The
published Memoirs and Records of the Survey, moreover, have now
become too numerous and bulky for general use; and it is difficult for any
one, without much study, to gather the more important observationson the
geolegy of the country from amidst the mass of local details. Many papers
on Indian geology are also scattered through various Indian and European
periodicals. As a guide to all who have occasion to acquire a knowledge
of Indian geology, or who desire information from a love of the science,
some compendium of the observations hitherto collected has become abso-
lutely necessary ; and the present Manual has been drawn up, by direction
of the Government of India, to supply the deficiency.

It was originally desired by the Government that this work should be
prepared by the late Dr. Oldham, or that the compilation should have the
advantage of his supervision. As Dr. Oldham was the first Superintendent
of the Survey, and remained at the head of the Department from its com-
mencement in 1851 to 1876, he would, unquestionably, have been admir-
ably qualified to carry out the work ; and it was his own desire to do so, as
the completion of his labours in India. Failing health, however, and the
pressure of other duties, prevented him from even commencing the task ;
and when, at length, he was unable any longer to remain in the country,
the duty of preparing a Summary of Indian Geology was left to his
successor, At this time the only preparation that had been made for the
work was the partial compilation of a general Geological Map of the
Peninsula.

The double authorship was not entirely a matter of choice; although
undertaken, and carried out, most willingly by both the writers. Both
have been engaged in the work of the Survey almost from the commence-
ment ; and as each has, in the course of his service, examined very large
vi PREFACE.

areas of the country, the combination secures the description and discus-
sion, from personal knowledge, of a much larger portion of India, At the
same time the advantages of wider experience and thought may not be
found an adequate compensation for want of uniformity and occasional dis-
crepancies—the natural results of divided authorship. To secure, so far
as possible, the responsibility of each author for the facts and opinions
stated, the initials of each are affixed in the Table of Contents to the chapters
contributed by him. Every such chapter has been read and revised by
the other writer; but the alterations have in no case been of more than
trivial importance; so that each chapter may be practically taken as an
individual contribution. The number of subjects is so large, and the con-
nection between them, in many cases, so slight, that the lack of uniformity
will not, it is hoped, seriously detract from the usefulness of the Manual.

In addition to the subjects discussed in the present work, it was, at first,

proposed to add an account of the Economical Geology, and to treat in a
special chapter of the known Mineral Resources of India. But the length
to which the Manual has already extended has rendered it advisable to
postpone this very important subject, and to reserve it for a separate
volume.
Although many of the details in the work now issued Lave not previ-
ously been published, and although the discussion of the observations involves
several new deductions and suggestions, the book is, in the main, a compi-
lation ; and it is quite possible that, especially in treating of areas and
formations of which the authors have no personal knowledge, full justice
has not always been done to the views of original observers. It has, in
several instances, been thought more important to point out possible
causes of error than to endorse opinions which, although very possibly
correct, are not sufficiently supported by published data to be accepted
as conclusive. In all such cases full references to previous publications
have been furnished ; and an examination of the details given in the latter
will, 1t is hoped, serve to correct any errors of interpretation’ on the part
of the authors of the present work.

The numerous and large areas left-blank in the annexed Map show,
at once, how far the present publication falls short of completeness, and
how imperfectly the promise implied in the title is fulfilled. A note upon
the Map further explains that large portions of it have been coloured
from very imperfect information, from sketch surveys or rapid traverses
affording no sufficient opportunity fora proper study of the formations.
It had, however, become imperative, as a duty to the public, for reasons
already mentioned, to bring together a summary of the work accomplished
since the commencement of the Survey; and it was equally essential, for
the Survey itself, that some general record of the results obtained up
to date should be compiled. These objects could only be attained by
PREFACE, Vii

attempting a general Map and Review of the Geology of India; but the
reader must not forget that the present attempt is more of the nature of
a progress report than of a finished work.

The Map, it is feared, will be found defective in several other respects.
Under the circumstances it was impossible to prepare a special reduction
of the topography ; and, amongst the Maps of India available in the Sur-
veyor General’s Office, there was, practically, no choice but to accept
that on the scale of 64 miles to the inch, then well advanced towards com-
pletion, as a basis for the geological details. The scale is inconveniently
small for all parts of the country that have been geologically mapped in
any detail, and the mountain ranges have not been inserted ; so that many
features discussed in the text are not indicated. But the most serious
drawback is in the names of places. Many towns of importance are
omitted, owing to the small scale; and other names of interest, for pur-
poses of geological description, such as those of fossil-localities, or of
villages near important sections, are wanting. Nor is this all. The spelling
of Oriental names is a well-known cause of perplexity ; and the confusion
has been increased by the unfortunate circumstance that, while one sys-
tem has been adopted by the Great Trigonometrical Survey, and employed
in all the maps, including those of the detailed Topographical Surveys,
issued by the Department, an entirely distinct system has been employed
by the Revenue Survey, by whom the maps of all the best known parts of
the country have been prepared. Under the first system, each letter in
the Indian language is represented by a corresponding letter in the
Roman character ; diacritical marks and accents being employed to distin-
guish such consonants or vowels in the latter as are required to represent
two or more sounds, and the Italian or German sounds of the vowels
being used, instead of the English. Under the second system, an attempt
is made to represent the original sound by English spelling ; double vow-
els being largely used, but no diacritical marks. The imperfection of the
latter plan is manifest ; because, in the first place, the sounds, of the vow-
els especially, in English, are variable, and incapable, in many cases, of
representing those of Oriental languages ; and, secondly, the representation
of the true names by supposed equivalents is arbitrary, depending chiefly
on the ear, often very imperfectly trained, of the transcriber. When maps
of large areas, as inthe present case, are compiled, the mixture of names,
spelt according to two different systems, is inevitable. The attempt at a
general revision of the nomenclature, however desirable, would have
involved serious delay.

Of late, the Government has adopted a compromise in the question of
spelling, and lists of the principal places in each province have been
issued; the familiar and well-known names being spelt in the manner that
has become customary by usage, whilst transliteration is employed in all
vill PREFACE.

other cases, with the exception that no diacritical marks are used for
consonants. This system is obligatory for all official publications ; and it
has, consequently, been adopted in the present work. In some cases,
however, the lists for particular provinces have not been published in time
to be available ; and in the following pages it is not unfrequently necessary
to mention places not contained in the lists, and the proper vernacular
pronunciation of which is unknown to the writers. In such cases, an at-
tempt has been made to spell the name according to the recognised system ;
but it is only fair to warn the reader that no dependence can be placed
on many names of places, specially upon those in the south of India, when
taken from old maps.

In the preparation of the Map a large share has been taken by various
Officers of the Geological Survey, all of whom have contributed. The
colouring and printing have been carried out at the Surveyor General’s
Office, under the superintendence of Captain Riddell, R.E., to whom the
authors beg to express their obligations for the labour he has given to the
work, and for the assistance he has afforded to them personally.

In the plates of fossil plants and animals at the end of the work some of
the most common and characteristic forms of organic remains found
in India are represented. The plants have been selected and arranged by
Dr. Feistmantel, and the tertiary Mammalia by Mr. Lydekker. All the
plates are lithographed by Mr. Schaumburg, whose work will answer for
itself, The majority of the figures are from original drawings, or from the
“ Paleontologia Indica ;” the remainder are copied from other works; but
these copies have, in many cases, been compared: with specimens.
PREFACE TO THE SECOND EDITION.

 

HE first edition of the Manual of the Geology of India was found to sup-

ply a want so much felt that it soon went out of print, and for several

years has only been procurable from the dealers in second-hand books.
It had, besides, become out of date in many parts, and the Government of
India decided that the time had come for the preparation of a new edition.

So great have been the strides made in our knowledge of Indian
Geology in the last fifteen years that it has been found possible to entirely
change the arrangement of the book and to adopt the more scientific and
orderly course of describing the rocks in chronological order, instead of
breaking the book up into a series of descriptions of separate districts,
as was found to be inevitable in the preparation of the first edition.
There is not, unfortunately, everywhere the same certainty regarding the
proper position in a chronological arrangement of particular groups of
beds. Frequently the true homotaxis is unknown, and there is a conspicu-
ous instance of the mistakes which may then be made, in certain rocks
of the Arakan Yoma, which were supposed to be triassic at the date when
this second edition was written, but have been shown to be eocene as the
pages were passing through the press, and consequently rocks which
should have been classed with the tertiary were described in the chapter
devoted to the carboniferous and trias.

Many districts have remained untouched since the publication of the
first edition, and in all such cases, and wherever indeed no serious modifi-
cations of the original text have been necessary, it has been allowed to stand
practically as in the original publication. All these passages have been
carefully revised and generally more or less condensed, while many minor
alterations needful to adapt them to the altered scope and arrangement of
the work have been made, as well as such alterations as have been neces-
sary on account of the advance of our knowledge. In the table of contents
the portions in which the first edition has been taken as the basis of the
text are distinguished by a different type from that which refers to the
portions which are new or have been entirely re-written.

As there is now aspecial volume devoted to the economic aspects of the
geology of India, not tomention the Dictionary of Economic Products and
the Handbooks of Commercial Products of the Imperial Institute, it has
x PREFACE,

been decided to exclude all references to economic geology in this work.
The references would necessarily have to be too brief for commercial pur-
poses, while they would have confused the stratigraphical descriptions and
increased the bulk of the volume. It is, therefore, better that these sub-
jects should be relegated to the books specially devoted to them.

For the rest, the remarks in the preface to the first edition, relative to
the geographical basis of the Geological Map of India and the spelling of
place names, still hold good. As regards the latter, the Imperial Gazetteer
of India has been adopted as a standard by the Government of India and
the Royal Geographical Society. The spelling in that work has conse-
quently been followed, and as regards places not mentioned in that work
the same system of spelling has been adopted so far as possible. The
scale of the Geological Map adopted for this edition is smaller than that of
the map issued with the first edition. It is hoped that the alteration will
make the map more convenient to consult, while the amount of detail
that can be exhibited remains practically the same, and the impossibility of
finding a map showing all the places mentioned in the text has been
rectified, so far as possible by an index of place names giving their geo-
graphical co-ordinates.

The date of the map will be seen to differ from that of the title-page
of the book; this is due to the fact that it was necessary to go to press
with the map before the text was completed. The map consequently repre-
sents the state of information available in the office of the Geological
Survey on the 31st December 1891, while the text dates about four months
later, and subsequent additions and corrections have been made in foot-
notes up to the beginning of this year,

Finally, I have to express my obligations for the help so readily ren-
dered in the preparation of this work, specially to Prof, Suess, for the loan of
the illustration block on p. 202, and to Dr. W. T, Blanford, who, in addi-
tion to other assistance, has been good enough to read the proofs of the
passages referring to the fossil tertiary and pleistocene mammailia.

ee
TABLE OF CONTENTS.

 

(Ordinary type refers to matter which has been adapted from the first edition,)

Heavy type refers to matter which is new or entirely rewritten in the present edition.]

CHAPTER I. :

PHYSICAL GEOGRAPHY.

Limits of area treated, three great divisions, Indo-Gangetic alluvium (1), penin-
sular, extra-peninsular. Contrast of peninsular and extra-peninsular regions, Extra-
peninsular rocks and peninsular structure west of Ardvallis (2); extension of peninsular
rocks into Assam hills and eastern Himdlayas; nomenclature of Indian Hill ranges,
difficulties, Peninsular mountain ranges (3), merely remnants of plateaux isolated by
erosion, Aravallis the only important peninsular range composed of disturbed rocks (4),
hills of Southern India possibly isolated by marine denudation, former continuity
of hills of Central India, absence of connection between direction of peninsular
mountains and disturbance of rocks (5), Ardvalli range an exception, but disturbance of
great antiquity. Extra-peninsular ranges directly due to disturbance of rocks, difficulties
of nomenclature (6), consideration of Himalayas postponed. Hills of western frontier,
separation of Salt range, two principal directions of strike in hills west of Indus (7),
hills of eastern frontier and Burma, Salt range and Assam Hills. Drainage system of
extra-peninsular hills always crosses axes of special elevation (8), “‘ tangis” of western
frontier, gradual elevation of hills with occasional interruption of drainage. Simi-
larity of hills on eastern frontier and differences introduced by climate. River system of
Peninsula (9), easterly trend of drainage, antiquity of east and comparativer recent origin
of west coast. Tertiary changes of extra-peninsular coast line, changes of level on
peninsular coast. Sahyddri range, resemblance to sea cliff (10), Cremnoconchus allied to
marine form, Subaerial erosion, elevation of the Konkan tertiary rather than post-
tertiary. Possible isolation of Southern Indian hills by marine denudation, Supposed
submerged cliffon Mekran coast, submerged forests of Bombay and Tinnevelli coast.
Changes of level in Cutch and Arakan (11), in Andaman islands. Coral archipelagoes off
west coast, erosion and accretion of land on east coast (12), in Cambay ; silting up of Rann
of Cutch (13), Evidence of glacial period, former extension of Himalayan glaciers, boulders
of the Potwér. No physical proofs inthe Peninsula, other evidence. HimAlayan plants
and animals on higher hills of Southern India (14), migration not due to greater
dampness (15). Volcances, Barren Island (16), Narcondam, Puppadoung (17) other
extinct volcanoes of Upper Burma ; termination of the Sunda chzin. Doubtful cases of
volcanic eruption; in Bay of Bengal (18), Londr lake. Salses or mud volcanoes (19),
Burma, Rdmri and Cheduba (20), ignition of marsh gas by lightning (21), Assam, Balich-
istan (22,,
xii CONTENTS.

CHAPTER II.

METAMORPHIC AND CRYSTALLINE ROCKS.

Great extent, imperfection of data available (23), gneissose rocks of two ages, older
or Bundelkhand gneiss, leading characters and distribution, newer or banded gneiss
appears to pass into transition rocks, doubtful validity of separation, a large part probably
newer than Bundelkhand gneiss (24). La Croix suggests presence of groups of successive
origin in Southern Indian gneisses. Bundelkhand gneiss in Bundelkhand. characters (25),
generally massive, schistose gneisses in southern portion of area (26), absence of
limestone ; pegmatite veins, absence of accessory minerals, quartz reefs, absence
in area of schistose gneiss (27), foliation of quartz reefs and- occasional presence
of serpentine; newer series of quartz veins; trap dykes (28), newer than quartz
reefs; relation to newer rocks. Newer gneiss of Singrauli (29), abundance of
accessory minerals, well marked banding and foliation (30), quartz reefs, dome gneiss,
comparative rarity of trap dykes, pegmatite veins (31). Chutiad Nagpur banded gneisses,
relation to transition rocks of south-west Bengal ; inliers of older gneiss, trap dykes,
unconformable junction with transition rocks, Gneisses of Talcher and Mahdnadi valley
(32), schists and conglomerate in gneiss. Nagpur and Nizam’s dominions. Gneisses of
Nellore, Carnatic, massive gneisses, schistose gneisses (33),.Bezwada gneiss containing
Murchisonite. Mica mines at Inikurti (34). Gneiss of Southern Maratha country and
Bellary; massive gneiss with granite iatrusioas, resemblance to Bundelkhand gneiss, quartz
reefs and dykes (35), absence of limestone, and rarity of accessory minerals. Gneisses of
the Konkan. South Malabar, three distinct forms (36), indications of successive origin.
Waindd and Nilgiris (37). Limestone and granite intrusions in gneiss of Coimbatore and
Trichinopoli, Salem district, magnesian schists (38), La Croix on Salem gnei:ses, Scuth
Arcot and Trichinopoli. Madura and Tinnevelli, supposed stratigraphical succession (39).
Distribution and composition of ancient trap dykes in Southern India. Ardavalli range,
apparent passage from gneiss to transition rocks. Limestones and slates associated with
gneiss near Jobat. (40) Extra-peninsular gneissose rocks. Afghanistan Himalayas.
Kashmir, supposed presence of gaeiss of two ages, doubtful correctness of
interpretation (41). True gneiss of Zanskar, Rupshu and Kumaun Himalayas Gneiss
of Darjiling. Porphyritic gneissose granite, description (42). mode of intrusion, similar
granite of Hazacza, distribution ia Himalayas (43), cate of intrusion. White granite and
syenite in Himalayas. Assam gneiss (44), outliers in Brahmaputra alluvium, and in Siwdlik
zone. Gncissose rocks of hills east of Burma, metalliferous veins (45). Resumé, imperfec-
tion of data, antiquity of the Bundelkhand gneiss, other gneisses largely newer, but in part
due to deformat:o. of the older gneiss (46).

CHAPTER III.
TRANSITION SYSTEMS.

Unfossiliferous older sedimentary rocks of Peninsula. Difficulty of classiGcation, divi-
sible into two, partly contemporaneous, groups systems. Newer may be classed as older
palzozoic, older as Transition (47), meaning attributed to the term. The Dharwar
system, distribution, petrology (48), volcanic beds, and deformation of pebbles by com-
pression, hematite schists and auriferous quartz reefs, unconformity to granitoid
gneiss (49), apparent local passage into gneiss, relation to Cuddapahs, complete isolation
of Dharwars, rrobable relation to transition systems of Central India, outlier of Dharwars
CONTENTS. xili

in Nellore (50), possible confusion with disturbed Cuddapahs. Bijdwar system, uncon»
formable to Bundelkhand gneiss, bottom beds (51) irregularity, total thickness, absence of
disturbance in Bundelkhand, prebable disturbance under the Vindhyan area. Beds
at the Ken, sandstone and conglomerate associated with gneissose rock (52), possibly a
metamorphosed arkose, Bijdwars of lower Narbada valley, relation to gneiss (53),
exposures near Bagh and Jobat, older sedimentary rocks associated with gneiss;
outcrops in upper Narbada valley (54), Bijdwars of Son-Narbada watershed, metamor-
phosed condition; Son valley (55) an older system also seen; section in Rer valley,
abrupt junction of gneiss with Bijdwars, and passage into slates of the Rer valley (56).
Transition rocks of Behar hills, disturbance (57), distinctness of boundary, metamor-
phosed condition and parallelism of stratification with gneiss; less metamorphosed condi-
tion of rocks in detached hills (58). Junction with gneissose rocks in the Shaikhpdra ridge
and at Luckeeserai, analogy to section in the Ken, probable equivalence of Behar transi-
lions to Rer valley slates (59). Shillong series of Assam, resemblance to Behar transitions,
two divisions, conglomerate at base of upper quartzites (60), relation to gneiss. Eruptive
diorite and granite (61), mode of intrusion of the granite. Transition rocks of Manbhim
and Singbhim (62), unconformity to gneiss inliers, associated eruptive rocks, metalliferous
veins (63). Chilpi gh4t or Saletekri series of Central Provinces. Gwalior system, uncon-
formity to Bundelkhand gneiss, distinctness from Bijawars (64), Par sandstone and Morar
group, absence of disturbance, original limitation (65), organic remains, voicanic rocks,
porcellanic beds, (66) limestone beds, hamatite shalcs, outlier at Hindaun, antiquity of
the Gwalior system (67). Ardavalli system, limitation and doubtful separation, apparent
passage into gneiss, unconformably covered by the Delhi system, conglomerates near
Udaipur (68), Jasper beds in Aravalli system, granite intrusions. Delhi system, distribution,
subdivision (69), Alwar quartzite, Delhi system near Nimach, unconformity at base of Delhi
series (70), reported passage into gneiss. Bedsin contact with Gwaliors of Hindaun (71),
post-Gwalior age of Delhi system ; flexible sandstone. Chiniot and Kirana hills, antiquity
of rocks (72). Champaner beds, apparently belong to the Delhi system (73). Maldni series,
felsites, doubiful age, formerly considered Lower Vindhyan, unconformity with Upper
Vindhyans, resemblance to Aravalli felsites of Tosham (74), classed as transition. Possi-
ble contemporaneity of part of Gwalior and Delhi systems with the Cuddapahs, relative
ages of the transition systems. Extra-peninsular area, Vaikrita system of Hundes and
Spiti, schistose beds of Nepal (75). Daling series of Sikkim, apparent infraposition to
gneiss, misleading appearance of section (76).

CHAPTER IV.

OLDER PALHOZOIC (CUDDAPAH AND VINDHYAN) SYSTEMS OF THE PENINSULA,

Absence of fossils, nomenclature, age possibly in part contemporaneous with transition
systems. Two principal areas. Southern India; Cuddapah system (77), extent, distur-
bance of strata on east, Nallamalai range, subdivision (78), unconformities between
different groups, boundaries of basin, original limit of deposition on west. Pdpaghni
group, Gulcheru quartzite, contact with gneiss, basal conglomerates; Vempallilimestone
(79), volcanic beds ; Cheyair group, two areas, Pilivendala subdivision, uncunformity to
Papaghni group, Nagari quartzites, basal conglomerate,Tadputri beds, volcanic rocks (80),
porcellanic beds, Pullampet beds, pseudo-fossils ; Nallatmalai group, Bairenkonda quart-
zites, Cumbum slates (81), Eshwarakupam hill; Kistna group, Irlakonda quartzites,
Kolamnala shales, Shrishalam quartzites. MKalddgi series, Cuddapah age, distribution
XIV CONTENTS.
(82), relation to gneiss, subdivisions, disturbance (83). Karndl scries, separation from
Cuddapahs (84), subdivision ; Banaganpalli group, diamond gravels ; Jamalamadugd
group, Auk shales (85), Narji limestone ; Paniam group; Kundair group, Nandidl shales,
Koil Kuntla limestones, supposed fossils, original limitation. Karnal series in the Palnad,
diamond sandstone (86), difficulties of correlation, disturbance of the beds, Bhim series,
subdivisions (87), relation to metamorphics, basal diamond bearing (7?) conglomerate,
slight disturbance, breccia formed iz situ, doubtful fossil (88), probable original continuity
with Karnuls. Cuddapah outliers north of the Kistna. Pakhal series, subdivisions ; Pakhal
division, resemblance to Cheyair group; Albaka division (89), Penganga beds, formerly
regarded as lower Vindhyan, identified with Pakhal series and Cuddapahs (90) similarity
of disturbance. Cuddapahs of Bastér and Mahanadi valley, Swllavai series (gt) sub-
divisions, relation to Pakhal series, regarded as Vindhyan. Barrier between Vindhyan
and Cuddapah areas. Vindhyan system (92), origin of name, extent of deposits; lower
Vindhyans, distribution (93), Son valley, Bundelkhand, and Chitor areas, outliers (94) ;
subdivisions in Son valley, relation to gneiss, porcellanic and trappoid beds, Rohtas group
(95); Bundelkhand area, subdivisions, general resemblance to Son valley rocks, Rohtas
group recognised, silicious brecc’a (96), peculiar bottom rock of eastern portion of ex-
posure, supposed xzidus of the diamond (97). Original limitation of lower Vindhyans,
overlap by upper Vindhyans in Bundelkhand. Chitor area, conformity of lower and upper
Vindhyans, basal conglomerate. [Doubtful propriety of uniting lower and upper Vindhy-
ans (98), unconformity, nature of disturbance of lower Vindhyans, apparent conformity.
Upper Vindhyans, subdivisions (99), general composition ; KX dimur conglomerate, charac-
teristics of the groups, diamonds (100), Bhander limestone, supposed fossils (101), connec-
tion between variations of subdivisions and boundaries of the basin, surface features of
Vindhyan area, general absence of disturbance, local disturbance at Jhalra Patan
(102), Dhar forest. Faulted boundary of the Vindhyans to the Ar4vallis, ana'ogy to
main boundary of the Himalayas (103), analogy of Vindhyans to Gangetic alluvium and
Siwéliks,'age of the Aravalli mountains (104). Outliers of supposed upper Vindhvans
in Bundelkhand, peculiarities and probable mode of origin (105). Vindhyans west of the
Aravallis, relation 'to Ardvalli mountains, glacial beds of Pokaran of doubtful Vindhyan
age, relation to;Vindhyans of main area (106). Diamond bearing rocks of India, of
approximately equallage, diamonds derivative, original source of the diamond (107). Re-
ae i age of rock systems, probable equivalence of Gwalior and lower Cuddapah
rocks (108).

CHAPTER V.
OLDER PALHOZOIC SYSTEMS OF THE EXTRA-PENINSULAR AREA.

Marine fossiliferous deposits. Salt tange, subdivisions (109) ; Salt marl, absence of
stratification, salt and gypsum deposits, impurities of the salt, gypsum beds (110), sup-
posed hypogene origin of the salt marl, appearance of intrusion (111), evidence sie con-
clusive. Kohat salt and gypsum, supposed age and origin, gypsum due to hydration of anhy-
drite, anhydrite inclusions in quartz crystals, general conclusions (1 12); Purple sandstone ;
Neobolus Feds, cambrian fossils; Magnesian sandstone (113) ; Salt pseudomorph zone,
Central Himélayas, classification of beds, Haimanta series (114), passage to Silurian, silurian
fossils (115). Cuter HiaAlayas, unfossiliferous slates mapped as Silurian, in fier Sarbouté
ferous ; Attock slates. trias and cretaceous fossils, Kashmir (116) ; Jenaace series, volcanic
beds, esemmblance to Haimantas ; Deoban limestone, distribution (117). Supposed
organic remains ; Baxa series ; Devonian fossils in Thibet 3 Silurian of Burma (118).
CONTENTS ay

CHAPTER VI.

CARBONIFEROUS AND TRIASSIC ROCKS OF ESTRA-PENINSULAR INDIA,

Passage between palzwozoic and mesozoic. Salt range (119), glacial boulder bed,
fragments derived from the south, facetted pebbles (120), Australian fauna, upper carboni-
ferous age (121), distribution of fossils in Productus beds (122), speckled sandstone.
Productus beds, classification (123); lower Productus beds, change of fauna, absence of
Autralian forms, European affinities (124), permo-carboniferous age, changes of land and
sea, Salt range fossils in China ;. middle Productus beds (125), relationships of fauna,
appearance of mesozoic types (126); upper Productus beds, permian facies of fauna, pale-
ozoic ammonites (127) ; Chidru group, great change of fauna, abundance of mesozoic types ;
probable age. Trias of Salt range (128). Central Himalayas, crinoid limestone, possible
glacial beds (129), Kuling series. Trias, subdivisions, lower trias (130), supposed rhztic
and lias, provisional character of conclusion (131), palzontological break between silurian
and carboniferous, absence of rhztic and lias. Carbonaceous system of Simla Himalayas,
Simla slates (132), Blaini group, glacial boulder beds, infra-Krol carbonaceous slates (133),
Krol or Boileaugunj quartzites, felspathic grits of western Garhwal. Upper carbonaceous
slates. Volcanic beds. Kashmir carboniferous (134), glacial boulder bed, carbonaceous
slates and volcanic rocks, fossils (135), affinities of fauna; Trias. General similarity to car-
bonaceous system and contemporaneity of origin (136); connection of Simla and Kashmir
areas, absence of fossils south of snowy range. Former correlation of the Blaini groups
consequent errors (137). Supposed trias fossils in Simla district. Outliers in Kagan.
Hazara, carboniferous, infra-trias of Sirban Mountain (138), Tanol series, Trias of Hazara.
(139). Carbon trias of Pir Panjal and Dhaoladhér, inliers of limestone in tertiary zone.
Karakoram range, Syringosphaera beds (140), Afgh4nistan and Sulaiman range. Supposed
carboniferous of Tenasserim. Maulmain group (141). Limestone of Tenasserim and
Mergui Archipelago. Unfossiliferous slates of doubtful age; Khwaja Amran range, not
tertiary (143), serpentine intrusions, Trias fossils in Baluchistan ; Arakan Yoma (144), axial
beds, supposed triassic age ; since proved to be nummulitic (144). Negrais rocks, description
(145), serpentine intrusions (146), distribution. Manipur (147), resemblance to rocks of
Simla area. Disang group of upper Assam (148).

CHAPTER VII.

THE GONDWANA SYSTEM.

Origin of name, extra-peninsular outcrops (149), distribution inthe Peninsula. Fluvi-
atile origin (150), possible lacustrine origin of the Talchir group, relations tu older and newer
rocks (151). Connection of outcrops with existing river valleys (152), faulted boun-
daries (153), supposed deposition against cliffs, evidences of disturbance (154). Barrenness
of Gondwana rocks, contrast of lower and upper Gondwana floras (155), classification
of the Gondwana rocks (156). Talchir group, glacial boulder bed (157), glaciated rock
surfaces in Penganga valley, trappoid beds, resistance to weathering (158), flexible
sandstone, fossils, possible lacustrine origin (159), evidence of existence of glaciers.
Karharbari group, separation from Damudas (160), description, conformity with Talchirs,
distribution (161), fossil plants. Damuda series, subdivisions (162), flora (163); Bara-
kar group characteristics, structure of the coal (164), Ball coal, relation to Talchirs;
Ironstone shales (165) ; Raniganj group (166) ; Damudas of Sdtpura area, Motir group,
Bijori group (167), flora, Godavari valley and Chhatisgarh, Kdmthi group (168),
xvi CONTENTS.

description and flora, MAngli beds (169), fossils. Almod group. PAnchet group (170),
fossils (171). Mahadeva series, Pachmarh{ group (172), Denwa and Bagra groups (173),
other exposures of Mahédeva rocks. Dubrdjpur group (174), Rajmahal series, volcanic
rocks (175), foci of eruption, dykes of RAjmahdl age (176), original. extent of volcanic
rocks; interval between Damuda and RAjmahdl perivds; flora of the Rajmahdl group
(177), contrast between Damuda and Raéjmahdl floras; outliers on east coast, Athgarh
basin (178), Ellore area, subdivisions, flora of the Golapilli group (179) Ragavapuram
and Tripetty groups (180), marine fossils and range in time ; - Ongole area, three-fold
division, equivalent to groups of Ellore area (181); Sripermatur and Sattavédu groups;
Trichinopoli area (182); flora of the outliers (183) ; comparison with other groups. Kota-
Maléri groups (184), fauna and flora (185), presence of a Panchet fossil, correlation with
the Denwa group. Chikid4la group (186). Jabalpur group, flora (187), relative age.
Umia group (188), flora of the Umia and Katrol groups ; Kathiawér plant beds (189);
relation of Kathiawar and Umia beds to Jabalpur group (190).

CHAPTER VIII.
HOMOTAXIS CF THE GONDWANA SYSTEM.

Controversy now extinct (191), affinities of the Damuda and Rajmahal flora (192).
Heterogeneous character of the floras, difficulty of determining relationship of fossil plants,
Alliances of the PAnchet flora (193) of the Umia and Jabalpur floras (1¢4), palzontological
contradiction in the Umia group. Affinities of the Gondwana faunas (195), Gondwana
flora in Tongking (196). Gondwanas of Afghanistan (197). Coal measures of Australia,
Bacchus marsh beds, glacial origin, correlated with Talchir beds ; sequence in New South
Wales, marine carboniferous (198), glacial boulder beds, contemporaneous with Bacchus
marsh beds, flora of the Stony creek and Newcastle beds (169), affinities with the Damuda
floras, equivalence of Newcastle and Barakar groups. Hawkesbury group, recurrence of
cold (200), indications of cold in the Panchet group, probable equivalence. “South Africa,
Karoo series (207), characters and distribution (202), classification, Karharbari and Damnda
plants in the Ecca and Beaufort floras, glacial boulder bed in Ecca group (203), reptilian
fauna of the Beaufort beds, Australian facies of Stormberg flora (204), Uitenhage series,
affinities with RAjmahal flora (205). Correlation of the rock groups, equivalence and upper
carboniferous age of the glacial beds and of the Barakar, Beaufort, and Newcastle groups,
permo-carboniferous age (206), Panchet, Stormberg, and Hawkesbury beds, trias. RAjmaha]
group, doubtful age (207), minor uncertainties of the correlation (208), range in time of the
Gondwanasystem. Evidence of a former land connection with Africa (209), close connection
of Gondwana and African floras necessitates land connection (210), evidence of marine
provinces in jurassic and cretaceous periods; bearing on doctrine of permanence of
oceanic areas (211), and onthe constitution of the earth's interior. Probable changes of
latitude, carboniferous glacial beds within the tropics (212), recent evidence of secular
changes of latitude (213), Fisher’s theory of the constitution of the earth (214).

CHAPTER IX,

Marineg Jurassic Rocks.

Absence in peninsular area. Cutch (215), general distribution and subdivision (216),
classification (217); Patcham group (218); Chari group, subdivisions, macrocephalus
beds (219), Dhosa oolite, general distribution (220), relations of the Cephalopoda ; Katrol
group (221), Kantkot sandstone, distribution and relation of the Cephalopoda (222), Umia
CONTENTS. xvii

group, general description (223), palzeontological contradiction between flora and fauna
(224), upper oolitic age. Correspondence of horizons with Europe (225), general results
of the examination of the Cutch Cephalopoda, Jurassic rocks of Western RAjputané
(226), Balmer sandstone, Jaisalmer limestone, Bedesar and Parihar groups (227), Abur
group or Kuchri ammonite bed. Salt range, connection in Cutch jurassics (228). Hima-
layas, Spiti shales, Giumal sandstone, Jurassics of Hazara (229). Doubtful jurassics of
Western Garhwal, Tl beds. (230.)

CHAPTER X.

MARINE CRETACEOUS ROCKS OF THE INDIAN PENINSULA,

History (231), distribution and relation to other rocks (232), classification. Utattr
group, general description (233), coral reefs, littoral character of upper beds, derivation
of sediment from the north (234), fossil wood, distribution, fauna (235), middle cre-
taceous, presence of older forms; Trichinopoli group, general description, granitic
pebbles (236), source of the pebbles, littoral character of deposits, unconformable to Uta-
tur group (237), distribution, fauna, jurassic types (238); Ariyaldr group, possible sub-
division (239), conformity to Trichinopoli group, distribution (240), relation to older and
newer rocks, mode of formation, richness of fauna (241), upper cretaceous age, jurassic
types, abundance of gasteropcda, occurrence of Megulosaurus an oolitic reptile (242).
Richness of South Indian cretaceous fauna (243), paleontological anomalies, creta-
ceous fossils of Sripermatdr (244); distribution of invertebrate fossils, in the groups
(245). Affinities of fauna with Khasi hills cretaceous fauna (246), and with South African
cretaceous fauna (247), former continuity of coast line. Western India, Bagh beds
(248), general description, nodular limestone, Deola and Chirakhan marl, coralline lime-
stone (249), fauna of the groups (250), correlation, cenomanian age, contrast to fauna of
South Indian cretaceous (251), large proportion of European forms, cre‘aceous fauna of
Arabia, contrast with South African fauna (253), a land barrier between the marine prov-
inces. Nimar sandstone, doubtful cretaceous age, possibly Mahadeva. Wadhwan sand-
stones of Kathiawar (253); equivalence with Nimar sandstones, andage. Stratigcaphical.
relation to Deccan trap (254).

 

CHAPTER XI.

Deccan TRape.

Extent (255), name, peculiarities of scenery (256), vegetation, petrology, abundance
of amygdaloids (257), columnar structure, volcanic ash (258), bole beds, mineral constitu-
ents (259), secondary minerals, zeolites (260), horizontality of bedding, thinness of separ-
ate flows (261), sedimentary beds, thickness of the series (262). Lametd group, general
description (263), relations to underlying rocks, and to overlying trap, distribution, rarity
of fossils (264), fauna. Intertrappeans (265), general description (2(6), alternation with
traps, small extent of individual beds, distribution (267), fossils, fresh water origin.
RAjamahendri outcrops (268), infratrappean, fossils, relation cretaceous of Southern Indian
marine origin, doubtful correlation with Lameta group (269), intertrappeans, description
fossils, estuarine origin (270), relationship of fauna. Intertrappeans of Bombay, position
in trap series (271), confined to uppermost beds, description (272) fossils, deposited in

B
RVIT CONTENTS.

marshes (273). Origin of the traps, characteristics of subaqueous eruptions, absence
of stratified ash beds, or marine organisms in the Decean traps (274). Uneven sur-
face of underlying rocks, subzerial erosion of Bagh beds, fresh water erigin of inter-
stratitied sedimentary beds (275), hypothesis of subaqueous origin in a large lake rejected
(276), difficulty of accounting for horizontality. Sources of eruption, evidence of ash
beds, distribution of dykes (277), RAjpipla hills, supposed vents in the Konkan, trachytic
cores (278), distribution of dykes outside trap area (279), lavas flowed far from their
sources. Age of the Deccan trap, unsatisfactory evidence of Rajamahendri beds, inter-
trappean fauna (280), relation to eocene of France and Laramie of America, relation to
cretaceous of Narbadé valley (281) and to tertiary in Surat, Cutch, and Sind, general
conclusion (282). Recapitulation, history of the Deccan trap period (283).

CHAPTER XII.

CRETACEOUS ROCKS OF THE EXTRA PENINSULAR AREA.

Continuity of cretaceous and tertiary deposits, breaks in the succession only local. Isc.
lated exposures of lower cretaceous recks (285), Cutch, Chichdli pass, Hazara. Cretaceous
of Sind (286), Hippurite limestone (287), Cardita beaumonti beds (288), age, volcanic beds.
Baltchistan (289), Hippurite limestone, Belemnite beds, unconformity, Dunghan group (290)
paleontological anomaly of fauna; section near Khelat (291) ; Suldim4n range. Petroleum
(292). Afghanistan, Kandahar section, volcanic beds, basic (293), and syenitic intrusions ,
Turkistdn section. Central Himalayas, Chikkim seties(2¢4). Thibet. Assam, Khasi hills,
coal beds, Chera sandstone (295), original limitation, Garo hills, Jaintia hills (296). Arakan
Yoma, Ma-i group, rocks included and extent. Supposed cretaceous of Tenasserim (297)

CHAPTER XIII.

TERTIARY DEPOSITS (excluding those of the Himdlayas),

Peninsular tertiary confined to west coast. Travancore (299). Ratnagiri. Surat and
Broach, laterite beds (300), nummulites, upper gravelly beds (301, no nummulites. Ter-
tiary of extra-peninsular area, difference between upper and lower. Tertiaries of Sind, classi
fication (302). Ranikot group, description (3¢3) age; Balichistén area, Ghazij group,
Pseudo-conglomerate (302), coal seams, Original limitation ; conformity of tertiary with
cretaceous. Kirthar group in Sind (305), Rohri hills, local unconformity to Rantkot group
change of character to the south (306), relation to Ranfkot, evidence of the echinoderms «
Balichist4n area, Spintangi group, nodular structure, gypsum beds (307); outliers in Western
Rajputana. Nari group in Sind, subdivision (308), possible original limitation, change
of character to the south (309), possible fluviatile origin of upper Nari age. excposires ty
Baluchistan (310). Gaj group in Sind, absence of nummulites, passage inte Manchhar
series, change of character to the south (311), general conformity and local unconformity.
Nari group, fossils (312): Manchhar series, equivalent to Siwdliks, subdivision differ.
ence between lower and upper (313), local unconformity to lower beds, general conformit
marine beds tothe south (314), distribution ; Mekrdn group (315), fossils, probable cee
equivalent of the Manchhars. Fossils of the Manchhar series (316) character of the
vertebrate fauna (317). Balichistan area, unconformity between Siwilil and nummulitic,
Siwaliks of the Quetta valley plains, disturbed river gravels (318), vertebrate, and aberrant
fresh water molluscan, fossils in Bugti hills, Tertiaries of Cutch (310), fala on to Deccan
CONTENTS, XIX

trap, classification (320) sub-nummulitic group 3 gypseous shales ; nummulitic limestone
(321) 3 arenaceous group ; argillaceous group, equivalent to Gj; upper tertiary (322),
Kathidwér, eocene outcrops, miocene beds, upper tertiaries of Perim island(323), Mammalian
fauna, Dw4rk4 beds (324). Afghan Turkistdn. Kohat tertiaries, classification (325), rock
salt and gypsum (326), difference from Salt range Salt Marl, nummulitic beds (327), upper
tertiaries, Eocene volcanic beds of Thal. Assam, Garo hills nummulitic (328), increase
of limestone to the east ; Khasi hills, Cherapunji coal (329), relation to cretaceous fossils
(330), a limit of deposition, easterly extension of nummulitics. Coal measures of upper
Assam (331), probable eocene age, Upper tertiaries, Garo hills, marine beds (332), disturb-
ance, outliers on high ground (333), Na4g& and Manipur hills, upper Assam (33+) Dis-
tribution of upper tertiaries, original extent, special upheaval in Manipur (335). Tertiaries
of Burma, classification, nummuiitics (336), general description (337), coal bed, exposures
west of the Arakan Yoma; Pegu group (338), Sitsyahn shales, Prome beds, volcanic bed,
extent (339) possible cretaccous inliers, possible representative west of Arakan Yoma.
Fossil wood growp (340), fossil wood, vertebrate fossils (341). shark’s teeth, probable fresh
water origin, former extension (342); Upper Burma, volcanic outbursts. Tertiary coal
bearing beds of Tennasserim (313). Andaman and Nicobar islands, Archipelago series
(3 14).

CHAPTER XIV,
TERTIARIES OF THE HimAtayas (tncluding the North-Western Punjab),

Two areas, central and outer Himdlayas. Indus valley, glacial beds (315), doubtful
tertiary age, nummulitic limestone, volcanic beds (346), Original extension, outliers 5
(247), tertiaries of Hundes. Doubtful tertiaries north of Sikkim. Outer Himalayas (348),
classification, Sirmur series, Subathu group (349) ferruginous bottom bed. relation to
older rocks; Dagshdai group (359), Kasauli group, passage from marine to fresh water
conditions ; Subathu inliers of Jammu hills (351) Salt Range tertiaties, nummulitic
limestone, Curditatbeaumonti beds. Hill nummulitic limestoae of Hazara (352), inliers
in newer tertiaries. Correlation of Eocene beds, east and west of the Jehlam (353). Salt
range nummulitics older than Subathu, unconformity at top of Salt range nummulitic lime-
stone, possible equivalence to hil! nummulitic limestoae, older than Subathu (354). Murree
beds, age, mammalian fossils from suposed Miyree beds (355). Siwélik series ; lower
Siwaéliks. Nahan group ‘336), lithological similarity to Dagshai group, distinctness, limi-
tation (357), fossils; middle and upper SiwAliks, subzerial origin of Siwdliks (353) thickness.
Age of the Siwdliks, Molluscan (359) and mammalian fauna,pleistocene and miocene types,
general character of fauna (361), large proportion of recent mixed with miocene genera
(362), evidence of reptiles, fishes, and birds, pliocene age (363); relation to Manchhar fauna
of Sind. Pikermi beds (364), fauna, affinities to Siwdlik, miocene species (365}, pliocene
age. Migration of miocene mammals of Europe to Africa (367), and to India, non-exist-
ence of Himalayas in miocene period. Relation of Siwdlik to recent fauna, disappear~
ance of dominant types (367), influence of the glacial period.

CHAPTER XV.

LaTERITE.

Laterite, importance, order of description (369). composition, character (370), storias
ceous aspect (371), lithomarge, detrital nature, recementation, barrenness (372). High level
and low level laterite, distinction of distribution, no distinction of character (373), difficulty
XX CONTENTS.

of ascertaining detrital or non-detrital origin of laterite. Distribution of high level late-
rite (374), laterite of Rajmahal hills (375), denudation and antiquity of the high level
laterite (376), distribution of low level laterite (378), origin of laterite, chemical change,
transference of iron oxide, hypothesis of origin by alteration of basalt (379), passage of
basalt into laterite, deficiency of iron in basalt (380), absence of amygdules in laterite, only
certain lava flows capable of alteration into laterite, hypothesis not universally applicable
(381), difficulties of supposing sedimentary origin, resistance to denudation, detrital origin
of Bundelkhand laterite (382), former continuity high level laterite of Bundelkhand and
the Deccan. Laterite of Malabar, vesicular variety (383), pellety variety of detrital origin,
vesicular laterite formed in situ. Iron ores of Ulster, hypothesis of origin in marshes by
action of vegetation (334). Summary of theories, definition of laterite, laterite partly formed
by decomposition of underlying rocks, difficulties, large proportion of iron (385), time an
element in its formation, thickness of some patches due to deposition in marshes, climatic
conditions of formation (386), Geological age in part tertiary (387), partly still being
formed. Laterite of east coast, not true laterite, mode of origin, stone implements (388),
post tertiary origin, denudation, ‘antiquity (389).

CHAPTER XVI.
PLEISTOCENE AND RECENT DEPOSITS (exclusive of the Indo-Gangetic alluvium).

Extent, difficulty of separation from tertiary (391), classification, Cuddalore sand-
stones, extent (392), fossil wood, age and origin (393), Warkalli beds, possibly tertiary (394).
Miliolite of Kathidwar. Cave deposits (395). Alluvial deposits, valley plains of Narbada
and TApti, alluvium of eastward flowing rivers (396), Narbadé alluvium (397), occupies a
rock basin, fossil fauna (398), stone implement, relations of fauna (399), fluviatile origin
(400) ; Tapti. and Purna alluvium, salt wells. Godavari alluvial gravels (401), fossils,
Kistna alluvium (402), fossils and flint implements ; diamond gravels; East Coast allu-
vium (403), sub-fossil shells, thickness, lignite at Pondicherri (404), submerged forest
at V4limukam; smooth water anchorages of West Coast (405); alluvium of West
Coast (406), submerged forest at Bombay, coastal alluvium of Narbadé . and
TApti (407); Gujarat and Kathidwar, littoral concrete (403). Lake. deposits. Soils (409),
red soil; Regur (410), composition (411), fertility, distribution (412), theories of origin
(414). Peat. Blown sand (415), Leré or red sandhills. Extra-peninsular area ; hills west
of the Indus (416), disturbed subrecent beds, gravel fans, the Kazez (417), loess, pseudo-
lacustrine deposits. Potwar alluvium, erratic blocks (418), raised Indus gravels, floods of
the, Indus (419), fossils in alluvium. Himdlayas; Kashmir, Karewas (420); Nepal (421).
Hunde, mammalian fauna (422). Eastern hil!s, Manipur, upper Chiadwin valley (423).
Burma, Engdain tract cf older alluvium (424) Irawadi delta (425.) ,

bis

CHAPTER XVII.

Tae Inpvo-GANGETIC PLAIN.

Extent (427), unity of Gangetic and Indus plains, no evidence of marine conditions in
Upper India (429), probable recent presence of sea in Sind (430), older and newer allu-
vium (431). Thickness of alluvium (432), Calcutta boring, pebbles, peat bed (433); Borings
at Lucknow (434), Agra, Umballa, Sabsalkot (435). Fossils in Jumna alluvium,
Explanation of the terms kankar (436), bhibar (437), tardt, bhangar, khddar, bhur (438),
CONTENTS. XX1

Alluvium of the Brahma putra in Assam (434). Delta of the Ganges and Brahmaputra,
windings of rivers (4!0', backwardness of eastern delta, change of course of the
Brahmaputra (441), ‘swatch of no ground.’ The Mddhupur jungle (442), age of the
Ganges delta, former continuity of Gangetic and Punjab rivers (443). Alluvium of
Upper India (444), cause of distinction between bydngar and kha lar (445), course of the
Ganges (440), changes of level. Re or Kalar, origin of Reh salts (447), cause of con-
centration, influence of canals. Salt wells (448). Alluvium of the Punjab; lost river of
the Indian desert (449), former independent course of the Sutlej, the Saraswati of the
Vedas (450); recent origin of the Khadar. Alluvium of the Indus in Sind, passage of the
Indus through hills at Sukkur (45!), changes in the course of the Indus. Indus delta
(452). Rann of Cutch (453), changes of level (454). Desert of western Rajputana, two
types of sandhills (455), longitudinal type (456), origin, derivation of the sand (457).
Other tracts of sandhills (458).

CHAPTER XVIII.

THE AGE AND ORIGIN OF THE HIMALAYAS.

Geographical limitation of the Himalayas, views regarding their orography (459), moun-
tain chains of the north-west termination (460!, main range probatly an aggregate of
Separate chains. Three orographical regions, Tibet (461), snowy range and lower Himé-
layas, Sub-Himalayas, lesser elevation due to lesser upheaval (462). Drainage system,
rivers cross main range, cause, cutting back of transverse valleys (463), direct evidence.
Stratigraphical zones correspond ‘with orographical (464). Tertiaries of the sub Hima-
layas, distribution (465), local conformities and unconformities (466), formed during disturb-
ance; the main boundary fault (467) an original limit of deposition (458) ; sub-montane
recent deposits, relation to present river valleys, a Similar relation in upper SiwAliks
(469); Himalayas existed in pliocene times. Siwaliks deposited in an area of subsidence,
encroachment of hills on area of deposition (471). Fisher's theory of m ountain formation,
modification in the case of the Himalayas, agreement with observed facts (472). Theory
tested, evidence of the borings (476). Himalayas non-existent in eocene period (477),
absence of conglomerates in lower Siw4liks (478), tertiary age of the Himalayas, evidence
of the Tibetan fauna (47y). Opposing views; greater antiquity, Middlemiss, successive
Loundaries of the Sub-Himdlayas (481). comparative disturbance of tertiaries and older
1ocks, transverse str ke (482), definition of the age of the Himalayas, disturbance in Hima-
layan rocks partly pre-Himdlayan, recent origin, Howorth (483), supposed absence of
glaciation, reason for abseuce of ice sheet (484), post tertiary increase of height, differential
movements of surface, evidence of Hundes pleistocene fossils (485), desiccation of Tibetan
lakes, o-igin of Tibetan lakes (486), Himalayas at their maximum development. Sup-
posed coanection between sedimentation and mountain ranges (487), sedimentary deposits
cf the north-west Him4la’as, absence in Eastern Himalayas (438), divergence in two
areas, position aad elevation of the Himdlazas not due to sedimentation (459).

CHAPTER XIX.

GEOLOGICAL HISTORY OF THE INDIAN PENINSULA.

References to changes of land and Sea in previous chapters. Earliest periods, ob-
scuity, an ancient land surface, changes of land and Sea (490), none of the main features
xxii CONTENTS.

of existing geography pre Cuddapah. Vindhyan epoch, rise of Ardvallis, other contem-
poraneous hill ranges (491), origin of East Coast. Silurian seas in extra-Peninsular
area (492). Carboniferous and Permian, glacial period, peninsula part of an Indo-African
continent, cretaceous land connection with Africa (493), Deccan trap eruptions. Tertiary,
break up of Indo-African land connectioa, encroachment of land oa sea (494). Origin of
the Western Ghats movements of elevation, antiquity of the Peninsular drainage system,
a possible case of reversal of drainage. Western Ghats the most recent and conspicuous
feature of the Peninsula (495).

LIST OF ILLUSTRATIONS.

Simvo, 22,300 feet, from the north-west (eastern spur of Kanchanjanga) . Frontis piece.

Hill Ranges of India » : ‘ . . j . . - toface page 8
Volcanoes of Burma and the Bay of Benga . H 7 + . ss » +20
Permo carboniferous and Permian (Productus lin-stone)fossils. Two plates . in » 126
Triassic fossils . . . . . . i . ; 3 ‘ ‘i » 130
Talchir-Karharbéri plants . ‘i * . ‘ ‘ . ‘ , fi » 158
Damuda plants. Two plates . . . . ‘i < ‘ 4 2 i » 162
PAnchet fossils 5 i . : ‘ ‘ . ‘ . é ‘ 4 » 170
Rajmahdl plants. Two plates . 7 . . . . . . . ” » 176
Jabalpur fossil plants é . . . . . . . . . ‘i » 186
Cutch (Umia) fossil plants < : < 5 . . : ‘ - i » 190
Jurassic fossils . . . . . . 7 ; ‘ : . y » 224
Cretaceous fossils . . ° ; . . . 3 . * + » «242
Intertrappean fossils . ‘ 2 : : é 3 é ; 5 i » 274
Miocene fossils . : ‘ ’ . . . - . c ” » 314
Siwélik fossils, Two plates . . s . - ‘ . . 3 3 1, 362

Map of the Indo-Gangetic alluvium . ‘ ‘ 7 ° ‘ ‘i ‘ a » 458
Sections across the Himdiayas and the Sub-Himdlayan Zone. 4 : 5 » «6482
Map of the Himalaya mountains. . . . . ‘ - End of Chapter XIX,
Geological Map of India . . . . . . . . . End of volume

WOODCUTS IN TEXT,

Fig. 1. Pagoda on the sea shore near Tranquebar. . a : : . Page 12
Fig. 2. Part of Tinnevelli coast showing present cosition of old sea ports. . i 13
Fig. 3. Mica mines at Inikurti. . . . . . A . 5 . 5 34
‘Fig. 4. Section showing the relation between the Gwalior and Vindhyan

systems at the junction of the P4r and Vindhyan Scarps. . . » 67
Fig. 5. Sketch section illustrating the relation of the Cuddapah and Karntl rocks,

after King . . . . . . . . . . ” 84
Fig. 6. Conularia Warthi, Waag. . ‘ s : . é ‘ : 3 » 120
Fig. 7. Productus semireticulatus, Mart. ‘ . 5 . ‘ ‘ | «E24
Fig. 8. Section of Richthofenia laurenciana, Kon, . ee ak 5 : » «125
Fig. 9. Daonella ( .alebia) lommeli Wissm , , , : ji j j 130
Fig. 10. Syringosphera verrucosa, Duncan. , i = i 140
Fig. 11. Estheria mangaliensis, Rupert Jones . ‘ : : . : : : 170

Fig. 12. Radiating c .lumnar trap, Rajmahdl hills, 8 . . : 5 » 195

Fig. 13. Geological Sketch of South Africa ee, “ . ; : . saa
Fig. 14. Trigonia ventricosa, Krauss 3 ; i. ios
: : ‘ : ‘ : é P i » 22

Fig. 1§. Hill composed of Deccan trap, near Harngaon, north of Nimawar, Nar-
bada valley ; : wig

 
Fig.

Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.

Fig.
Fig.

Fig.
Fig.

17.
18,
19.
20.
21.
22,
23.

24.
25.

26.
27.

CONTENTS.

Radiating basaltic columns in a dyke near Gijri, north-west of Maheswar,
Narbadé valley . . . ei : ‘
Cardita beaumonti, D'arch et Haime . * . :
Hill of rock salt, at Bahddur Khel, after Wynne . ;
Section through the Bahadur Khel salt locality, after Wine, .

Section across the Gauli plate, after Foote : j ; oe ™
Diagram illustrating the theory of the Karez . : , 3 7
‘Diagram illustrating the relations of Bhdbar and Tardi z oot

Sandhills of the transverse type, after the topographical —— of
Rajputdna 6 . ‘ -

Sandhills of the fonsteadtwal type after ‘the Sind Revenue Sarvey 5 .

Section at the head of the Sara valley, eastern Kumdun, showing overlap
of the main boundary fault by the upper Siwaliks . .

Diagram to illustrate Rev. O. Fisher's theory of mountain fomaaties

Diagram to illustrate the theory of the elevation of the Himdlayas . .

XXili

page 258

288
326
327
382
47
438

456
456

468
472
473
A MANUAL

OF

THE GEOLOGY OF INDIA

CHAPTER I.

PHYSICAL GEOGRAPHY.

Scope of the work—Threefuld division of British India—Contrast between extra-peninsular and
peninsular areas—Mountain ranges of the Peninsula—Extra-peninsular mountain ranges
—Drainage of extra-peninsular ranges—River system of the Peninsula—Evidence of
changes of level in the Peninsula—Changes of coast line—Glacial epoch in India—
Volcanoes—Doubtful cases of volcanic action—Salses, or mud volcanoes.

HE limits of the area, whose geology is treated of in this book, coincide

with the limits of the jurisdiction of the Governor-General of India. In
some few cases references to the geology of adjoining countries will be-
found, but such are not many, and are all imperfect as they depend on
observations which were made during hurried. traverses and under cir-
cumstances precluding the idea of detailed geological work. In spite of
these drawbacks the results have often been important, interesting and
impossible to ignore, but there is generally less reason to congratulate
ourselves on the knowledge obtained of countries across the border than
to deplore our ignorance of large tracts within it.

The general shape and principal features of British India, the great
triangular promontory with the pear-shaped island of Ceylon south-east of
its extremity, the great range of the Himalayas to the north, and the large
area of Burma to the east, running down into the narrow strip of Tenasserim
along the east coast of the Bay of Bengal, are well enough known, as well
as the principal political divisions of the empire. But, for geological put-
poses, the important point to be noticed is the chreefold division of this
area into, rst, the great Indo- Gangetic alluvial plain, comprising the Punjab

B
2 GEOLOGY OF INDIA—PHYSICAL GEOGRAPHY. (Chap. J.

and Hindustan proper, with Bengal and the eastern prolongations up the
valleys of the Brabmaputra and the Barak ; 27d, the triangular area of the
Peninsula, lying to the south ; and 9rd, the extra-peninsular area, compris-
ing the hilly country west, north, and east of the Indo-Gangetic plain.

Nor is this division an arbitrary one. The geological history of the pen-
insular and extra-peninsular areas has been radically different. Since
the latter end of the palzozoic era the former appears to have been an
area of dry land; no sedimentary formations of marine origin have been
found except near the present sea coasts, and there they thin out against
the older rocks on which they rest, in a manner suggesting that the shore
line cannot have been very far removed from the present position of the
coast when they were being deposited. In the extra-peninsular area, on
the other hand, marine deposits range through the palzozoic and mezozoic
eras, and only in the latter part of the terliary period is there any great
development of deposits formed on dry land.

Structurally too the two areas differ greatly. The Peninsula has under-
gone no great compression since the close of the palaozoic era, and the
beds all lie at low angles of dip. In the extra-peninsular area the con-
ditions are totally different; the rocks have everywhere undergone great
compression and disturbance since the commencement of the tertiary
period, a disturbance which ranges in degree from the comparatively
regular, though high dipping, folds of the Balachistén and Punjab hills, to
the complicated overfolds and thrust faults of the Himalayas.

This difference in geological history finds its expression in the differ-
ence of the present contours of the two areas. In the extra-peninsular
area we have mountain ranges which coincide with regions of special eleva-
uion, that is, the. courses of the principal chains, and often of the minor
ridges, are governed by their structure and are the direct result of the
compression, and consequent disturbance and elevation they have under-
gone. Asa result of this, the valleys are deep, narrow, and steep-sided,
the rivers and streams rapid and torrential in their nature, and, as a
rule, evidently actively at work in deepening their valleys. In the penin-
sular area, on the other hand, the mountains are all remnants of large
table-lands, out of which the valleys and low lands have been carved. The
valleys, with a few local exceptions, are broad and open, the gradients of
the rivers low, and the whole surface of the country presents the gently
undulating aspect characteristic of an ancient land surface.

Such, broadly speaking and subject to some minor exceptions, are the
contrasting characteristics of the two areas. In the country lying west
of the Ardvallis, between them and the Indus, there is a tract of geo-
graphically debateable ground, which exhibits a combination of the
characteristics of the two areas. The rocks exposed are very largely
secondary and tertiary beds of marine origin, agreeing in this with those of
Chap. 1] MOUNTAIN RANGES OF THE PENINSULA, 3

the extra-peninsular area, while in their low undulating dips and absence
of any marked degree of disturbance, they approach the type of the penin-
sular area. On the north-east again beds, belonging to formations which
are characteristically peninsular, are found in the Himalayas of Sikkim
and north of the Assam valley and in the hills intervening between the
Brahmaputra and Bérak rivers. We will find the explanation of these ex-
ceptions to the geological contrast between the two areas in the great
structural disturbances which took place during the tertiary period, and
profoundly modified the outlines of that ancient land surface of which
the Peninsula proper is but a remnant.

The nomenclature of Indian mountain ranges is still a difficulty, it being
a rare exception that any definite term is applied to a mountain chain,
throughout its extent, by the people ofthe country. In many parts of India
peaks and passes have names, but the ranges have none and, even if names
exist, their application is not unfrequently vague. Thus, the ancient name
of ‘ Vindhya,’ applied to the hills separating Hindustan proper or the Gan-
getic country from the Deccan (Dakshin or south), has now, by common
consent, been restricted to the hills north of the Narbadd, but it appears
almost certain that the term originally applied also to the ranges now known
as Sdtpura, south of the river, and it is very probable that the latter hills were
more especially indicated by the term “ Vindhya” than the former. The
term “Sdtpura ” again was of very indefinite application and probably in-
cluded other ranges besides that to which it is now restricted. The names
here applied are those employed by the latest writers on Indian geography,
but some of them are by no means generally adopted on maps.

The most important mountain ranges of the Peninsula are the Sahy4dri,
or Western Ghats, running along the western coast from the Tapti river to
Cape Comorin, atthe southern extremity of the Peninsula; the Sdtpura,
running east and west on the south side of the Narbadd valley, and
dividing it from the drainage areas of the Tdpti to the westward, and the
God4vari to the eastward; and the Aravalli, striking nearly south-west to
north-east, in Rdjputdna. The so-called Vindhyan range, north of the
Narbad4, and the eastern continuation of the same north of the Son valley,
known as the Kéimur range, are merely the southern scarps of the Vindhyan
plateau comprising Indore, Bhopal, Bundelkhand, etc. The ‘plateaux of
Hazdribagh and Chutid Nagpur (Chota Nagpore) in south-western Bengal
appear to form a continuation to the eastward of the Sdtpura range, but
there is no real connection between these elevations and the Sdtpura chain.
They are formed of different rocks and there is no similarity in the geo-
logical history of the two areas, so far as itis known. In many maps a
range of mountains is shown along the eastern coast of the Peninsula, and
called the Eastern Ghats. This chain has not the same unity of structure
or outline as the Western Ghats. It is composed to the southward of the

B 2
4 GEOLOGY OF INDIA- PHYSICAL GROGRAPHY. (Chap. I.

south-eastern scarp of the south Mysore plateau, on the east of the Yella-
konda range along the eastern margin of the Cuddapah transition basin,
and further north of the south-eastern scarp of the Bastdr- Jaipur pla-
teau, north-west of Vizagapatam, and of several short isolated ridges of
metamorphic rocks, separated from each other by broad plains and having
in reality but little connection with each other. There are also several
minor ranges, such as the Rajmahdi hills in western Bengal, the Indhyddri
between the TApti and Godavari, the Nallamalai (Nullamullay) near Cud-
dapah, north-west of Madras, and the little metamorphic plateaux, such
as the Shevaroys, Pachamalai, etc., scattered over the low country of
the Carnatic, south-west of Madras.

The peculiarity of all the main dividing ranges of India is that they
are merely plateaux, or portions of plateaux, which have escaped denudation.
There is not throughout the length and breadth of the Peninsula, with the
possible exception of the Ardvalli, a single great range of mountains
that coincides with a definite axis of elevation, not one, with the exception
quoted, is along an anticlinal or synclinal ridge. Peninsular India is, in
fact, a table-land worn away by sub-aerial denudation, perhaps to a
minor extent on its margins by the sea, and the mountain chains are
merely the dividing lines, left undenuded between different drainage areas,
The Sahyddri range, the most important of all, consists to the northward
of horizontal or nearly horizontal strata of basalt and similar rocks, cut
into a steep scarp on the western side by denudation, and similarly eroded,
though less abruptly, to the eastward. The highest summits, such as
Mahdbaleshwar (4,540 feet) are perfectly flat-topped, and are’clearly un-
denuded remnants of a great elevated plain. South of about 16° north
latitude, the horizontal igneous rocks disappear, the range is composed
of ancient metamorphic strata, and here there is, in some places, a distinct
connection between the strike of the foliation and the direction of the hills,
but still the connection is only local and the dividing range consists either
of the western scarp of the Mysore plateau, or of isolated hill groups,
apparently owing their form to denudation. Where the rocks are so
ancient as are those that form all the southern portion of the Sahyddri, it is
almost impossible to say how far the original direction of the range is due
to axes of disturbance ; but the fact that all the principal elevations, such
as the Nilgiris (Neilgherries), Palnis (Pulneys), etc., some peaks of which
rise to over 8,000 feet, are plateaux, and not ridges, tends to show that
denudation has played the principal share in determining their contour.

The southern portion of the Sahyddri range is entirely separated from
the remainder by a broad gap, through which the railway from Madras to
Beypur passes west of Coimbatore. The Anamalai, Palni, and Travan-
core hills, south of this gap, and the Shevaroy and many other hill groups
scattered over the Carnatic, may be remnants of a table-land once united
Chap. J.] MOUNTAIN RANGES OF THE PENINSULA. 5

to the Mysore plateau, but separated from it and from each other by
ancient marine denudation. Except the peculiar form of the hills, there
is but little in favour of this view, but on the other hand there is nothing
to indicate that the hill groups of the Carnatic and Travancore are areas
of special elevation.

The whole of the western Sdtpuras, from their western termination in
the Rdjpipla hills to Asfrgarh, consist of basaltic traps, like the Sahyddri.
It is true the bedding is not horizontal, but the dips are low and irre-
gular, and have no marked connection with the direction of the range.
The central SAtpuras, comprising the Pachmarhi or Mahddeva hills, from
the gap in the range at Asirgarh to near Narsinghpur, are composed chiefly
of horizontal, or nearly horizontal, traps, but partly of sandstones and of
metamorphic rocks, and there is here again, as in the southern Sahyddri,
some connection between the strike of the foliation in the latter and the
direction of the ranges. The highest peaks, however—those of Pachmarhi
(4,380 feet)—are of horizontal mesozoic sandstones. Farther east still
the SAtpuras consist entirely of horizontal traps, terminating in the plateau
of Amarkantak, east of Mandld4. East of this plateau there is, north of
Bildspur, a broad expanse of undulating ground at a lower level, and
farther to the eastward again rises the metamorphic plateau of Chutid
NA4gpur, capped in places by masses of horizontal trap and laterite. ‘These
formations were apparently once continuous, across the low ground near
Bil4spur, with the same strata on an equal elevation at Amarkantak,
Similar outliers occur on the Bundelkhand plateau, north of the Narbadé,
all tending to the same -conclusion—that the low valleys of central India
are merely denudation hollows, cut by rain and rivers out of the original
plateau of the Peninsula. The chief exceptions to this lavw—the instances
in which the strike and dip of the rocks appear to have produced import-
ant effects on the contour of the country—are to be found amongst the
metamorphic and transition formations.

If is true that some small ridges are formed of azoic and mesozoic sand-
stones, in places where the beds of these systems have been disturbed,
but the only important lines of disturbance in either appear to be due to
older axes of metamorphic foliation, and it is a rare case to find that the
strike of the sandstones appears to have much effect upon the directions of
the hills and valleys. A possible exception occurs in the Damodar valley
in Bengal, but even this is a disputed case, and the subject will be dis-
cussed in the chapters relating to the Gondwana system.

This remarkable absence in the Indian Peninsula of.any evidence of
disturbance in late geological times—a feature which abruptly distinguishes
the whole area from the remainder of Asia—will be further noticed in the
sequel; at present it is sufficient to remember that the principal mountain
chains of the Indian Peninsula are, with one exception, not coincident with
6 GEOLOGY OF INDIA—PHYSICAi GEOGRAPHY. (Chap. I.

axes of disturbance or elevation, and to note the contrast in the extra
peninsular area. ;

The Aravalli differs from the other great ranges of India in being
entirely composed of disturbed rocks, with the axes of disturbance cor-
responding with the direction of the chain, The formations found in the
Arévalli range belong to the transition rocks, and are of great antiquity ;
for the most part they are much altered, they are quite unfossiliferous,
and there is evidence which renders it probable that the elevation of the
range dates from a period anterior to the deposition of the Vindhyan
rocks, themselves of unknown age but almost certainly not of later date
than carboniferous, whilst the fact that these Vindhyan rocks are found
almost horizontal in the neighbourhood of the Ardvalli range, on both
sides of the chain, shows that here, as elsewhere in the Peninsula, the
forces which have affected the extra-peninsular area in Jater geological
epochs have not beer. felt.

Passing to the other side of the Indo-Gangetic plain—no matter whether
the region reached be to the westward in Sind and the Punjab, to the
northward in the Himdlayas, or to the eastward in Chittagong and Burma
—the mountain ranges, with the exception of the Salt-range and the Assam
range, are everywhere composed of disturbed and contorted beds, and the
disturbance has invariably affected rocks of late geological age. The
amount of alteration may be small or great, the hills may consist of simple
anticlinal folds as in Sind, or of the most complicated inversions as in parts
of the Himalayas, the strike of the bedding may vary from east and west
to north and south, but two characters are constant—great disturbance
affecting all the tormations, and the coincidence of the direction of the
ranges with synclinal and anticlinal axes.

‘The nomenclature of extra-peninsular mountain ranges is compassed
with the same difficulty as those of the Peninsula, owing to the absence of
local names except for individual ridges, peaks or passes, and has been
further complicated by a want of unanimity among geographers as to the
true limit of the term ‘mountain range.’ Different geographers have recog-
nised from two to seven distinct ranges in that great system of mountains,
collectively known as the HimAlayas, which rises to the north of the Indo-
Gangetic plain, and the opinions regarding the true western limit of the
principal range have varied from that which regards it as ending in the
Simla spur to that which looks on it as continuous with the Hindu Kush,

These contradictory Opinions are all more or less correct, according
to the limited point of view of the individual author, but if we look below
the accidents of surface contour to the underground structure of this great
mass of hills, we find in their geological structure and composition that they

owe their elevation to a great series of earth movements, which must be
Chap. J EXTRA-PENINSULAR MOUNTAIN RANGES. 7

regarded as_a single and continuous system of disturbance, and this
. Structural unity is now generally held to unite the separate chains into
a single system of mountains to which the term of ‘range’ is inapplicable,
unless we give it a wider application than is usual. The details of the
physical geography of this great system of mountains, which stretches from
the Indus to the Brahmaputra, have so important a beariut on the history
of its elevation that they will be deferred to a subsequent chapter.

At its north-western extremity the great snowy range of the
Himdlayas bends round into the Hindu Kush, which runs south-westwards
along the southern side of the upper Oxus valley. On their southera
boundary the strike bends round to southwards at the valley of the Jeh-
lam, and from this termination of the Himdlayas proper there extends
through Afghdnistén and Balichistén a complicated series of hill ranges ot
whose detailed geography very little is known.

The proper nomenclature and classification of these hill ranges is a dif-
ficult matter to determine, and, in the present state of our knowledge, both
geographical and geological, it is impossible to arrive at a satisfactory result.
An attempt-has been made to classify these ranges according to the “sys-
tem of disturbance” they belong to, which would hardly need mention
were it not for the eminence of its author and the nature of the conclu-
sions arrived at. In his great work ‘Das Antlitz der Erde,” Prof. Suess
has regarded the Salt-range of the Punjab as forming part of the foot hills
of the Hindu Kush, and has united all the ranges between ;it and what is
known to geographers as the Hindu Kush into one range, on the ground
that they belong to the ‘‘ Hindu Kush system of disturbance.” This is
still a matter for proof. We have no knowledge that there is a unity in the
disturbance of the rocks composing these hills, comparable to that seen in
the Himdlayas, and until such knowledge is forthcoming it would be use-
less to adopt a system of nomenclature so opposed to that ordinarily cur-
rent, and so certain to lead to needless confusion. As regards the Salt-
range we shall shortly see that it ought to be classed by itself, and the
individuality of the Safed Koh range, as depicted on our maps, is so marked
that it is difficult to believe that it has not alsoan individuality of structure,

In one sense the whole of the ranges west of the Indus may be classed
together, for there can be little doubt that they were contemporaneously
elevated, and that the greater part of this upheaval, if not the whole,
took place within the tertiary era. But, structurally, they may be divided
into two classes, whose strike is about north and south and east and west,
respectively, and these two alternate with, and pass into, each other in
a manner that is at present not understood.

The most northerly range is the Safed Koh which runs eastwards from
the neighbourhood of Kabul, forming the south side of the valley of the
Kabul river, till it ends in British territory. Further south the Sulfi-nda
8 GEOLOGY OF INDIA—PHYSICAL GEOGRAPHY. (Chap. I.

range is found striking north and south along the western frontier, and at
its southern limit this bends round into the east and west running hills of
the Bugti country, a strike which extends to near Quetta. The hills again
take a southerly bend, and the Brahuik and Kirthar ranges run north and
south. Further west the strike again changes, and in western Balichistan
the ranges run east and west.

At their eastern extremity the Himdlayas are met, in a manner that
has not been worked out, by a series of hills which at first strike south-
westwards, afterwards bending round to a more southerly direction in
the N4g4 hills, where the principal ridge is known as the Patkoi. They run
through the Manipur country southwards, till they are continuous with the
range of hills, known as the Arakan Yoma, which lies between the Irawadi
valley and the Bay of Bengal.

Besides these principal systems of hill ranges there are two minor ones
to be noticed which, despite a total difference of the rocks-they are.com-
posed of, show a considerable similarity in their structure and geographical
position, subtending as they do the angles between the southern margin
of the Him4layas and the hills which meet them at either extremity. The
first of these is the Salt-range in the Punjab, the second that set of hills,
called for convenience the Assam range, inhabited by the Géro, Khasi
and Jaintia tribes. Jn both cases the hills are composed of a plateau
with a steeply scarped face to the south, along which there is an axis of
abrupt folding, accompanied by more or less faulting. This similarity of
structure and position, in spite of great difference in the rocks of which
they are composed, would seem to show that their elevation is a direct
result of the same great series of movements of the earth’s crust which
resulted in the elevation of the hill ranges forming the extra-peninsular
limits of the Indo-Gangetic al!uvium, but they cannot be regarded as belong.
ing to any of the mountain systems whose re-entering angles they subtend.

To the east of the Irawadi valley there is a great series of mountain
chains, stretching southwards through the Shan states tillit terminates in
the Malay peninsula, of whose geography and geology even less is known
than of the hills beyond the western frontier of India.

The drainage system of the extra-peninsular hills everywhere shows the
peculiarity that, though the valleys often run along the strike of the hills for
long distances, the streams and rivers always sooner or later break across
the axes of maximum elevation. This peculiarity is very noticeable in the
case of the Himdlayas, whose river system must be deferred to a future
chapter, but is in some respects more conspicuous in the case of the hills
west of the Indus.

Here the ridges are mostly formed by the hard limestone cores of anti-
clinal folds through which the streams flow in narrow precipitous-sided
 

 

 

 

 

a

rae
Ag

«

%,

VAHYVS

’

o |»
oe aun Pe
wa ao ot

S
NM

fk

 

 

 

 

 

 

Ay
4
ry

My rom werd
Ms seule,

    

oo

  

 

 

 

  

 

 

 
  

 

 

 

 

 

 

 

 

 

 

 

 

y = A
x > oye Mle,
D “ r ext ee
ya 2 WA
v

HILL RANGES OF

INDIA.

———

 

Scale of English Miles

2 50190 zoe 200

 

 

 

ued > ;
:
$ aut
o a
|
\
: g
3
* ar 5 1
% al
8 Pp
x7 75
“sf
° BR
4 &
° i:
z
ea)
ie
i]
STi
o4
Z.

Andaman I¥ / v

=
?

 

 

 

 

 
Chap. 1.) EATRA-PENINSULAR DRAINAGE SYSTEM. 9

gorges, locally known as ‘tangi.’ ‘They form a very characteristic feature
on the scenery of the hill country west of the Indus, and are commonly
attributed to some disruptive force, which opened a way for the stream to
flow through the mountain. Such is not, however, the true explanation, not
only can it be disproved in many cases by the continuity of the solid rock
across the stream bed, but it is inadequate, as it fails to account for the
broad open valleys invariably found above the gorges. The fact is that they
have been gradually cut down by the streams that flow through them, and
the same absence of vegetation which increases the erosive power of rain
wash on soft clays enabling broad valleys to be formed where they are
exposed, diminishes its action on the hard limestones, at the same time it
reduces chemical action to a minimum, and the absence of moisture almost
entirely deprives the night frosts of their power to disintegrate the rock,
As a consequence, the steep sides of the ravines cut by the streams, where
they meet with compact rock, remain standing almost perpendicular, while
in the intervening stretches of soft clay the valley widens out.

It must not be supposed that the whole country was elevated to the
height of the crests of the ridges through which these “tangis”’ wete cut.
It will be shown in the sequel that the compression, contortion and con-
sequent elevation of the hills was taking place at the same time that the
valleys were being excavated, but sometimes the rate of elevation was too
great for the streams, and areas of closed drainage were formed in which
extensive alluvial and xolian deposits have been accumulated. These are
particularly common in Baltchistdn, where they are usually occupied by a
broad expanse of wind-blown loess.

In the eastern hills the same features are to be seen, but, owing to the
greater rainfall and dense vegetation, as well as to the different type of rock
forming the hills, there is not the same abrupt alternation of broad open valley
with deep and narrow gorges as on the west. The chemical action of the
humic acids developed in the jungle-clad svil has smoothed off the steep-
ness of the sides of the gorges, while the vegetation has protected the softer
clays from being so easily washed away. As another result of the more
rapid erosion of the stream beds there are no areas of closed drainage, but
there are broad alluvial valley plains, such as those of Manipur and the Kubo
valley, where differential movements of elevation of the beds of the streams
have checked their velocity and compelled them to deposit their solid burden.

The river system of the Peninsula, omitting the drainage into the
Ganges and small streams flowing to the west coast, is nearly all-taken
by six large rivers, of which two, the Narbadé and ‘apti, drain the
north-western portion and escape into the Gulf of Cambay, while the
drainage of all the rest of the Peninsula, even from the crests of the
Western Ghats within sight of the sea, flows eastwards by four great deltaic
10 GEOLOGY OF INDIA~PHYSICAL GEOGR \PHY. (Chap. I.

rivers, the Mahdnadf, Godavari, Kistna, and Cauvery,—the only other
streams of any importance being the northern and southern Penner.

This easterly trend of the drainage is probably of very ancient date, as
there are patches of littoral marine deposits along the east coast, ranging
as far back as the close of the jurassic period, which show that, since that
period at least, the eastern coast cf the Peninsula has maintained very much
its present position. On the west coast no marine sediments older
than the upper tertiary are known, if we except the cretaceous beds of the
lower Narbadd valley. At the close of the Deccan trap period, that is the
commencement of the tertiary era, dry land must have extended consider-
ably west of the coast line; south of the trap area the evidence is only
negative, but the absence of any large valleys draining in this direction
suggests that the present position of the shore line is of more recent origin
than that of the east coast, and that the earth movements which gave rise to
it were either too slow, or more probably not of a nature, to change the
easterly course of the drainage.

There can be no doubt that, beyond the limits of the Peninsula, there have
been very great changes in the distribution of land and sea since the com-
mencement of the tertiary era and, even in the latest part of it, the great
disturbances which the rocks have undergone must have been accompanied
by great changes of shore line. But when we come to the post-tertiary period
and enquire whether, on the whole, there has been elevation or subsidence,
the evidence is contradictory. In the alluvium of the Gangetic delta, and
near Pondicherri, beds of peat, at various levels below the surface of the
ground, show that there has been subsidence, but this is the usual, if not in-
variable, condition in a delta, and it is more than probable that al! the large
deltas along the coast are being gradually depressed.

Along the non-deltaic portions of the coast evidences of sub-recent eleva-
tion are found in coral reefs and marine deposits raised above the present
level of the sea. The low level laterite of the east coast lies on a gentle
slope of the older rocks, unaffected by subaerial erosion, such as is formed
by the sea, and must have been deposited either before or shortly after this
was raised above sea-level.

The escarpment of the Sahyddri range—a remarkable feature of the hills
parallel to the western coast of the Peninsula—has frequently been noticed
as furnishing evidence of arise of land. Throughout the trap country of the
Bombay presidency, the Western Ghats rise from the Konkan in an almost
unbroken wall, varying in height from 2,000 to 4,000 feet, cut back in places
by streams, projecting here and there into long promontories, but preserv-
ing throughout a singular resemblance to sea cliffs. This resemblance,
however ceases to a great extent to the southward, where the metamorphic
rocks replace the hotizontal basaltic traps. The escarpments of the Mdlw4
plateau, north of the Narbadd, and of the Deccan plateau, south of
Chap. I.) CHANGES OF LEVEL IN THE PENINSULA. i

Khandesh, although far inferior in elevation to the scarp of the Sahyddri,
resemble the latter too closely in appearance to justify the assumption,
without further evidence, that the cliffs of the Western Ghats are of marine
origin, The parallelism of the Sahyddri escarpment to the sea-coast is
suggestive of a connection between the two, and this connection is strength-
ened by the facts that a thickness of at least 4,000 feet of bedded trap has
been removed from the surface of the Bombay Konkan, and that the plane
of marine denudation, already mentioned as supporting the low level laterite,
extends in places nearly to the foot of the scarp. The circumstance that
the hills of the Sahyddri are inhabited by certain fresh-water mollusca
belonging to the genus Cremnoconchus, which is unknown elsewhere
and is so closely allied to Indian forms of the littoral marine genus
Littorina as to render it probable that both are descended from the same
ancestors, also tends to strengthen the view that the Sahyddri mountains
were formerly washed by the sea. But it is certain that great denudation
has taken place since the scarp was a sea cliff, and it is far from improb-
able that, if the sea ever extended to the base of the Western Ghats, the
epoch belonged rather to tertiary than post-tertiary times. It is also
possible that the isolation of the different hill ranges of Southern India,
and the denudation of the Palghat Gap, south of the Nilgiri plateau, are
due in part to ancient marine action of the same date as the formation of
the Sahyadri escarpment. In this case, as in so many others connected
with Indian geology, all that is now possible is to suggest probable inter-
pretations of phenomena, and to leave them for future exploration to
confirm or contradict.

On the other hand, a sudden deepening of the sea, at a distance of
10 to 20 miles from the shore, along the Mekrdn coast, has been sup-
posed (o represent a submerged cliff. More positive evidence of recent
subsidence is to be found in the occurrence of a number of trees
imbedded of mud, in the spot where they grew, at a depth of 12 feet
below low-water mark on the east side of the island of Bombay,! and in
the submerged forest at the western end of the Valimukam bay on the
Tinnevelli coast described by Mr. Foote.®

Local alterations of level, accompanied by earthquakes, are known to
have occurred on at least one occasion, namely the great earthquake of
Cutch in 1819, when a considerable area in the Rann of Cutch was suddenly
submerged.’ A more doubtful instance is the elevation and subsidence
which is said to have taken place on the Arakan coast in the middle
of the last century, presumably during the great earthquake of 1762. A
raised beach which is g feet above sea-level at Foul island and 22 on the

1 Records, X1, 302, (1878). 90, (1823); Lyell, ‘Principles of Geology.”
2 Memoirs, XX, $2, (1883). See also A. B. Wynne, Memoirs, IX/ 26,
3 McMurdo, Trans, Lit. Soc., Bombay, II, (1872).
12 GEOLOGY OF INDIA—PHYSICAL GEOGRAPHY. (Chap. i.

north-west of Cheduba island, has been attributed to the effects of this
earthquake which further north is said to have caused the permanent sub-
mergence of 60 square miles near Chittagong.?

Away from the sea ccast, the Andaman and Nicobar islands have
certainly at one time been connected with Arakan, and the intricate chan-
nels and long ramifying fjords which penetrate the great Andaman and
adjoining islands indicate a considerable submergence. Along the coast
there are, however, indications of minor oscillations of level, both upwards
and downwards, within the recent period, the last movement being probably
one of subsidence.®? Off the west coast of India the coral archipelagoes of
the Laccadive aud Maldive islands probably mark the site of submerged
land, though this is a matter still under dispute.

Besides the changes produced by rise and fall of the sea-level as
compared with that of the land, there have been minor modifications of

 

Fig. 1.—Pagoda on the sea-shore at Tranquebar.

the shore line due to erosion and accretion of land. St. Thomé, a short
distance south of Madras, is said to have formerly been situated 12
leagues inland and, 40 miles further south, the town of Mahébalipur is
said to have been overwhelmed by the sea. Still further south, erosion
of the sea beach at Tranquebar is well attested by old records as well as the
destruction of a large portion of an old pagoda, whose eastern gate tower
had been partially destroyed in 1859* and has probably now been com-
pletely removed.

1 Phil Trans., VII, 251, (1763); G. P. Hal- | Oldham, Records, XVIII, 143, (1883).
stead, Four. As. Soc., Bengal, X, 433, (1841) ; 8 T. J. Newbold, Four, Roy, As. Soc., VIII,
F, R, Mallet, Records, X1, 190, (1878). 250, (1846).

3S. Kurz, “Report on the Vegetation of the 4 W, King, Memoirs, 1V, 362, (1864).
Andaman Islands,” Calcutta, 1870; R. D. :
Chap. 1. CHANGES OF COAST LINE. 13

Evidence of the advance of land is to be found on the Tinnevelli coast,
where the deseited town of Korkai, now five miles inland, has been iden.
tified with the “ Kolkoi Emporium” of the classical geographers. About
600 B.C. this town was the capital of a kingdom and apparently an
important sea-port, By the time that Marco Polo visited this coast in
1292 A.D., the advance of the land had necessitated the abandonment of
the old port and the establishment of a new one at “Cail,” a town which
also has decayed and was forgotten till its site was discovered and re-

a
d

 

Fig. 2.—Fart of Tinevelly coast shewing present position of old seaports.

cognised by Bishop Caldwell in the modern village of Kayal, and made
public in Colonel Yule’s edition of the travels of Marco Polo.”

A similar advance of the shore line is said to have taken place on the
east coast of the Gulf of Cambay, and it is said that the Rann of Cutch
was once a gulf of the sea with sea-ports on its shores, and that remains of
ships have been found imbedded in the mud*. The Rann is now a sort of
debateable land, being flooded during the south-west monsoon and a dry
barren mud flat during the rest of the year; the change, which has
indubitably taken place, if not so recently as has been supposed, was
doubtless due to silting up, partly aided by a slight elevation of the land.

The evidences of alterations of level along the sea coast, which have
been detailed above, point toa slight elevation of the land during the
post-tertiary period, though too small to have any appreciable influence
on the climate.

1 rst edition, Vol. II, p. 307, (1871). Ixix, Ixxxv, (1868) ; A.B. Wynne, Memoirs,
2 Trans. Geog. Soc., Bombay, XVM, pp. lvi, | 1X, 26, (1872).

’
I4 GEOLOGY OF INDIA—-PHYSICAL GEOGRAPHY. (Chap. IL

This is important in its bearing on the evidence that there is of the
cold of the glacial period having been felt in the peninsula of India
There is no physical evidence, so far as is known, of a geologically recent
cold epoch, and some geologists have doubted whether India was affected
by the glacial period. In the Him4layas there is everywhere abundant
evidence of the glaciers having extended to lower levels than they
reach. Grooved and polished rock surfaces have been found now at as
low a level as 7,500 feet in Pdngi,! and in a higher latitude large boulders
are found imbedded in the fine silt of the Potwar at an elevation of
less than 2,000 feet above the sea. Besides these there are many cases of
large erratic blocks and supposed morraines which have been referred by
some observers to glaciers and by others to river action, ‘The positive
and unmistakeable proofs of a period colder than the present are suffi-
cient to enable us to discard all the more doubtful evidence, and more
recent investigations have shown that it cannot be.attributed, as was once
suggested,’ to a former greater elevation of the Himdlayas than they now
attain,

These indications do not point to a sufficient diminution of tempera-
ture of the Himalayas to make it probable, or even possible, that there
should be any actual physical proofs of the glacial period having been felt
in the Peninsula, and it is of importance to ascertain whether there is any
collateral evidence of a cold period having affected India in later tertiary
or post-tertiary times, it being remembered that a general refrigeration
of the earth’s surface, sufficient to produce an arctic climate in Europe,
would not diminish the temperature of the Indian Peninsula beyond the
average of the temperate zone at the present day.

The argument is, briefly, as follows. On several isolated hill ranges,
such as the Nilgiri, Anamalai, Shevaroys and other isolatcd plateaux in
Southern India, and on the mountains of Ceylon, there is found a tem-
perate fauna and flora, which does not exist in the low plains of South.‘
ern India, but is closely allied to the temperate fauna and flora of the
Himalayas, the Assam range (Géro, Khdsi, and Nagé hills), the mountains
of the Malay peninsula, and of Java. Even on isolated peaks, such as
Paérasnath (4,500 feet high in Behar) and on Mount Abi in the Aravalli
range, several Himalayan plants exist. It would take up too much space
to enter into details; the occurrence of a: Himdélayan plant like Rhodo-
dendron arboreum, and of a Himdlayan mammal like Martes flavigula
on both the Nilgiris and Ceylon mountains, will serve as an example of a
considerable number of less easily recognised species. In some cases there
is a closer resemblance between the te mperate forms found on the peninsular

1C, A. McMahon, Records, X1V, 310,(1881).) 3 H.B. Medlicott, Memoirs, LI, pt. ii p. 155:
3 W. Theobald, Records, X, 140, (1877). (1864), , :
Chap. 1] GLACIAL EPOCH IN INDIA. 15

hills and those on the Assam range! than between the former and Him4-
layan species, but there are also connections between the Himdlayan and
peninsular temperate regions which do not extend to the eastern hills.
The most remarkable of these is the occurrence on the Nilgiri and Anamalai
ranges and on some hills further south, of a species of wild goat (Capra
hylocrtus), belonging to a sub-genus (Hemitragus), of which the only other
known species, C. jemlatca, inhabits the temperate region of the Himé-
layas from Kashmir to Bhutd4n, This case is remarkable, because the
only ‘other wild goat found completely outside the Palzarctic region is
another isolated form on the mountains of Abyssinia.

The range in elevation of the temperate fauna and flora of the
oriental region in general appears to depend more on humidity than
temperature, many forms which are peculiar to the higher ranges
in the Indian hills being found represented by allied species at lower
elevations in the damp Malay peninsula and archipelago, and some of
the hill forms are even found in the damp forests of the Malabar coast.
The animals inhabiting the Peninsular apd Singalese hills belong, for the
most part, to species distinct from those found in the Himdlaya and
Assam ranges. In some cases even genera are peculiar to the hills of
Ceylon and Southern India, apd one family of snakes is unrepresented
elsewhere. There are, however, numerous plants and a few animals in-
habiting the hills of Southern India and Ceylon, which are identical with
Himalayan and Assamese hill forms, but which are unknown throughout
the plains of India.

That a great portion of the temperate fauna and flora of the Southern
Indian hills has inhabited the country from a much more distant epoch
than the glacial period may be considered as almost certain, there being
so many peculiar forms. It is possible that the species common to Ceylon,
the Nilgiris, and the Himdlayas, may have migrated at a time when the
country was damper without the temperature being lower, but it is
difficult to understand how the plains of India can have enjoyed a
damper climate without either depression, which would have caused
a Jarge. portion of the country to be covered by sea, a diminished temper-
ature which would check evaporation, or a change in the prevailing winds.
The depression may have taken place, but the migration of animals and
plants from the Himdlayas to Ceylon would have been prevented, rather
than aided, if the southern area had been isolated by sea, so that it may
be safely inferred that the period of migration and the period of depression

-1 Only one species of plant, however, is men- | mus nilagiricus, has the same distribution,
tioned by Hooker and Thomson (“Introductory | and the genus Streptaxis is found in Burma,
Essay to the Flora Indica’, p. 238) as being found | the Khdsi hills,and the Southern Indian ranges,
both in the Khdsi hills and Nflgiris, but not in | but not in the Himdlaya west of Bhutan,
the Himalayas. One land-shel! at least, Budi- | Several other instances might be quoted.
16 GEOLOGY OF INDIA—PHYSICAL GEOGRAPHY.

(Chap. 1.

were not contemporaneous. A change in the prevailing winds is improb-
able so long as the present distribution of land and water exists, and
the only remaining theory, to account for the existence of the same
species of animals and plants on the Himalayas and the hills of southern
India, is depression of temperature.!

The Indian empire can boast of one.volcano, which is at present
dormant, but has been in active eruption within the century. Barren island
in the Bay of Bengal is not only a perfect model of a volcano, but is
classical in the history of geological controversy. It has been repeat-
edly referred to and described by geological writers, but the earlier refer-
ences were all more or less inaccurate, and it is only within the last few
years that a careful description of it has been given by Mr. F. R. Mallet?

. The volcano of Barren island has an irregularly circular form of about
2 miles in diameter, composed of an outer rim, rising to a height of
from 700 to 1,000 feet above the sea and surrounding a slightly ellipti-
cal amphitheatre, whose larger axis runs north-east and south-west with
a length of about g furlongs, From the centre of this a cone of re-
markably regular form rises to a height of 1,015 feet above the sea. The
depression, where the slope of this cone and the inner slope of the
amphitheatre meet, has a maximum elevation of 300 feet above the sea,
and is almost entirely occupied by three distinct lava streams, It is
evident that the island was once much higher than it is at present, and the
outer rim marks the limit of the crater, produced by some great paroxys-
mal eruption which blew away all the upper portion of the old cone.
The bottom of this old crater must have been much betow the level
of the sea and, for a time, before the new cone attained its present
dimensions, the sea must, as is represented in the older descriptions of
the island, have flowed round its base, between it and the foot of the inter-
nal slope of the amphitheatre, but there is no authentic record of any one
ever having seen this stage. In 1789, when Blair saw the island, the sea
did not penetrate into the amphitheatre, which had all been filled up to
above sea-level. The volcano was then in active eruption, throwing out
blocks and scorize, and it may be that it had been visited at some earlier

1 The above is a meagre and condensed
account of a very interesting subject, which
requires further enquiry. One possible objec-
tion may be answered at once, It is true that
many of the temperate damp-loving forms of
the Nilgiris and Ceylon hills are forest forms,
and it may be urged that they might have
migrated when the plains of India were covered
with forest. But, judging from what remains
of the forest on the plains of the Carnatic,

Deccan, Central Provinces, etc., the flora, even
when the whole was forest, differed so widely
from that of the hills, that it is improbable that
any general diffusion of hill species could have
taken place without a change of climate.

ina subsequent chapter reference will be
made tothe probable influence of the glacial
epoch on the Siwdlik mammalian fauna.

2 Memoirs, XXI, 251, ff, (1885).
Chap. 1.) VOLCANOES. 17

period, before the hollow was completely filled up, and that the only record
left of this visit is to be found in the erroneous description: which was at
one time current in text-books.

It must be remembered that the portion of the volcano above sea-level,
which is all that has been referred to in the description, is but an insigni-
ficant portion of its whole bulk, Soundings, taken by Captain Carpenter,
show that the cone rises from a depth of 800 fathoms below the sea, and
that the total height is consequently some 6,000 feet at Satie or was
8,000 feet before the upper part of the outer cone was blown away.!

At the time of Blair’s visit there appears to have been no lava stream
in the gap where the outer rim is breached, but in 1832 the lava was
there, and still so hot that the water in contact with it was boiling. Since
that period the flow has cooled down and the temperature of the water,
which percolates beneath the lava and issues as a spring on the sea shore,
has steadily diminished at each visit, till it was no more than 110° F. in
1886,2 It seems certain, therefore, that this lava flow was poured out later
than 1789, and probably within the present century.

Seventy-five miles north-north-east of Barren island lies the island of
Narcondam, indubitalily of volcanic origin like the former, but composed
almost, if not quite, entirely of hornblende andesite lava with little or no
volcanic ash, It is not certain whether this volcano ever had a crater,
as it may have been of the so-called endogenous type, formed by the
quiet extrusion of lavas unaccompanied by any crater-forming materials.
The complete obliteration of the crater, ifthere ever was one, is in itself an
indication of the period for which the volcano has been extinct, and in any
case the deep ravines, with which its sides are scored, are an equally
eloquent testimony of the time during which subaerial denudation has been
uninterruptedly at work, so that this volcano has probably been longer
extinct than either of the two that follow.

About 50 miles north-north-west of Yenangyoung and 25 to 30 miles
east-south-east of Pagdn, both large towns on the Irawadi, the extinct
volcano of Puppa? rises toa height of about 3,000 feet above the undulating
country composed of pliocene sands and gravels. The mountain has
preserved its original form to some extent, but the crater has been greatly
broken down by denudation, and the rim completely cut away at one
point, where the drainage from the interior has made itself a means of
exit. The peak consists of ash breccia, but lava flows, mostly trachytic,
form the lower slopes and the surfaces around the base of the volcano.
Among these flows are some ofa very beautiful porphyry, with crystals cf
pyroxene.

The horizontal beds of gravels and sands around the base of the vo!cano

1 Records, XX, 46, (1887). 8W.T.Blanford, Your, As. Soe, Bengal,
2 Records, XX, 48, (1€87). XXXI, 215, (1862),
Cc
18 GEOLOGY OF INDIA—PHYSICAL GEOGRAPHY. — (Chap. J.

contain fossil wood and ferruginous concretions, and apparently belong to
the pliocene fossil wood group. They are capped by the lava flows,
contain pumice and volcanic fragments, and, in one place, a bed of ash
breccia was found interstratified with them. It appears highly probable,
therefore, that the volcano was active in pliocene times, but it may have
continued to emit lava and scoriz at a later period.

Far to the north the extinct volcano of Hawshuenshan near Momien in
Yunndn has been described by Dr. Anderson'; and near Kanni, on the
Chindwin, Dr, Noctling observed basalt breaking through the pliocene
sandstones and forming a cone on their surface, but no detailed notice of
this locality has been published.

In this connection mention may be made of a mass of trachyte which
is found about four miles east by north of the village of Byangyi on the
Rassein river and some 30 miles south of Bassein town. It is about six
feet in diameter, there is no reason to suppose that it has been trans-
ported from a distance and no similar rock is known anywhere else in
the province. No rock is seen in contact with the trachyte, but unaltered
shales and sandstones of upper nummulitic age are seen not far off, dipping
at low angles. Close to the block itself are some fragments that have
flaked off, and among them a piece of shale, which had a somewhat baked
appearance on one side, was found. This, taken in conjunction with the
fact that it lies on the same general line as the volcanoes just described,
suggested the idea that it forms part of an intrusive neck,? but the true
relations of the mass are obscure.

It will be noticed that these old volcanoes lie along a line which, if con-
tinued to the south, would be continuous with the general direction of the
great chain of volcanoes running through the islands of Java and Sumatra
in the Malay peninsula, and this suggests that they form the northern termi-
nation of what is known as the Sunda chain of volcanoes. The observation
is interesting and important in view of the fact that this chain has been
supposed to find its final expression in the pseudo-volcanic phenomena on
the Arakan coast which are described below.

But before passing on to this subject it will be well to notice some
doubtful cases of volcanic action in the Indian Peninsula and on its shores.
In 1756 a submarine eruption is said to have taken place off the coast of
Pondicherri, which threw up large quantities of ashes and pumice and
formed an island half a league long and of the same breadth, No exact
details of locality are given, but the account is a very circumstantial one?
and, unless a pure fiction, must refer to a true volcanic eruption, It may

1Report on the expedition to Western 2 Memoirs, X, 330, (1873).
Yunnan, Calcutta, 1871, p. 87. * “Asiatic Annual Register,” 1758, reprinted
in Four. As. Soc., Bengal, XVI, 500, (1847),
Chap. I. ] LONAR LAKE. < 19

be noticed that the Admiralty chart of the Bay of Bengal marks a sound-
ing of 5 fathoms in east longitnde 80° 42’, north latitude 12° 46’, with the
remark ‘Doubtful’; the position would agree sufficiently well with that
indicated in the account, and the depth is that which would be produced
by the action of the waves.

This alsois the best place to notice a very curious crateriform
lake, situated in the interior of the Indian Peninsula, near the village
of Londr, about 40 miles east by north of Jélna in the northern part of
the Nizdm’s territory, and about half-way between Bombay and Ndgpur.
The surrounding country for hundreds of miles consists entirely of Dec-
can trap and in this rock there is a nearly circular hollow, about 300 to
400 feet deep and rather more than a mile in diameter, containing at
the bottom a shallow lake of salt water without any outlet, whose water
deposits crystals of sesquicarbonate of soda. The sides of the hollow
to the north and north-east are absolutely level with the surrounding
country, whilst in all other directions there is a raised rim, never exceed-
ing 109 feet in height and frequently only 40 or 50, composed of blocks
of basalt, irregularly piled, and precisely similar to the rock exposed
on the sides of the hollow. The dip of the surrounding traps is away
from the hollow, but very low.?

It is difficult to ascribe this hollow to any other cause than volcanic
explosion, as no such excavaticn could be produced by any known form
of aqueous denudation, and the raised rim of loose blocks around the
edge appears to preclude the idea of a simple depression. It is true that
there is no sign of any eruption having accompanied the formation of the
crater, no dyke can be traced in the surrounding rocks, no lava or scoriz
of later age than the Deccan trap period can be found in the neigh-
bourhood. The raised rim is very small, and cannot contain a thousandth
part of the rock ejected from the crater, but it is impossible to say how
much was reduced to fine powder and scattered to a distance, or removed
by denudation.

Assuming that this extraordinary hollow is due to volcanic explosions,
the date of its origin still remains to be determined. That this is long
posterior to the epoch of the Deccan traps is manifest, for the hollow
appears to have been made in the present surface of the country, carved
out by ages of denudation from the old lava flows. To all appearance
the Londr lake crater is of comparatively recent origin, and if so it
suggests that, in one isolated spot in India, a singularly violent explo-
sive action must have taken place, unaccompanied by the eruption of

2 Malcolmson, Geol. Trans., and series, V, 562, | view of the lake);—G. Smith, Mad. Your, Lit.
(:840); Newbold, Y¥our., Roy. As. Soc., 1X, 40, | Sei, XVII, 1, (1856). See also Records, I,
(1848) (with this paper there is a fairly executed | 63, (1868), where other references are given.

C2
20 GECLIGY OF INDIA—PIYSICAL GEOGRAPIY. {Chap. IL

melted rock. Nothing similar is known to occur elsewhere in the Indian
Peninsula,

Associated with true volcanoes in name at least, even if, as is held by
many geologists, in nothing else than name, are mud volcanoes, of which
two principal groups are known, in Burma on the east, and Baluchistan on
the west, of India, respectively. ;

Of the Burman ones the best known are those of Minbu on the Irawadi,
and those of the islands of Ramri and Cheduba on the Arakan coast.’ A few
others are reported, but they are small and isolated, and consist only of
temporary outbursts. ‘

The Ramrf mud volcanoes are more interesting than the others, since
they alone, so far as is known, are subject to paroxysmal eruptions of great
violence, and from them alone stones have been ejected and flames emitted.
Some of the principal phenomena may be briefly described here. There
are about a dozen or rather more vents in Ramrf island itself, more than half
that number in Cheduba, and a few in the other neighbouring islands. Near
Kyauk-pyd in Ramrf, six occur ina line, within a distance of about a mile
and a half along the summit of a low broad ridge.

Many of the vents consist of truncated cones, built up of the dried mud
ejected by outbursts of gas. The crater, filled with more or less liquid or
viscid mud through which the gas escapes, occupies the top of the conical
hillock. The majority, however, of the Ramri mud volcanoes consist
of mounds, composed on the surface of angular fragments of rock and
having scattered over them a few small mud cones with craters at the top,
varying in height from a few inches to eight or ten feet. When gas ceases
to be omitted from a vent, the mud is rapidly washed away by rain and
there remains a low mound, composed of angular fragments of rock which
were ejected together with the mud, and the repetition of a similar process
accounts for the formation of the mounds. The mounds in Ramrf are from
50 to 100 yards in diameter, with a height of from 15 to 30 feet, two, of
exceptional size, near Pagoda hill in Cheduba, being 200 to 250 yards across.
The cones in which the mud is viscid are very steep, being built up partly
of small quantities of mud, spurted out by the evolution of gas so as to form
a hard rim round the mud crater, partly of mud poured out from the crater
down the slopes through broken portions of the rim,

Besides the gas and mud,a_ small quantity of petroleum is usually
discharged from the vents. The gas consists mainly of marsh gas (light
carburetted hydrogen), probably mixed with some of the more volatile
hydrocarbons usually associated with petroleum. The mud is simply the

For a description of the mud volcanoes of | volcanoes are described, with full references to
Minbu by Dr. Oldham, see Yule’s Nazrative | earlier accounts, by Mr. Mallet in Records, XI,
of the Mission to the Court of Ava in 1855, 188, (1878). Sketches of the cones are given
appendix, p. 339. The Ramrfand Cheduba mud | in both cases. a
; ‘ KAKHYENS
Shillon ¢ : : J
Hoov shuen shore ” MoM

TIPPERAH
HILLS

M {A H

Kut
\ _J{eMandelay
pecUmmer ap oora

Ava

)

S 5 ae Puppadhoung

 

Ko fgrtaban
? aulmein
NA :
; of
"Mart mK
2

?

Great Coco
uU

i ¥s Nareondoum

) Andaman I?

# Barren [%

Port. Blair

 

Voleanoes of Burma & the Bay of Bengal.
21

Chap. I. j SALSES AND MUD VOLCANOES.

grey shale or clay of the tertiary rocks, mixed with water containing some
salt in solution.

The association of petroleum with large quantities of marsh gas, and the
frequent emission of, usually saline, water and of gas in abundance from
borings for mineral oil are too well known to require the recapitulation of
details. Both petroleum and gas are known tobe found in many localities
amongst the tertiary rocks of extra-peninsular India and Burma. Not un-
frequently both gas and mineral oil issue with water in the form of a bub-
bling spring being, perhaps, forced to the surface by the pressure of the gas,
Whether a spring of this kind forms a “mud volcano” or not depends
evidently on the nature of the beds traversed on the way to the surface.
As a rule the mud in these vents is either of the same temperature as the
air or a little higher, but in the Ramri craters a higher temperature has
been recorded alter the more violent eruptions.

These paroxysmal eruptions appear to occur at irregular intervals,
are at times very violent and appear frequently to accompany earth-
quakes. Mud and stones are shot out with great force, accompanied by
large quantities of inflammable gas, which in many cases takes fire
and lights up the country for miles around. Some of these eruptions
of ignited gas have taken place at sea off the coast of Ramri, and in one
case a small island was formed near False island, south of Ramrf and
south-east of Cheduba, but it was soon washed away again. The ejected
stones are in all cases fragments of the tertiary rocks, chiefly shale or sand-
stone, some of them being from half a cubic foot-to a cubic foot in size anda
few are larger, but the majority range fram half an inch to five or six inches
in diameter. The ignition of the gas is ascribed by Mr. Mallet to frictional
electricity,! and nof to the high temperature at which the various ejecta issue,
the fact that the stones, and even fragments of lignite, thrown out during
eruptions are, as a rule, entirely unchanged by heat, proving that the gas is
not in a heated state previous to emission. A very few rare and excep-
tional fragments of burnt and reddened shale have been found, and these
have probably been calcined by the flaming gas.

It should be noticed that even the fiery eruptions of the mud volcanoes
of Arakan have nothing in common with the igneous outbursts of true volca-
noes. In the former gas is emitted at, in all probabi'‘ity, a low temperature
and is ignited, or perhaps occasionally exploded, when mixed with atmo-
spheric air. In the latter, red-hot lava and scoria are ejected, and the
appearance of flame is due either to the high temperature of the substances

1 Records, XI, 202. Mr. Mallet points out that , produce great friction, is similar to that of
the principle of the hydro-electric machine, in | violent evolution of gas from such vents as
which the production of electricity of high ten- | those of Ramri. He also notices the well-
sion is due to the issue of partia!ly condensed | known fact that lightning often accompanies
steam through small orifices of such form as to | volcanic eruptions.
22 GEOLOGY OF INDIA—FHYSICAL GROGRAPIIY. [Chap. I.

projected into the air, or to the reflection of masses of glowing lava by
condensed vapours or by cloads of volcanic dust.

One mud volcano is known in eastern Assam, but none have been found
in the Himalayas or in the Peninsula of India. On the Balichistdn coast
in Lus a number have been described, which do not appear to be sub-
ject to the same paroxysmal eruptions as those of Ramri and attain a
greater size, the largest being over 300 feet in height. How far this is due
to the absence of paroxysmal eruption or to a greater constancy of the
vents, and how far to the rainless nature of the climate, is not clear, but in
all other respects they resemble those already described.
CHAPTER IL.

METAMORPHIC AND CRYSTALLINE ROCKS.

Gneissose rocks of the Peninsula of two ages—The older gneiss—The newer gneiss—Bundel-
khand gneiss—Newer gneiss of Singrauli—Contrast with gneiss of Bundelkhand—Chutié
Nagpur —Orissa—Central Provinces—Haiderébid—Nellore—Bellary—The southern Kon-
kan— Malabar —Nilgiris—Trichinopoli and Arcot—Madura and Tinnevelli — Instrusi ve trap
of Southern India—Gneissose rocks of the Ardvallis—Extra-peninsular India--Afghdn-
istin - Him4layas—Assam—Hills east of the Irawadi and Malay Peninsula.

The oldest rocks in India, and consequently the first to be considered
in the arrangement adopted in this book, are those belonging to the great
series of gneissose rocks, which covers a larger area in the Peninsula of
India than all the newer formations put together, and forms the foundation
on which they are built up. In spite of the great interest of the numerous
problems presented by these rocks, they have as yet received but little
attention from the Geological Survey, its energies having been mainly
directed towards those districts where the most important results were to
be expected in the shortest time, and we have consequently but little
detailed information regarding the gneissose rocks, except in the imme-
diate neighbourhood of the sedimentary formations.

How imperfect is our present knowledge of these rocks may be judged
from the fact that it has been found impossible to distinguish, on the accom-
panying map, between true gneiss and intrusive granite. It is only within
late years that the frequency with which granite has acquired gneissose
characters has been recognised, and it is highly probable not only that
some large exposures, of what has been described as granitoid or massive
gneiss, are composed of gneissose granite, but that what have in many cases
been described as beds of gneiss, interbedded with non-felspathic schists, are
in reality veins of granite which have been intruded along the foliation
planes, and subsequently acquired a parallel arrangement of their consti-
tuent minerals. Under these circumstances it is impossible to attempt
anything like a complete or connected account of the gneissose rocks,
and the following pages must be taken as the merest sketch, whose main
use will be to show how much has yet to be learnt.

In spite, however, of the imperfect state of our knowledge, certain
leading features have come out from the investigation, so far as it has
24 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. IL

one, and the most important of these is the recognition cf the fact that
the gneissose rocks do not belong to a single system, but are comprised
in two or more distinct systems differing in age from each other.

The oldest of these, often described as the Bundelkhand gneiss
from its having been first recognised in the country of that name, is
characterised by its massive structure, with the foliation generally but
obscurely developed, and the extreme rarity of accessory minerals. It
has been recognised in Bundelkhand, in the gneissose inliers of Sing-
bhéim and in the Bellary district, and it is possible that the massive
granitoid gneisses of other parts of Southern India may belong to the same
system, In the first and last named districts, and apparently also in the
second, it formed the floor on which the oldest distinctly sedimentary beds
of the Peninsula were deposited, showing that it must, even at that
remote period, have been an ancient rock which had undergone a great
amount of sub-aerial denudation. These facts indicate a greater unity of
character than can be found in the very diverse characters of what are
believed to be the newer gneisses, and seem to show that the Bundelkhand
gneiss may be considered as a true rock system.

The second, or newer, type of gneiss exhibits a much better deve-
loped foliation as a rule, it is at times interbedded with schists and is
distinguished from the first by the abundance and variety of the accessory
minerals it contains. The rocks of this division are looked upon as newer
than those already mentioned, firstly, because no unconformable contact
of original deposition, between them and the sedimentary beds of the
older transition systems, has been observed, secondly, because, where
they are in contact with the latter, bands of gneiss have in several places
been observed apparently interstratified with the distinctly sedimentary
beds, and thirdly, because the general type of rock is, according to theories
that are widely held, newer than that of the Bundelkhand gneiss, All
three of these arguments are open to dispute. As regards the first it
may be observed that the presence of a contact of original unconformity
with the transition beds has of itself been held sufficient to show that
the gneiss belongs to the older class, as regards the second, there is no evi-
dence in the recorded observations to show that the supposed interbedded
gneiss is not gneissose granite intrusive along the planes of bedding, or
a foliated arkose, and as regards the third, the theories are by no means
completely established. Still, though absolute proof is wanting, there is a
distinct balance of evidence in favour of the more foliated gneisses being of
newer origin than the massive forms, and this is in accordance with the type
which the foliated gneisses and their accompanying beds of crystalline
schist exhibit, intermediate between that of the granitoid gneiss on the
one hand and the distinctly sedimentary, though metamorphosed, beds of
the transition systems on the other.
Chap. If. THE BUNDELKHAND GNEISS. 25

In the only critical and detailed examination of the gneisses of the
Peninsula that we have at present,’ Lacroix has suggested the existence
of groups of successive origin among the gneisses of the Salem district
and Ceylon. His investigations were based entirely on specimens whcse
localities had been but imperfectly recorded, and the suggestion still lacks
that confirmation from detailed observations in the field which alone could
render it authoritative. It is, however, probable, as will appear in the
sequel, that there are more than two successive series of gneisses repre-
sented in the Peninsula of India.

After these introductory remarks it will be best to proceed to a brief
review of the principal points of our knowledge of the gneisses in
different districts of the Peninsula, commencing with that of Bundel-
khand. .

This gneiss forms the basis of lower Bundelkhand, as distinguished
from the higher portions of that district lying on the adjoining Vindhyan
plateau. On the north-north-east border, for 200 miles, the gneiss is
gradually covered by the superficial deposits, forming outlying and mar-
ginal portions of the Gangetic plains, at an elevation of 500 to 600 feet
above the sea. Elsewhere the area is very sharply bounded by a scarp
of overlying formations, whether Vindhyan or transition. Along the base
of the scarp to the south-west the elevation varies between 1,000 and
1,200 feet above the sea, the scarp itself rising to 1,900. The gneiss
sometimes forms hills, but the general features of the ground are flat, un-
dulating uplands, sparsely cultivated, including shallow valleys and plains
of alluvial land.

~ Gneiss is by far the commonest rock.? It may be described as coasist-
ing of some six minerals—red orthoclase felspar, a white plagioclastic
felspar (probably oligoclase), quartz, hornblende, chlorite, and mica, In
any one place the gneiss may contain only two of these, or it may in-
clude all of theia. With regard to texture every variety is met with,
from a homogeneous felstone, in which no individual mineral can be dis-
tinguished even with a lens, to a coarsely porphyritic rock, including
felspar crystals more than two inches long. The orthoclase nearly always
forms the main mass, and exceeds in quantity all the other minerals to-
gether. Its ordinary colour is a darkish red, but now and then it is much
paler, and almost or quite white. In such cases it is not easy to distin-
‘guish the felspars on a fresh fracture, when both occur, bat the difference
becomes apparent on a weathered surface. The orthoclase has a
vitreous, or somewhat pearly, lustre and translucent aspect, the other is
1 Bull. Soc. Frang. Minéral., X11, 83, (1889); | 7 The characters of the rocks of this area are

Records, XXIV, 155, (1891). chiefly taken from the unpublished reports of
Mr. F, R.' Mallet.
26 GEOLOGY OF INDIA—GNEISSIC ROCKS, {Chap. II.

dull and quite opaque, having been superficially altered into kaolin. The
plagioclastic felspar is a comparatively uncommor mineral, when present
it plays quite a subordinate part, and occurs in much smaller crystals than
the orthoclase. As often as not, no free quartz can be detected in the
gneiss, and it very rarely indeed occurs in large quantity. The rock is
usually hornblendic, but sometimes the hornblende is partially or entirely
replaced by chlorite and mica. The mica is of more than one species, both
uniaxial and biaxial, and of varying colours, green, brown, black, and
silvery-white, the last being rare and chiefly found in the southern part
of the area.

Foliation is seldom well developed, the rock being very commonly to
all appearance perfect granite, but no direct evidence has been obtained,
beyond this, of the existence of true granite, and sometimes, by close
examination, traces of foliation may be detected in rock which at first
sight appears quite devoid of it. ‘This foliation has generally an east-
north-east direction, but varies to north-east and east-south-east. The
planes are more or less vertical.

The subordinate varieties of rock, all combined, are of very insigni-
ficant extent in comparison to the gneiss, One of the most prominent is
hornblende rock, which sometimes rescmbles trap very closely, parts
of it being as fine-grained as the intrusive dykes of diorite, or the over-
flowing basaltic trap, and weathering into similarly rounded lumps, but
this variety passes into a much coarser kind in which the felspar and
quartz are well separated. The most trappean looking portions, more-
over, contain thin strings and films of epidote, which have not been
observed in the trap. Another well-marked variety contains about equal
quantities of hornblende and white or greenish-white felspar, in crystals
about an eighth of an inch long. This rock sometimes contains, in addi-
tion, an inconsiderable proportion of reddish felspar, quartz, and green
mica, and very minute specks of iron pyrites.,

Besides the hornblende-rock, talcose, hornblendic, chloritic, quartzose,
and even argillaceous schist, and the combinations of these with each other,
occur with the gneiss, Mica schist has not been observed. Schists are of
very rare occurrence in the gneiss generally, but all the above varieties are
to be found in some force in the southernmost part of the area, in the
Maraura region, This peculiarity of distribution is so marked that it
was thought that the schistose strata might here be separable from the
gneiss, a suggestion which is much encouraged by the fact that the great
quartz-reefs, elsewhere so prevalent in the gneiss, stop short of this ground,
but it has not as yet been found possible to draw a line between the
gneissic and the schistose subdivisions. Gneiss of the usual type is still
a prevalent rock in the schistose area, and is the most southerly rock seen
at Shahgarh. It seems, too, to be truly associated with the schists. We
Chav, IL] BUNDELKHAND GNEISS—QUARIZ REEFS, 27

may perhaps at least infer that in this region we have the top of the
gneissic series of Bundelkhand.

It is a noteworthy fact that over the whole of this large area of gneiss
not a single bed of limestone has been detected.

Pegmatite veins, from a few inches to a foot or two in breadth, are
very common. [f these were intrusive, it might be expected that they
would be somewhat uniform in composition, irrespective of the nature of
the surrounding rock, but it has been invariably observed that the felspar
of the vein is the same as that of the rock adjacent, whether the latter is
orthoclase or plagioclase, or includes both, the chicf difference consisting
in the larger crystallisation and in the usual absence of the third mineral
(hornblende, etc.) in the veins. It is therefore presumable that the veins
were formed by segregation at the time of the crystallisation of the gneiss.

The almost total absence of accessory minerals in these rocks is
remarkable, Mr. Mallet only mentions epidote sparingly in the horn-
blende rock, schorl in some of the small quartz veins of the Maraura
region, small grains of ilmenite in some of the pegmatite veins, and
strings of altered kyanite in the quartzose rock of Dhdnkua hill, Sirall
pieces of galena have been sent from Jhdnsi for analysis, but their
locality is not known, and they may not have been procured from the
gneiss. Iron ore has been extensively burrowed for at Dhawara, it is a
decomposed earthy condition of the banded haematite and quartz. The
absence of any trace or tradition of gold in connection with the quartz
reefs is noteworthy.

All over lower Bundelkhand long narrow serrated ridges composed
of quartz-reefs form a most striking feature of the landscape. ‘They
run in straight lines, generally with a north-cast to south-west direction,
sometimes attain an elevation of 600 feet over the surrounding country,
and are exclusively confined to the gneissic series. They are pretty
equally distributed over the ground, with the exception noticed in the
southern region, which, moreover, lies right across the strike of the reefs
to the north-east. 7 ,

‘The prevalent strike of these quartz-reefs is about nerth-east by
north, but varies from north-north-west to east north-east. The
breadth of the veins varies from a few feet up to 100 yards, in the reef
west of Bhagwa. Some of them are traceable in a direct line for more
than 60 miles, the local interruptions which occur being sometimes due
to removal by denudation, sometimes to strangulation of the vein itself.
Other reefs, again, though of full thickness and very prominent at the
‘surface, are short and end abruptly. There is a good case of this at
Dehri, where a broad reef, 300 feet high, is only about a mile long, no
trace of it occurring in the gneiss to the north or south. The narrow
gaps by which the minor streams in many places cross the reefs give
28 GEOLOGY OF INDIA—GNEISSIC ROCKS. CChap. IL

peculiar facility for the formation of lakes, as a very short dam is often
sufficient to pond back a large surface of water, and many of the numer-
ous artificial lakes in Bundelkhand are formed .in this way.

The reefs are often affected by joint-planes, which sometimes give an
appearance of horizontal bedding to the mass. When parallel with the
direction of the reef itself they suggest, in a more puzzling manner, the
impression of a bedded mass. Now and then the quartz is much shattered.
At Deokalli and Bagpura samples might be taken for the Bijéwar horn-
stone-breccia, a description of which will be found on a subsequent page.

Foliation is often developed, owing to the metamorphic condition of
the reefs themselves, and it usually extends to the contiguous gneiss,
which is generally so amorphous. When the foliation in both rocks has
the same direction as the reef itself, the quartz mass might be taken to
be interbedded with the gneiss. Generally the foliation is oblique to the
direction of the vein, while still the same as that of the gneiss.

Many of the reefs are of greyish-white quartz. Frequently they con-
tain a large amount of impure serpentine, and occasionally they are formed
almost entirely of this material. The more northerly of the two ridges at
Dallipur is an example. At Rajdpur a band of nearly black serpentine
occurs, apparently a continuation of the quartz vein to the north. In
many cases the gneiss is serpentinous for some distance on each side of a
vein, there being no distinct separation between the two rocks, but sei-
pentinous gneiss has not been observed except near a quartz vein.
Steatite takes the place of serpentine in a few veins.

From his observations, as above sketched, Mr. Mallet concludes that
the reefs were probably formed before the metamorphism of the gneis.
was accomplished.

Many of the quartz-reefs, as well as the gneiss itself, are traversed by
more recent, and much smaller, veins of pure white quartz, the thickest not
much exceeding one foot in breadth. They are very frequently crystalline
and drusy in the centre, and they are always sharply distinct from the rocks
they traverse. Their direction is very irregular.

The gneiss of Bundelkhand is also remarkable for being traversed by
extensive trappean intrusions, none of which penetrate any of the younger
formations. These dykes, of true igneous rock, are more numerous than the
quartz-reefs, and exhibit nearly as much regularity in their course, their
prevailing direction being about north-west by north, so as to cut the reefs
obliquely at an angle of about 70°. Some few run east of north, or due
west. Many are of considerable size, a breadth of too feet being not
unfrequent, while some are much wider. They are often persistent for
great distances. The commonest type is an extremely hard and tough
close-grained greenstone (diorite), in which the hornblende and the white
felspar are sometimes clearly separated. The rock often weathers into
Chap. IJ,] BUNDELKHAND GNEISS— CONTIGUOUS FORMATIONS. 29

large rounded blocks without any tendency to exfoliation. The small
dykes are of a more earthy texture.

It is very rare to find any intersections of the dykes and reefs that
can be taken as conclusive of their relative age. It is not so uncommon
to find a dyke running close up toa reef on both sides without cutting
it, but such an occurrence might easily happen, even though the reef were
the older, as it may have offered a greater resistance to splitting. One
good case of the converse carries much more weight. Mr. Mallet records
an occurrence where a quartz vein, striking east-20°-north, abuts against
a strong dyke running west-20°-north, traces of the quartz being found
also on the other side. But he considers the general argument from
the condition of the two rocks to be independently eee erO the trap
has certainly not undergone metamorphism, whereas the reefs as certainly
have.

The history of this comparatively small area of gneiss would be very
interesting. It has served as a shore or a bed for each of the great
adjoining formations. The Bijdwars and the Gwdliors lie upon its margin
north and south, but no detached outliers of either are found within its
border, so that it may have been a well-elevated aréa at the period of their
formation. The same may be said of the lower Vindhyan deposits. It
is not so with the upper Vindhyans, of which the outliers are numerous,
and lie at considerable distances from the scarp of the basin. In the east
these form a portion of an ascending slope, the base of the capping
Vindhyan sandstone being highér in the outliers than in the scarp. In the
north-west it is curiously the reverse, the gneiss reaches high up all along
the western scarp, but the outliers of Vindhyan sandstone to the eastward
rest at the general level of the low country.

The next overlying formation is the Deccan trap, remnants of which
are found on the low ground in the southernmost part of the area, and
traces of the infra-trappean (Lametd) conglomerate occur more exten-
sively in the same position, That this portion of the scarp-bounded area
can have been so occupied almost necessarily implies that the whole of
the gneissic ground must, at the period of the Deccan trap, have had
a configuration very like what it has now, and, the source of the eruptive
rock being pr<sumably to the south or south-west, the lava must have
poured from the plateau to the lowlands. Trap does, in fact, occur
continuously from one level to the other in the Madanpur gorge, but its
condition suggests no resemblance to a lava stream,

The newer type of gneiss is well developed south of the Vindhyan
basin, in Rewd, Mirzdpur, and Behar, It has been there examined by
Mr. Mallet, from whom the account of the Bundelkhand area has been
taken, and consequently we have not to allow for discrepancies of
30 GEOLOGY OF INDIA—GNEISSIC ROCKS. [Chap. IL.

observation! He gives the following tabular abstract of the constituents
of the gneiss in Singrauli, a petty principality now absorbed in the Rewa
state and adjoining districts :—
1.— Minerals occurring as constituents of the gneiss :—
Quartz, orthoclase, oligoclase, muscovite, biotite, hornblende, epidote.

2.—Occurring in beds in the gneiss :—
Limestone, dolomite, corundum, magnetite, quartz as quartzite and
quartz-schist, hornblende as hornblende-rock, tremolite-rock and

jade, mica as mica-schist, epidote.

3.—Occuiring in veins in the gneiss :—
a.— Quartz in veins and reef-quartz, stilbite (?)
é.—In pegmatite veins ‘as constituents): orthoclase, oligoclase, quartz,

mica,
-c.—In epidotic veins : epidote, quartz.
4.—Accidental minerals in the gneiss :—
Magnetite, ilmenite, schorl, garnet.
5.—Accidental minerals occurring in the subordinate beds (2) of the gneiss :—
a.—In the limestone—magnetite, pyrite, hamatite, serpentine, chryso-
tile, phlogopite (?), wollastonite.
6.—In corundum bed—schorl, euphyllite, diaspore.
¢.—In jade bed (associated with corundum)—corundum, rutile (?), schorl,
euphyllite.
6.— Accidental minerals occurring in the veins, etc., in the granite veins :—
a—%n the quartz-veins— micaceous iron, tremolite, augite, epidote, schorl,
muscovite.
b.—In the quartz-reefs—galena, cerussite.
c.—In the pegmatite veins—schorl, garnet.

To this list may be added the minerals found by Mr. Mallet in the
same zone further to the east, in Hazdribdgh: lepidolite, tourmaline,
beryl, apatite, leucopyrite, and tinstone. Zircon is also said to occur.

The contrast between the minerals named in this table and the con-
stituents of the Bundelkhand gneiss is very striking. The most marked
differences are‘the abundance of the disseminated quartz, the compara-
tive frequency of limestone, and dolomite, and of mica schist, and the
general occurrence of accessory minerals in the Bengal gneiss.

The structural characters present another noteworthy point of con-
trast between these two gneissic series. In Bundelkhand the rock is
generally homogeneous and amorphous, the foliation obscure and con-
stantly in more or less vertical planes, as if due to the causes which
produce cleavage. In the Bengal gneiss bordering the Bijdwar basin on
the south, the'foliation coincides with what appears to be the original lami-
nation and bedding. It has a general east-north-east strike, corresponding
with that of the main rock-boundaries, but the alternating strata frequently
roll about at low angles of dip, or are crushed together confusedly, the
foliation cunstantly agreeing with the lie of the beds. A third point of

1 Mr. Mallet’s work tas been only partially published, Records, V, 18; VI, 42 ; VIL, 32, (1872-74).
Chap. IL J NEWER GNEISS OF SINGRAULI AND BENGAL. 31

contrast is in the relation to the overlying transition formations, which
rest nearly horizontal and undisturbed on the gneiss of Bundelkhand, and
are uniformly disturbed, metamorphosed and subjected to granitic intru-
sions where they occur in contact with the gneiss of the Bengal area.

Quartz-reefs have been described in this gneiss also, but to a very
subordinate extent, and their origin as veins is in many cases open to
question. A common mode of occurrence of this quartz, or quartzite, is close
to the boundary of the slate and gneiss series, but it does not coincide with
their junction, and it is not in any sense a contact-formation separating
sharply distinct types of rock, It occurs in the strike of the foliation and
stratification, and may well be an altered quartzite.

There is a rock common inthis northern area of the Bengal gneiss,
perhaps occurring most typically within the zone mainly occupied by the
transition series. It is known as dome-gneiss, from its weathering into
great hemispherical or ellipsoidal masses of bare rock, the only divisional
planes being concentric layers of exfoliation’. The domes are often
several hundred feet high, and fourm a very peculiar object in a landscape.
Foliation is always more or less traceable, and, in every respect of texture
and composition, the rock is the same as that of the thin bands alternating
with schists in the adjoining ground, Both are often porphyritic, the
dome-gneiss generally so, containing large ill-formed (rounded) crystals
of felspar, There can be no doubt that the peculiar form exhibited by
this rock is due to the occurrence of large masses of more homogeneous
composition than usual, ‘but the question is how these conditions were
produced, whether we must not suppose a partial degree of plasticity to
have;been attained, and whether the rock is not in a manner intrusive.
At the Kaldpahdr and the Bhiaura hills on the northern fringe of the
Hazdribdgh plateau, and the Mandar hill of the Bhagalpur district in the
same yeological region, there are very typical examples of the dome-gneiss.

The comparative rareness of trap-dykes in the Bengal gneiss is another
point of contrast with the Bundelkhand area. In some parts they are
pretty frequent, perhaps most so in the vicinity of the basins of Gondwana
rocks, where they are often continuous into such basins, thus fully establish-
ing their comparatively recent date, but they are by no means generally
distributed.

-Pegmatite is not uncommon in the gneiss of Singrauli. Mr. Mallet
does not consider this to be intrusive, its composition varying with the
rock it traverses, as was explained in tl:e case of the pegmatite of
Bundelkhand. In northern Hazdribdgh, however, he describes the extensive
occurrence of intrusive pegmatitic granite, ramifying in the most intricate
manner, in veins and dykes of from $ inch to 50 yards wide, through both
the gnciss and the transition schists, and maintaining its composition

VV. Ball, Memoirs, V1, 132, (1867) ; XVIII, 95, (1881).
32 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. II.

irrespective of the enclosing rock. It is composed, in order of crystallisa-
tion, of tourmaline, mica, felspar, and quartz, all four being generally
present, but their proportions vary greatly. Its texture is very uneven,
the cearsest forms being often found in comparatively narrow dykes. It is
in this rock that the mica-mines of Behar are worked. Not unfrequently the
pegmatite assumes the curious form known as graphic granite.

The gneiss of the Chuti4 Ndgpur districts, up to the basin of trans-
ition rocks in south-west Bengal, is more or less freely interbedded
with micaceous hornblendic and silicious schists, and occasional bands of
the porphyritic granitoid variety. Patches also occur of less highly meta-
morphic schists,

The junction of the Chutid NAgpur or Bengal gneiss with the trans-
ition rocks of Singhbhtim, is described by Mr. Ball* asa great fault. But
within this basin of submetamorphic rocks there are extensive inliers
of a gneiss which.seems to be of an older date than that of Chutid Nagpur,
and is apparently equivalent to the Bundelkhand gneiss. It is very uniform
and granitoid, there is a total absence of the thin-bedded gneiss, schists,
etc., which abound in the main gneissic area to the north, and we again find
a remarkable abundance of trap-dykes, forming two intersecting systems,
having north-westerly and north-easterly courses, respectively. In contact
with this Chutiaé Nagpur gneiss the transition strata exhibit a minimum
of alteration and disturbance. Mr. Ball describes them,. at and near
Chaibdsd4, as. sandstones and mudstones resting immediately on the rough
weathered surface of the granitic gneiss. There are local faults along the
boundary, but itis certain that the original relation of the two series is like
that between the Bijdwars and the Bundelkhand gneiss,as already described.

Further south, in the Tdlcher country, the ordinary newer type of
metamorphic rocks again prevails. The following rough classification of
them is given by Mr. Blanford? :—

Gness, a.- Hard, coarse, and felspcthic, becoming sometimes lithologically a
perfect granite.
», 6.--Soft foliated, quartzose or micaceous.
5 Sees but sometimes soft, containing garnets, frequently decom:
posed.

Hornblendic gneiss or schist, soft and foliated.
Quartz-schist or schistose quartz, occurs frequently in bands separated by softer

micaceous layers
The variations in composition coincide with the planes of foliation, the
prevailing direction being west-north-west to east-south.east.
Higher up the Mahdnadf valley in the neighbourhood of Sambalpur,
Mr. Ball® observed syenitic ard protogenic gneiss as common, horn-
blende-rock and schist as somewhat rare, strong quartzites forming the

1 Memoirs, XVIII, 88, 130, (1881). ? Memoirs, I, 39, (1856). 3 Records, X, 181, (1877).
Chap. IT. . GNEISSES OF NELLORE. 33

most peculiar feature in the gneiss; mica-schist, quartz-schist, and shaly
slate (and in one instance, near Kdtikela north-east of Sambalpur, a con-
glomerate) were found associated with the gneiss. The strike in this
region would seem to be very variable—east to west, north to south, north-
west to south-east, and north-east to south-west, being all recorded.

On the same latitude, about Nagpur, Dr Blanford! has noticed the
general resemblance of the gneissic rocks to those of Bengal. Here
dgain there is much irregularity in the strike.

There is but little information available regarding the gneiss of the
Nizam’s dominions in Haider&b4d. Westward of Kamamet the massive form
is known to be largely developed in broad bands running about north-north-
west, with the more foliated typesintervening. Froma short distance east
of Kamamet the schistose forms are found,? and continue to the eastern
limit of the gneiss area’.

In the Nellore portion of the Carnatic, amd the coastal regions of
the Kistna and Godavari districts, Dr. W. King recognised four distinct
varieties of gneiss, occupying separate areas and apparently indicating
differeat periods of formation, which were distinguished as—*

4. Micaceous, talcose and hornblendic schists, with few quartz- oles

schists or quartz-rock

neisses,
3. Foliated gneisses with frequent ee padideie or auanee -rock eee
2. Gfey gneiss, sometimes porphyritoid . . . ‘ 3 Massive.
1. Red granituid gneiss : is . . . gneisses.

The red granitoid gneiss only occurs south-west of Venkatagiri and
westwards into the upland of North Arcot and Cuddapah. The gneiss is
generally a close-grained aggregate of quartz and felspar (orthoclase, oligo-
clase (?) and a little albite), hornblende being oftea scarcely discernible
anc mica even more rare. The quartz is of two forms, dull amorphous,
and glassy, and the felspar usually of a pale flesh colour,

The grey granitoid gneiss forms a band immediately east of this, and
extends northwards along the eastern edge of the main exposure of Cudda-
pah rocks. The actual passage from one type to the other is gradual, but
on either side of a narrow interval the rocks differ decidedly in type.
There is more variety in the rocks of this band than in the red gneiss, but
the prevalent form is a rather rough gneiss of quartz, felspar, and hornblende.
The rock is obscurely foliated, but the foliation can generally be detected.

The boundary between the granitoid and the schistose gneisses is said
to be tolerably defined. The latter are distinguished by their distinct
foliation, and consist of hornblendic, micaceous, talcose, and chloritic
schists, and well foliated and more massive gneisses. No attempt was made
to map the two subdivisions separately, but the relative age was regarded

1 Memoirs, 1X, 301, (1872). 3 W. King, Memoirs, XVIII, 201, (1881).

2 R. B. Foote, Records, XVIII, 28, (1885). « Memiors, XVI, 126, (1880).
b
34 GEOLOGY OF INDIA—GNEISSIC ROCKS. {Chap. IL

as that given in the tabular statement. The gneiss No. 4 is said to occur
in a band separating the granitoid gneisses from No, 3, and to overlie
both. It seems probable, however, that an outcrop of the Dharwar system,
which had not been separated when the survey was made, has been included
with the metam orphies.

In the northern portion of this area a fifth form of gneiss is known,
which has been described as the Bezwada gneiss! from the tows of that name,
where it is typically developed. It forms a band along the edge of the
alluvial area from the Kistna northwards to the Godavari, and, from its posi-
tion as well as the less degree of metamorphism it has undergone, is
believed to be newer than the other gneisses. With the exception of thin
subordinate bands of quartz-schist and quartzose gneiss, the usual rock
is of a dark brownish colour, composed principally of lustrous red murchi-
sonite, which sometimes so predominates that there are seams, and even
thick beds, of felspar rock, the murchisonite being then granular. Garnets
are frequently scattered through the rock and often extremely abundant.
Here and there graphite occurs in sufficient abundance to convert the rock
into a graphitic schist,

These rocks appear to extend northwards along the coast as far as
Vizagapatam’, where they attain an extensive development, and, besides
the types of rock seen further south, contain some bands of crystalline
limestone,

In the southern part of this area important mica mines have lately
been opened in some very coarse granite intrusions. At Inikurti the
crystals are as much as Io feet in diameter, sheets of 4 or 5 feet across

-_

 

Fig. 3. Mica mines at Inikurti.
have been obtained, free from adventitious inclusions which would spoil

1W. King, Memoirs, XVI, 206, (1880). [ 7 W. King, Records, XIX, +19, (1886).
Chap. II.) MASSIVE GNEISS OF BELLARY. 35

their commercial value, but the size of the crystals can perhaps be best
realised from the photograph reproduced in figure 3.

In the south Marathd country, and in the Bellary district, the massive
type of gneiss, resembling that of Bundelkhand, is largely developed,
The outcrops form broad bands intervening between the strips of trans-
ition rocks, which rest on an eroded surface of the granitoid gneiss.
There are two principal types of this,’ first a fine or medium-grained red-
dish or grey rock, sometimes so homogeneous as to be hardly distin-
guishable from a felsite, whose most remarkable accessory mineral is
pistacite, occurring in veinlets and films lining the joints, and also as grains
scattered through the mass of the rock. The other type is coarse-grained,
often highly porphyritic, and is principally developed in the west, while
the finer-grained types are more common in the east of the district, The
foliation in the coarse-grained porphyritic rock is very obscure, and the
prevailing felspar a pink orthoclase.

Schistose-foliated gneiss is found in a band lying between the massive
porphyritic rock of Bellary and the band of Dhdrwdrs to the south-west, the
principal form of rock is quartzo-hornblendic, but no sections showing the
contact of this rock with those on either side were seen.

Numerous veins of granite, composed of quartz, red or pink orthoclase,
and pistacite, are found intersecting the granitoid gneiss near Gooty, and
the schistose.bornblendic gneiss further south.

The crystalline rocks in Bellary district are, as a rule, very unlike the
mass of the gneisses in the east and south of the Peninsula, but bear
a very strong petrographical likeness to the Bundelkhand gneiss. The
resemblance is not only a petrographical one in hand specimens, but
also a very striking one in the features of the landscapes of parts of these
two widely remote regions—a likeness abundantly confirmed by compari-
son of good photographic views of the granitoids in both tracts.

The long, narrow, serrated edges of quartz reefs, which form such
frequent and striking a feature in the Bundelkhand landscape, are nearly
as common in Bellary district and other parts of the Ceded Districts, and,
like those in Central India, they not unfrequently attain to heights of
500 to 600 feet and upwards above the general level. The granitoid
gneisses. in Bellary district and the adjacent districts of Anantapur,
Karnil, and Cuddapah, are also traversed by very numerous trap dykes,
of great size and length, which often rise into bold crests and ridges, form-
ing very conspicuous objects in the landscape. The relation of these to
the gneisses and to the great quartz reefs is precisely the same as in the
typical Bundelkhand area,

) The description in the text is based on | Records, XIX, 1€0, (1886) and a forthcoming

R. B. Foote, Memoirs, XU, 37, (1876); Memoir on the Bellary district.
D2
36 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. IT.

A further noteworthy fact, in perfect agreement with the geological
structure of the Bundelkhand gneiss, is the total absence, as far as our
present knowledge goes, of limestones in the gneissic region of Bellary,
thé south Maratha country, the Raichur Dodb, and the districts of Anantd-
pur, Karnal, and Cuddapah. Yet another point in which the granit-
oid gneisses of Bundelkhand and Bellary show a strong similarity, is
in the extraordinary rarity of accessory minerals—a point in which they
differ much from the gneisses in the eastern Carnatic.

The gneissic rocks of SAwantwdri and Ratnagiri in the Konkan, as
the low country between the Sahyddri range and the sea is called, would
seem, from Mr. Wilkinson’s description, to be more varied than on the
Deccan plateau above the ghats. The distribution in separate bands of
more massive and more schistose characters does not occur. The beds con-
sist of true gneiss (¢.e. a well foliated quartzo-felspathic rock), micaceous
and hornblendic schists, quartzites, and altered micaceous sandstones, with
some subordinate bands of granitic and syenitic gneiss, also occasional
‘talcose, chloritic and actinolitic schists. The mass of porphyritic syenite
forming Wajhiri hill, five miles from Vengurla, is considered to be intru-
sive.” .

In south Malabar Mr. Lake* has described three principal types of
gneissose rocks, which appear tobe of different ages. They are, firstly, the
quartzose gneiss, composed principally of quartz and hornblende, or some-
times mica, the other minerals are frequently absent, and thick bands of
pure quartz occur; large masses and runs of very hornblendic rock are also
found, and some beds consist principally of quartz and haematite, the latter
apparently due to the decomposition of garnets. Secondly, the garnetiferous
gneiss, usually a granular rock, with much quartz, a little felspar, and horn-
blende in varying, though seldom, large, proportion; garnets characterise the
rock, and are occasionally very abundant, but usually subordinate to the
quartz, Thirdly, the felspathic gneiss, in which felspar forms the principal
constituent, quartz also occurs, but there is very little of any other mineral.
‘There are also exposures of granitoid and hornblendic gneissose rocks which
appear to be intrusive and do not cover any large area in the district.

The distribution of the three types of gneiss presents some features of
interest. The quartzose type of rock is found in two areas, which are pro-
bably connected with each other in the intervening unsurveyed ground. The
strike of the foliation is nearly north and south. The quartzose gneiss
is succeeded to the west by the garnetiferous, and this again bythe felspathic

' This and the two preceding paragraphs are | clature between different observers in these
extracted from Mr. R. B. Foote’s unpublished | rocks; what some might call an aliered
Memoir on the Bellary district. micaceous sandstone others would name a

2 Records, \V, 44, (1871). Some allowance quartz-schist.
must be made for discrepancies of nomen-| 4 Memoirs, XXIV, 209, (1890).
Chap, IL] GNEISSES OF SCUIHERN INDIA, 37

gneiss, in which the foliation runs nearly east and west, but bends round
the outcrops of the quartzose gneiss, in a manner suggesting very
strongly that they are of different ages and that the latter is the oldest.
The foliation is usually coincident with the banding in all three types, but
where the east and west foliation bends round the nerth aud south foliation,
there are occasional discrepancies.

Our next note upon these rocks refers to the south-east Waindd, on
the uplands of Mysore at the north-west base of the Nilgiris. In the little
map published with Dr. King’s report? on this ground the greater part of
the area is shown to be within the region of the steady east-north-east
strike which obtains in the Nilgiris and along the south-east edge of the
Mysore plateau, but towards the north there is an area of troubled dips,
centred round two masses of granitoid rock ferming the Mani malai and
Yedakal malai, north of which the foliation again passes into the normal
north-north-west strike of the Sahyddri. Dr. King treats these granitic
masses as doubtfully intrusive. This Nilgiri strike is noted as distinctly
that of the lamination and bedding of the gneiss as well as of the folia-
tion, the general dip here being southerly. Four belts of gneiss are
recognised in the south Waindd, the quartzo-hornblendic gneiss of the
northern face of the Nilgiris, and below (north) of it the Devala band
of highly felspathic gneiss with two minor belts of chloritic gneiss.
North of this is the quartzose and ferruginous band forming the Mar-
panmadi range, beyond which is a broad area of more varied gneiss.
‘Ihe auriferous quartz-reefs are perhaps most developed in the Devala
band, ‘Their lie is peculiar, the strike is north-north-west, correspon:ing
with that of the gneiss in the country to the north, and at right angles to
that of the rocks in which they occur, yet they generally have a low dip,
from 10° to 30°, always easterly. One small trap dyke occurs in the
Devala band, running east by north, nearly in the strike of the gneiss,

In the Nilgiris? massive (obscurely foliated) gneiss prevails, but it
is of a very different type from the massive gneiss of the south Mardthd4
country, which is granitoid and copiously felspathic. It is in the very
hornblendic variety of the gneiss, such as prevails over the northern
portion of the Nilgiri plateau, that the foliation is least marked.
The rock is described as hard, tough, and black, breaking with an even
fracture and consisting of an intimate mixture of quartz and horn-
blende, with some garnets. It was taken by early observers for syenite
and greenstone. A similar rock, but with a variable proportion of felspar,
is very common in the central parts of the hills. There are also several
strong courses of a quartzo-felspathic gneiss, which has been taken

3 Records, VIII. 29, (1875). | 2 Memoirs, 1, 218, (5858).
38 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. TI,

for graphic granite.! Locally this gneiss also contains garnets in great
\juantity.

A few thin dykes of trap have been observed in the Nilgiri hills, but
no granitic veins. Small irregular veins of white quartz are common, but
no reefs have been observed,

To the south, as well as to the north, of the Nilgiris the gneiss of the
low ground becomes well foliated and schistose. South of Coimbatore
a band of limestone has been observed in the metamorphic rocks. Granit-
ie veins are also common in this neighbourhood, being especially
conspicuous in the hill of Sankari Drug, but no intruded granite mass
of large dimensions occurs. Mr. H. F. Blanford, from whom the notes
on the Nilgiris are taken, describes? a band of granitic rock to the
north of Trichinopoli,“and points out that this band is possibly a con-
tinuation of the very similar rocks of Coimbatore.’ In Trichinopoli, as
to the westward, there is no massive intrusion, but the whole band
(about four to six miles wide) may be considered rather as a network of
veins, running generally in the planes of foliation of a shattered band
of highly foliated hornblendic gneiss, which is frequently twisted and
contorted in every direction, The veins consist of a largely crystalline
binary granite, mica occurring but rarely. The proportions of quartz and
felspar vary greatly, and these ingredients sometimes affect the structure
known as graphic granite. Mica is altogether a rare ingredient in the
gneiss of this region of the Peninsula. : ee

A considerable area of the gneissic rocks of Southern India, from the
Cauvery northwards, has been mapped in some detail. The geology has
been described by Messrs. King and Foote, and the leading features have
been made out, or at least suggested.3 The belt of granitic intrusion
along the north bank of the Cauvery, already mentioned, is on an anticlinal
axis.- Beds of variable gneiss and schists, with some limestone, dip from
it on both sides. To the north they pass under the great mass of rocks
forming the several clusters of hills in the Salem district, where, as in the
Nilgiris, a syenitoid hornblendic gneiss is very prominent. With it
are associated the various magnesian schists from which the magnesite of
the “chalk hills” is derived * and also the great beds of magnetite which
have made Salem famous as an iron-producing district. These are not
lodes, but regularly bedded masses of banded iron-ore and quartz, associated
with the gneiss. With the aid of the very conspicuous outcrops formed
by this rock, several great features of contortion have been made out,
proving the strata to be frequently repeated at the surface.

1 Recent unpublished observations by Mr. | * Recent observations, as yet unpublished
Holland indicate the probability that some of | by Mr. Holland, show that the magnesite is
these are intrusive. probably derived from the decomposition of the

2 Memoirs, 1V, 30, (1863). ultrabasic intrusive rocks which he has detected
3 Menoirs, IV, 269, (1864). | in this district.
Chap. IT.) GNEISSES OF SOUTHERN INDIA. 39

o

A collection of rock specimens from the Salem district, made by Les:
chenault de la Tour ia 1819, has recently been made the subject of an
importaat study by M. Lacroix, who found among them an interesting
series of acid and basic gneisses, scapolite bearing rocks, and mica
schists. The greater part of his results are of a lithological and mineral-
ogical nature, too isolated in their bearing on Indian geology to bear
summarising here, and his suggestion, that the specimens belong to three
separate stages, correspondiag to those known as {}, «? amd « of the
geological map of France, yet waits the confirmation of detailed work
in the field. The paper is, however, of importance, apart from the valu-
able results it contains, as showing how much remains to be done in the
study of the Indian gneissose recks.

In South Arcot, to the east of the Salem hill-groups, a considerable
area is occupied by rocks having a very granitic aspect, yet showing in
many placesan appearance of stratification and eccutring in great con-
tinuous ridges, which apparently form anticlinal and synclinal folds. The
rock is composed of quartz and white and pink felspar. It frequently
contains blocks, both angular and rounded, of hornblende schist. Alto~
gether, the nature of this reck and its position in the metamorphic series
are still open questions.

North of Trichinepoli a change takes place in the direction of the
strike of the metamorphic foliation analogous to that noticed in the
Waindd. The east-north-east direction changes rapidly into north-north-
east, parallel to the Coromandel coast. The regularity of the coast-line
is no doubt connected with this fact, and it is interesting to note how the
main structural features of the fundamental rocks thus determine the
actual configuration of the Peninsula, All the fossiliferous deposits,
and even the later azoic formations, are but patches on the weather-worn
surface of this most ancient gneissic mass,

In the Madura and Tinnevelli districts the gneissose rocks are of
two types,? one the ordinary granitoid gneiss, the other described as
granular quartz rock, and Mr. Foote believed that he had recognised no
less than six alternations of these in the northern portion of the Madura
district. Such supposed stratigraphical successions, however, in rocks that
have been so highly metamorphosed and disturbed as these, are open to
grave elements of doubt. Both micaceous aud hornblendic types of gneiss
are found, granite intrusions are not abundant, quartz veins are rare,
and trappean intrusions almost entirely absent. A few exposures of
coarsely crystalline limestone in the gneiss are recorded.

\Bull. Soe. Frang, Minéral., X11, 83, (1889). | XXIV, 155*166, (1891).
Translated by Mr, Fe R. Mallet, Records, 22. B. Foote, Memoirs, XX, 10, (1883).
40 GEOLOGY OF INDI{A—GNEISSIC ROCKS. (Chap. Il.
A few words may here be said regarding the distribution of the trap
dykes which traverse the gneisses of southern India. They are extremely
rare in Tinnevelli and south Travancore, and very few are seen in Madura,
Pudikattdi, south Trichinopoli, or in the band penetrated by granite
veins north of the Cauvery. In north Trichinopoli, Salem, and South Areot
the number increases, but in Coimbatore they are not numerous On the
Nilgiris they are few and small, and only one is known in the Waindd. In
North Arcot they are very numerous, and often large and of great length,
they continue to be numerous in south-west Cuddapah and Anantdpur,
becoming less so in Bellary and Karnél. In Chengalpat, Nellore, and
Kistna they are not common, and none very large.

In composition the trap is dioritic, usually of medium grain, though
both very coarse and very fine-grained varicties are met with. The in-
trusions are of very ancient date, and probably connected with the vol-
canic outbursts of the Cuddapah system. The ancient volcanic neck of
Wajra Karur, a notice of which will be found in a subsequent chapter,

may be connected with them.

Turning northwards to the Aravalli range, we have very few observa-
tions recorded, regarding the gneissose rocks. In some exposures
in the northern part of the range and on its eastern margin, beds
belonging to one of the transition systems are said to rest unconformably
on granitoid gneiss,’ but in the central portion of the range the contact
is said to be transitional and accompanied by an_ interstratification
of the transition beds with the gneiss. It is not impossible that there is-
gneiss of at least two distinct ages in the range, but the apparent trans-
ition is probably due to the disturbance the beds have undergone, and
the apparent interstratification to veins of gneissose granite intrusive
along the planes of bedding. True granite is known to occur abundantly
in the Ardvallis.

The gneiss of Mount Abd and its neighbourhood is said to be highly fels-
pathic, massive, and crystalline, but occasionally a few schistose beds occur.

At the southern end of the range, in the lower Narbada valley, near
Jobat, well foliated gneisses, with quartz and hornblende schists, are promi-
nent, while mica schists and granitoid rocks are rare. Limestone and
slates occut among the metamorphics, but whether the association is real,
or only due to the disturbance they have undergone, is not clear.®

In extra-peninsular India the gneissose and granitoid rocks occupy a
smailer proportion of the total area and, except in a small portion of the
Himalayas, have been even less studied than in the Peninsula.

1C. A, Hackett, Records, XIV, 27¢, (1881), 2 PLN, Bose, "Memoirs, XXI,7, (1884).
Chap. II.] GNEISsIC ROCKS OF THE HIMALAYAS, 4t

Of the gneissose rocks of Afghdnistan we have very little information,
They are known to occur in the Safed Koh and in the Hindu Kush, and
probably occupy the greater part of Kafiristén, up to the region of the
Pamirs.

In the Himalayas the few observations that have been recorded are
often of a misleading nature; owing to the difficulty there often is in
distinguishing between true gneiss which has become so metamorphosed
as tobe passing into granite, and granite which has assumed the foliated
structure characteristic of gneiss. The broad features of the distribution
of the gneissose rocks in the north-western portion of the Himdlayas
are tolerably known, and it is found that the area occupied by them
not only corresponds to the regions of special geological elevation,
but is approximately coincident with the principal ranges of high peaks,
the coincidence being due quite as much to the comparatively greater
resistance offered to denudation by the crystalline rocks, as to the fact
of these being areas of greater special upheaval.

In the region north-west of Kashmir, where the Himdlayas and Hindu
Kush meet, there is known to bea large extent of gneissose rocks, associated
with some schistose slates and crystalline limestones, but no lithological
details of importance are available.

In Kashmir territory the boundaries of the crystalline rocks have been
approximately mapped, and the rocks themselves cursorily described by
Mr. Lydekker,' but his description, being based on rapid traverses, contains
tew details of sufficient importance and certainty to be repeated here, and
fails moreover in always distinguishing between the gneissose granite
and granitoid gneiss. Mr. Lydekker believed that the gneissose rocks
were of two ages, —for the one, the older, he retained Dr. Stoliczka’s name
of ‘‘ central” gneiss, while the newer gneiss was regarded as conformable
with, and in part formed by the metamorphism of, the sedimentary unfossil-
iferous slates, believed to be of silurian age, so largely developed in the
Himdlayas. It is more than probable that the Himalayan gneiss, like that
ofthe Peninsula, belongs to more than one system, and is not all of one age;
but the second conclusion, though in consonance with a somewhat discredit-
ed theory of metamorphism, is open to doubt. ‘The apparent interstratifica-
tion of gneiss with schistose slates and limestone appears, in at least some
instances, to have been due to foliated granite, intruded along the bedding
planes, and the apparent complete metamorphism of the slates into gneiss
is probably due either to a complete obliteration of the sedimentary rocks
by a gneissose granite, or to a misinterpretation of the scattered observa-

“tions,

1 Memoirs, XXII, 265, (1883). It must be re- | widely separated traverses of the country
membered that the observations were made on | which has not yet been examined in detail.
42 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. IT.

It is certain, however, that in Zanskar, Ripshu, and the gneissose area
which extends along the main axis of the Himdlayas north of Kumaun, true
gneiss is largely devoloped. In places this is well foliated, and exhibits a
succession of parallel layers, of differing mineral constitution, which are
strongly suggestive of an origin by some process of sedimentation.’ But
often the gneiss is extremely granitoid, with the bedding very obscurely
exhibited, and it is at present uncertain whether this is merely the
result of a more advanced metamorphism of the bedded gneiss or indicates
a difference of age.

No careful and critical examination of these gneisses has been made as
yet. The principal minerals are quartz, orthoclase, and mica, in varying
proportions, but plagioclase felspars, schorl, garnet, and kyanite are not
uncommon accessory constituents, In some of the beds purphyritic crys-
tals of orthoclase are found, of lenticular shape with a well developed crys-
talline structure, exhibiting universally a twinning of the Carlsbad type,
with the plane of twinning coincident with that of the foliation,

The gneiss of the Ddrjiling district is of the foliated type, apparently
a true gneiss? composed of translucent colourless quartz, opaque white
felspar and dark brown and silvery mica, varying in texture from fine-
grained to moderately coarse. The gneiss frequently passes into mica
schist or felspathic mica schist, bands of quartzite are rare, while horn-
blendic rocks are extremely uncommon and in beds of insignificant
thickness. Almost the only accessory minerals are kyanite, schorl, and
garnet, the two last often forming large-sized crystals in the schists.

The well-bedded gneisses seen on the ascent from the Sutlej valley
to the Babeh pass, were originally classed by Dr. Stoliczka® with his
‘central” gneiss, which he considered as forming the original central
axis of the Himdlayan chain, on either side of which the subsequent de-
posits were accumulated. The name has, after being current for many
years, beeu abandoned of late, as it implies a theory which, to say the
least, has not been proved, and because a granitoid porphyritic rock,
which was included with these bedded gneisses and regarded as the
typical member of the system, has since been shown to be a true granite
in its mode of occurrence.

This rock consists of porphyritic crystals of orthoclase imbedded in a
fine-grained matrix composed of quartz, orthoclase, some plagioclase
felspar, and mica, both biotite and muscovite, as well as crypto-crystalline
mica whose exact mineralogical species is indeterminable, with magnetite,
garnet and schorl as accessory minerals. The larger masses of this rock

1 It is now well established that the apparent | beds themselves occasionally show palpable.
stratification of the crystalline gneisses is | indicatious of their detrital origin.
often deceptive. In the cases reterred to in | ? F, R. Mallet, Memcirs, XI, 43, (1874).
the text this cannot always be the case, for not | 3 Memoirs, V, 15, (1865).
only are.acres of surface of a single bed occa- | # C, A, McMahon, Records, X, 22; XV, 34
sionally exposed on the steep hill-sides, but the | XVI, 129; XVII, 53, 168; (1877-84).

x
Chap. ID] GNEISSOSE GRANITE OF THE HIMALAYAS. 43

exhibit slight traces of foliation, generally recognisable in the field,
but barely, if at all, in hand specimens. Besides the larger, slightly
foliated, intrusions it is found in sheets, generally intruded along the planes
of bedding, and traceable at times for distances of 20 miles with a fairly
constant thickness of asmany feet. Under these circumstances, it has a
remarkable resemblance to a truly interbedded rock, a resemblance en-
hanced by the well-developed foliation, parallel to the bounding sur-
faces, which these thin sheets invariably exhibit. But there is one con-
stant character which marks both the thin sheets and the larger, less foli-
ated, masses as intrusive, and that is the presence of numerous porphyritic
crystals of orthoclase, one and all twinned on the Carlsbad type, varying in
size from half an inch to as much as six inches across, and lying with their
axes pointing in every direction.

‘The small amount of disturbance with which the intrusion was
accompanied, is in many cases remarkable. It is especially conspicuous in
the case of the Chor mountain, which rises to a height of nearly 12,000
feet, 25 miles south-east of Simla. All round this mountain the stratitied
rocks are comparatively little disturbed, dipping inwards on all sides,
and in the centre of this quaquaversal dip the granite has risen, as if
it had simply dissolved and absorbed the rocks whose place it occupies.
There is some direct evidence that such really was the case, for in the
south-eastern portion of the intrusion, where it has replaced volcanic
beds of the carbonaceous system, the fine-grained matrix is coloured dark
green from the amount of hornblende it contains, a mineral which is
usually conspicuous by its absence. Even the thin sheets appear to have
been intruded to some degree in the same manner, for the further they
are traced from their parent mass the more micaceous, that is to say the
more impure, do they become, till the mica in seme cases becomes so
abundant that the rock splits easily into large flags. This contami-
nation of the rock only extends to the matrix, and the porphyritic ortho-
clase crystals are unaffected, showing that they had already solidified, and
crystallised out from the still fluid or pasty magma, when the rock was
intruded into its present position,

A similar porphyritic granite is found in Hazéra, where it has been
graphically described by Mr. Wynne,} it occurs in the core of a highly
compressed anticlinal fold in the Pir Panj4l and Dhaola Dhdr ranges,
and extends eastwards along the ranges bordering the Sutlej valley.
It is also found in the gneissic area to the north of this line, associated
with the true gneisses of Zanskar and Rupshu, and, to the south of the
main axis of intrusion, is found in the Chor mountain already referred to,
and in the Dudatoli mountain in Garhwal. Intrusions also occurin Kumdun

which have not yet been mapped in detail.

1 Records, XII, 118, (1879).
44 GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. II.

The date of these intrusions is still undetermined, and the evidence is
contradictory. On the one hand, we have the recorded occurrence of blocks
of it in the so called Panjél conglomerate of Kashmir,’ but on the
other, the intrusive masses of the Pir Panj4l and Dhéola Dhar ranges must
have been in a plastic condition subsequent to the period of deposition of
the beds in which these boulders were imbedded: The probable explana-
tion is that the boulders were derived from an ancient land area, composed
of a rock very similar to the porphyritic granite of the Dadola Dhdr, and
that in the great compression and disturbance which caused the elevation
of the Himalayas this rock was once more fused and intruded into the

position where it is now found.

Another very characteristic intrusive rock of the Himalayas is a white
granite, occurring in veins of various sizes and degrees of coarseness ct
texture, The granite consists of quartz, white felspar, which in at least
one instance is oligoclase,? and muscovite. Schorlis a very common acces-
sory constituent, and beryl, fluor spar, and garnet are found, The rockis a
very common and conspicuous one along the principal axis of the Himdlayas,
occurring in intrusive masses and innumerable veins, tamifying through
the gneiss and schists, and even penetrating the slates. Its intrusion
appears to have been, generally, of later date than the porphyritic granite,

Syenite is largely developed in the range north of the Indus and west
of Leh, where it is unconformably overlaid by the eocene beds of the
Indus valley. Another outcrop of syenite is known, east of Chakréta, in
Jaunsar, but this is intrusive in the slates, and is very probably of later
origin than the Ladakh rock.

From a geographical point of view, Assam and the Shillong plateau
could not be affiliated to the Peninsula, but, geologically, this would
appear to be their proper connection, since the prevailing rocks closely
resemble the gneissic and transition’ formations of Bengal, and differ
widely from the rocks of the adjoining mountains to the north and east.
The structural characters bear the same relation ; on the edges of the
Shillong plateau secondary and tertiary strata lie quite horizontally.
while much younger deposits have undergone intense disturbance in the
contiguous Himdlayan and Burmese regions. The plateau thus forms a
wedge-like mass of neutral ground, occupying an acute angle between
two regions of contortion.

The ground to which these remarks apply is not known to extend
north-eastwards, beyond the Dhaneswari (Dhansiri) river, though it is likely
that the gneissic rocks stretch for some distance under che alluvium of

oo XXII, 280, (1883), first incorrectly described as albite. SeeF R.
emoirs, Vy 14, (1865). The felspar was at Mallet, Records, XIV, 238, (1881).
Chap. II.) GNEISSOSE GRANITE OF THE HIMALAYAS. 45

upper Assam. The principal area is the continuous hill-mass, 250 miles
long, between the Dhaneswari and Brahmaputra. The whole of the lower
Assam valley may be included in the same geological region, for the numer-
ous hills protruding through the alluvium, north of the Brahmaputra, consist
of the same gneiss, and not of the Himalayan type of metamorphic rocks.

The most interesting of these outcrops in the low ground of the Brah-
maputra valley is one observed by Mr, Mallet! within 200 yards of the
tertiary sandstone at the base of the Himalayas, on the left bank of: the
Raidak river, in the western Bhutan Dwdrs. It is really within the sub-
Himalayan zone, being up a river-valley, inside the mean outer boundary
of the sandstones. The rock is thick-bedded hornblende-schist, a com-
mon type of rock in the Bengal gneiss, but one that is rare in the Darjil-
ing gneiss of the adjoining mountains. This is the only instance of close
proximity of the azoic rocks of the Peninsula to the Himalayas,

The only observations hitherto made on this Assam gneiss prove little
more than that it has a likeness to the Bengal rock, and that the general
strike is the same. Some granitic intrusions occur in the transition rocks
of the Shillong area, in connection with which they will be noticed.

Gneissose rocks are largely developed in the hill ranges which run
southwards from the eastern termination of the Himdlayas, to the east of
Burma and Tenasserim, but very little is known of their constitution.
More or less granitoid gneiss, hornblendic gneiss, crystalline limestone,
quartzite and schists of various kinds are found. In many places the
gneiss becomes a true granite, and much of the area is occupied by a rock
which has been described by various observers as an eruptive granite.
Some of the granitoid portions of the rock weather into remarkable
rounded masses resembling perched blocks, isolated from each other
by the disintegration of the intervening rock.? The hornblendic gneiss
appears to be less common than in the Peninsula of India, while crystal-
line limestone is not uncommon.

These gneissic formations are known to be metalliferous in several
places. Tin occurs in abundance in Tenasserim and the Malay Peninsula,
lead and silver mines—one of them at least, the famous Bau-duen-gyee,
of very large dimensions and highly productive—exist in the Shan States
north-east of Ava, while the valuable and productive ruby mines of
Mogouk and the less important ones nearer the capital, are situated in the
same series of metamorphic rocks.

From the foregoing brief account of the present state of our knowledge
of the gneissose rocks of India, some idea of its imperfection can be

¥ Memoirs X1, 44, (1874). Four. As. Soc. Beng., XXX\V, pt. 2, plates
2 Sketches of these are given by C. Parish, | vi, vii, viii, (1865).
46, GEOLOGY OF INDIA—GNEISSIC ROCKS. (Chap. IT.

gathered. As was stated in the opening paragraph of this chapter, the
attention of the Geological Survey has, with few exceptions, been only
incidentally devoted to the gneissose rocks, the exceptions being almost
exclusively surveys made at an early date, when the true nature and
origin of these rocks was far less understood than at present, The
analogy of other countries makes it impossible to doubt that results of
great importance and value will be obtained when the great gneissose area
of India conies to be examined in detail,

In the meanwhile there seems to be only one thing certain, that what
has been described as the older or Bundelkhand gneiss had already solidi-
fied, acquired most if not all of the characteristics it now has, and been
exposed to extensive denudation, at a period anterior to that of the oldest
of the distinctly sedimentary rocks of the Peninsula. How far the banded
gneisses are newer than this, or how far they are the result of subsequent
changes of the older gneiss, it is at present impossible to say. The re-
searches that have been made in other countries renders it almost certain
that, in many cases, what has been described as stratification in the gneiss
was not’ produced by any process analogous to sedimentation, but it isa
result of the deformation and other changes set up in the rock by pressure,
heat, and the intrusion of foreign igneous rocks, The recognition of this
will necessitate a profound modification of many of the passages in this
chapter, a modification whose extent and limit it is not yet possible to
Indicate.
CHAPTER III.

TRANSITION SYSTEMS,

Transition systems— Difficulty of classification— Definition of term—The Dhdrwdr system—The
Bijdwar system—A possibly older system—Transition rocks of Behar—of the Assam
hills—of south-west Bengal—The Gwalior system—The Ardvalli system-- The Delhi
system—Kirdna and Chiniot hills— Champ4ner beds—MalAni rocks—Relative ages of the
Transition systems—Transition rocks in extra-peninsular India—The Vaikrita system—
The Daling series.

Resting upon the crystalline gneisses, and intervening between them and
the oldest fossiliferous beds of the Peninsula, there are a number of rock
systems whose age it is impossible to determine with certainty, whether
relative to the European sequence or, in many cases, relative to each other.
This difficulty is due to the complete absence of any recognisable fossil,
an absence which in many cases is easily explicable by the disturbance
and partial metamorphism the beds have undergone, but equally often they
have undergone little or no disturbance, and are apparently admirably
fitted for the preservation of the remains of such animals as may have lived
when they were deposited.

Failing fossil evidence, we are compelled to fall back on lithological
characters and the degree of disturbance or metamorphism the beds have
undergone, and on these grounds they can be divided into a number of

‘separate systems, representing different periods, or at any rate different
areas, of deposition, and these systems can again be divided into two
groups,—an older which, as arule, has undergone considerable disturbance
and some degree of metamorphism, and a newer which, as a rule, shows
only gentle dips and a much less degree of alteration than the other. The
characters are by no means constant; the Gwalior rocks, which we will
have to include among the older group of systems, are almost undisturbed,
while the Cuddapahs, which will fall among the newer, are in places highly
compressed and contorted. Onthe whole, however, the distinction appears
to be valid, as it certainly is convenient for the purposes of description,
though it isimpossible to say how far the eras represented by the two may
not have overlapped each other.

The newer of these groups of rock systems we can, with some degree
of propriety, class as older palzozoic, and for the olderthe name ‘transition’
48 GEOLOGY OF INDIA—TRANSITION SYSTEMS. (Chap. It,

has become customary. This word is not intended to imply any theory
as to the mode of origin of the deposits, but merely to indicate their posi-
tion, intermediate between the older crystalline gneisses and the newer
sedimentary deposits, which have undergone a less degree of disturbance

and metamorphism.

The most recently recognised of the transition rock systems, though
probably the oldest in point of age, is that which has been described by
Mr. Foote under the name of Dhdrwdr system.’ As at present seen, it
occupies a series of long bands and elongated outliers of highly disturbed
beds, folded and faulted into the gneiss, which have a general north-north-
west and south-south-east strike, and extend from the southern limit of
the Deccan trap to the Cauvery valley, south of which only a few small
outliers are known. ‘Jhese qutcrops exhibit two types of structure. They
are either a sharply folded synclinal or series of synclinals, or else, show-
ing on one side, usually the west, a natural boundary of original contact,
they are cut off on the other bya fault of great throw, by which the softer
beds of the upper Dhdrwars have been dropped down, and so preserved
from entire removal.

The most westerly of these bands commences in the north as a series of
inliers in the Kal4dgi basin north-east of Belgaum. From the southern
edge of. this basin it runs down, with a-width of 10 to 16 miles, past Dhdar-
war to the Tungabhadra river. Here it spreads out and covers a large area
in north-western Mysore, sending one offshoot down to within 40 miles of
the city of Mysore, while another may possibly run southwards from Shimoga
down into the low country of south Kdnara. The next of these bands
starts at the southern boundary of the Kalddgi rocks and runs by Dambal
and Chitaldrég to Chiknayakanhalli, South of this the band bifurcates,—one
branch extends to Seringapatam, the other, running somewhat east of south,
crosses the Bangalore-Mysore railway. A third band runs along the north.
east boundary of the Kalddgi rocks, east of Bellary, and on to the Penner
river, Between this and the last-mentioned band there is a tract of
Dharwar rocks forming the. Sanddr hills and copper mountains west of
Bellary. A fourth band runs southwards through the Shor4pur district
past Maski to near the Tungabhadra,

Besides these larger bands there are a number of outliers, the most

4important of which are those which form the gold-tields of Kolar, and the
three bands, further east, which run out from under the Cuddapahs.

The rocks of the Dhdrwdr system are hornblendic and chloritic
schists, phyllites and conglomerates, associated with contemporaneous trap,
banded jasper, and hematitic quartzites. The degree of metamorphism

1The principal published descriptions of | the text are principally based on the unpub-

the Dhdrw4r system are in Records, X XI, 40-56, | lished Memoir on the Bellary district by
(1888), and XXII, 17-39, (1889). The notes in | Mr. R. B. Foote,
,

Chap. IIL.] THE DHARWAR SYSTEM. 49

they have undergone appears to vary considerably and to be connected
with the varying and often extreme degree of compression they have
undergone, Generally speaking, the beds are distinctly schistose and fre-
quently well-characterised schists, but in places they are described as are
gillites, which readily weather into soft shale. The dioritic traps, which are
usually found conspicuously developed in the lower part of the-series, are
replaced by hornblende schists on some of the highly disturbed sections,
and the pebbles in the conglomerates have occasionally been deformed
into long rod-like forms.

Conglomerates are of frequent occurrence at and near the base of the
system, some being of the ordinary type of true conglomerate, others of
that type, consisting of boulders scattered through a fine-grained matrix,
for which the name of ‘boulder beds’ has been suggested. The included
pebbles and boulders consist of various varieties of schist, quartz, quartz-
ite, grit, banded hornstone, and gneiss.

The hzmatitic quartzites are composed of alternating layers of quartz-
ite and hematite in proportions varying from a rich, pure, hematite iron
ore on the one hand, to a banded hornstone containing little or no
iron on the other. These hematite beds are everywhere found in the
lower portion of the system, and owing to their hardness, they stand
up as sharp tidges from the softer schists, which have been denuded away
from their sides, (hus acquiring a conspicuousness quite out of proportion
to their real importance, ‘hey are specially abundant in the Sandir state,
where there are vast supplies of a rich, pure, hematite iron ore, formerly
worked to a considerable extent, by the natives of the country. The
industry is now almost extinct, as much in consequence of the reckless
destruction of forests as the competition of imported iron.

‘lhe Dhdrwér system is economically important as it carries all the
paying gold reefs that have yet been discovered in Southern India, all the
known gold fields being on the Dhdrwér outcrops, and all trials outside
them having so far led to nothing but disappointment. The gold appears
to be most abundant in the outcrops situated in Mysore territory, where
gold-mining has, within late years, passed through the phase of fierce
speculation into a well-established and remunerative industry. The reefs
have been mined in prehistoric times by miners, whose workings, in spite
of their primitive appliances, penetrated to depths of over 200 feet in
places and, besides the abundant mines, old dressing floors can still
be found, with the mortars in which the quartz was crushed, generally
small hollows in which the quartz was pounded, but occasionally large”
saucer shaped depressions in which huge blocks of granitoid gneiss of a
ton and more in weight were rolled round and round.

The relation of the Db&rwdr system to the granitoid gneiss is one of
most unequivocal unconformity. Wherever a section showing the original

E
59 GEOLOGY OF INDIA—TRANSITION SysTeMS.  { Chap, III.

contact is found, the bottom beds of the Dhdrw4rs are found to rest on an
uneven eroded surface of the granitoid gneiss. Yet gneissoid beds are
occasionally found in the Dhdrwérs and a section, east of Memkal in Bellary,
is recorded, where several alternations of micaceous gneiss with quartzites
and hornblende schists areseen. The section is, however, exceptional, the
gneiss is described as differing more from the typical granitoid than from
the Dhdarwdr schists with which it is stratigraphically connected, and it is
possible that the interbedded gneiss may be metamorphosed arkose. There
is of course the alternative interpretation that the Dharwdrs are here
locally in contact with a gneiss newer than that to which they are so dis-
tinctly unconformable. /

In a northerly direction the Dhdrwdr beds of the central and western
bands run under the great spread of Deccan trap, but not before their
upturned and denuded edges have been seen to be unconformably covered
by the Cuddapah beds of the Kalddgi area, and to the south-east a similar
relation subsists between the Dhdrwdrs and the typical Cuddapahs of the
Cuddapah area. The Dhdrwdr system is thus completely isolated, both
geologically and geographically. The unconformable breaks above and
below are so great that they indicate nothing more than the necessity for
an utter separation of the system from any other occurring in its neigh-
bourhood, while the distance which separates it from the north of the
peninsular area is too great to allow of its identification on mere litho-
logical grounds with any of the transition systems there seen. Yet these
are the only ones on which we can venture even a guess at its correlation
and, so far as they go, the resemblance is greatest_in the case of the
Gwalior system; contemporaneous dioritic traps occur in both, while the
nematitic. quartzites of the Dhdrwdrs resemble the hzmatite beds of the
Gwaliors and not those of the BijAwars in their structure. On the other
hand, the small amount of metamorphism or disturbance which the Gwaliors
have undergone sharply distinguishes them, and it is probable that both
they and the Bij4wars are newer than the Dhdrwérs of Southern India,

There can be little doubt that, when the gneissic area of the Madras
presidency is more fully surveyed, other outcrops of Dharwar rocks will be
discovered. At present one, at least, can be indicated in the Nellore Car-
natic, where a distinct band of more eminently schistose rock! is said to
occupy the western edge of the field, the schists being talcose, mica-
ceous and chloritic, with frequent intercalations of hornblendic bands.
Interbedded quaitzites are common, and a laminated hzmatitic quartz
schist occurs south of the Swarnamukhi. Associated with these more foli-
ated rocks there is a development of trap, both as dykes and as a large
irregular mass of diorite and greenstone, which was regarded as intrusive
on the whole, but in certain cases? distinctly stated to be interbedded. This

! Memoirs, XV1, 133, (14880). | 3 Mencot1s, XVI, 150, 1€8, (1880).
Chap, IIL] THE BIJAWAR SYSTEM IN BUNDELKHAND. 51

association of rocks is strongly suggestive of the presence of an outcrop
of the Dharwar system, which had not been separated from the crys-
tallines at the time the survey was made.

A good deal of doubt attaches to the mapping of this area owing to
the occurrence of what appear to be outliers of true Cuddapah quartzites,
and the difficulty of distinguishing between the less altered and disturbed
Dhdrwdr quartzites and those of the Cuddapah system, where they have
undergone much disturbance and alteration. Between the Penner and the
Swarnamukhi, the narrow strips of quartzite appear all to be associated with
contemporaneous traps and schists, and are probatly Dhd4rwdrs. North of
the Penner there seems good reason for supposing that they are Cuddapahs,

Turning now to the northern part of the Peninsula, we find a great
system of transition recks, which has been distinguished under the name of
Bijawar, from the town of that name in Bundelkhand. By far the greater
patt of the area, over which these rocks originally extended, is now covered
up by the newer Vindhyans and Deccan trap and they are only exposed in
a series of outcrops, of varying size, which extend from Bundelkhand to
south of the Narbadé, a distance of about 1oo miles from north-north-west
to south-south-east, and from Jobat to the upper Son valley, some 500 miles
from west-south-west to cast-north-east.

The commonest bottom-rock of the Bijdwar system in Bundelkhand,
is a quartzite! that might locally be called a sandstone. It is generally fine-
grained, but sometimes, at the base, coarse and conglomeratic from con-
taining pebbles of white quartz. It rests quite horizontally, or with a
slight dip, upon a denuded surface of the gneiss, even in that most western
part of the area, where the uppermost portion of the Bundelkhand
gneissic series is supposed to be found.?

With this quartzite a hornstone-breccia and a limestone are intimately
associated. They sometimes replace the quartzite as the bottom rock, oz
else are interstratified with or overlie it. The hornstone is compact
quartz, more or less transparent or opaque, of yellow, brown, and red
tints, the angular fragments included in it being generally of white
quartz, and always paler than the matrix, In some cases, if not in all,
they are clearly not the result of fracture caused by contortion, for the
breccia mostly lies quite flat upon a firm support. Occasionally the
former continuity of the detached pieces is evident, the mass looking as
if thin bands of quartz had been shattered by concussion or shrinkage,
then re-cemented in place, and the interstices filled by a more jaspideous
form of quartz. The limestone, too, is highly silicious, the quartz appear-
ing both as thin layers and as shapeless, irregular segregations of chert,

t Memoirs, U1, 11, (1860), | ?Subra,p 27.
E 2
52 GEOLOGY OF INDIA—TRANSITION SYSTEMS, [Chap. III.

These bottom rocks of the Bijawar system in Bijdwar are very irre-
gular in their distribution, In some sections there is no quartzite, in others
no hornstone-breccia or limestone. The total thickness nowhere exceeds
2oofeet. This unevenness of the basement-bed tends to suggest the un-
conformity of the succeeding deposits, but no confirmation has been found
of this suggestion, on the contrary, sub-schistose shales, like those of the
upper part of the system, are sparingly intercalated with the limestone and
quartzite.

More or less earthy ferruginous sandstone, locally somewhat conglo-
meratic, is the prevailing upper rock, and is associated with incipiently
schistose rusty shales. The iron in these rocks is locally concentrated into
a rich hematite which has been extensively worked. Several thick, but
discontinuous, beds of dioritic trap occur in the bottom part of the
group.

The whole Bijadwar system in the typical Bijawar area is probably
not more than 800 feet thick. The strata generally either have a very low
south-easterly dip, or are quite horizontal, but in a few places to the south,
before they become covered up, they are seen to have undergone a consid- -
erable amount of crushing, which has not in the least affected the lower
Vindhyan rocks immediately overlying. The general immunity from dis-
turbance in this small area may be due to the original shallowness of the
deposits here, where they thinned out over the mass of gneiss, which
afforded an unyielding support against compression. It is probable that the
transition basin deepens rapidly to the southward, beneath the Vindhyan
rocks, and that the complete unconformity between the Bijawars and
the lower Vindhyans, as observed in the Son valley, rapidly replaces the
general parallelism of stratification that obtains in the Bijdwar area. East
of the Ken the Bij4war rocks soon disappear, being totally cut out by
the Vindhyans overlapping on to the old gneiss. From a little west of
Allahabad all the lower azoic rocks are concealed by the Gangetic allu-
vium, which stretches up to the base of the Vindhyan scarp.

At the Ken the character of the bottom Bijawar rocks changes rapidly,
the strong quartzite thins out suddenly, and a prominent rock, on the
continuation of its strike, is a peculiar, sharply cellular quartzite, much
quarried for quernstones, but the beds associated with this quartzite
are sandstones and shales like those of the upper part of the series. In
the river, and certainly below the horizon of the bottom quartzite of the
Bijawars west of the Ken, there are two or more steady outcrops of pebbly
sandstones, having the same low south-easterly dip as the adjoining Bija-
war strata, but occurring in the midst of thick pseudo-crystalline gneissic
rocks. It is important to notice these observations with a view to their veri-
fication or correction, for these sandstones seem to have escaped the notice
of the later observers, and they are important as fixing the allinities of the
Chap. ITI.] BIJAWAR SYSTEM IN THE NARBADA VALLEY, 53

associated gneissic strata with the transition series rather than with the
normal gneiss of Bundelkhand. Very similar rocks are found far to the east
in an analogous position at the base of the transition series in Behar, and
again extensively in the Ardvalli region, and the question arises as to
whether we must not recognise in the great gneissic series some rocks that
are not metamorphic in the full sense of the word, but are merely recon-
solidated granitic or gneissic detritus.

Proceeding from Bijd4war in a south-west direction obliquely across
the plateau, where the Vindhyans are for the most part covered by the
Deccan trap, we should strike the Narbadd about Handid, at the west end
of the wide alluvial plain, 200 miles long, which, in India, is specially desig-
nated as the Narbadd valley. West of Handid there is a considerable
area occupied by transition and gneissic rocks. They abut on the west
against the Vindhyan rocks of the Dhar forest area, but appear again in
the north of this area, and west of it about Bdrwai. These transition
strata have been fully recognised by Mr. Mallet,' as bottom Bijdwars, con-
sisting of quartzite hornstone-breccia and chert-banded limestone, identical
with those of Bundelkhand. No associated trap rock was observed.

These rocks are more disturbed here than in Bijdwar, but Mr. Mallet
describes their relation to the gneiss as the same, the quartzite being
often found quite flat and surrounded by vertical strata of the metamor-
phics. It is only possible to question this view by supposing that what
we take to be stratification in the metamorphics is a result of molecular
forces acting on lines of cleavage. This possibility has been forcibly
argued with reference to this very area, and is connected with the sugges-
tion that the two series may be very closely allied, the gneiss being more
or less a metamorphic condition of the Bijawars.

Upon the settlement of this question as to the relations between the
metamorphic and transition series, it will depend whether the gneiss of
the Dhd4r forest should be affiliated to that of Bundelkhand, or to the sup-
posed younger gneiss of Bengal. The composition of the Dhar forest
gneiss is in favour of the former, as well as the relation stated to subsist
between it and the Bijawars.

Here, as so often elsewhere, a doubt occurs as to the intrusive charac-
ter of the more granitoid varieties of the gneiss. Some hornblendic and
earthy schists of this area, as above Mortaka where the Indore railway
crosses the Narbad4, have been included with the gneiss, but it may be
questioned if they do not belong to a transition group older than the
Bijdwars. . . aimee

In the lower Narbadé valley the Bijawar formation has been recognised

1 Memoirs, VI, 199, (869). 2 Memoirs, V1, 200, (1869); XXI, 14, (1884).
|

|
|

eA GEOLOGY OF INDIA—TRANSITION SYSTEMS. (Chap, IIT-

specifically identical with the beds in the Dhdr forest, near Bagh and
Jobat, the two localities being separated by 80 miles of Deccan trap-

All the most charaeteristic rocks of the formation are well represented
at Bagh,)—quartzite, hornstone, breccia, and cherty limestone,—and here
again interbedded trap occurs, though. not found in the Dhé&r forest area.
Clay-slate, too, is more prominent, and sometimes becomes cenglomeratic
through the presence of pebbles, which are more or less drawn out in the
direction of the cleavage planes. The town of Bagh stands near the
south boundary of the small triangular area of Bij4wars, where it is covered
by cretaceous rocks, the other two boundaries, with the gneissic rocks,
being faulted. The area only extends 7 miles to the north-north-west,
and 5 miles to the east of Bagh. The rocks are highly disturbed and
cleaved, but the metamorphism is local and moderate.

Jobat is about 15 miles west-north west of Bagh, and stands at the
southern point of another small patch of transition strata. The conditions
are peculiar and puzzling. The only recognisable Bijawar rock is a very
typical one, a locally brecciated, ferruginous jasper, with veins of horn-
stone. It lies almost horizontally, forming a low scarped plateau. Along
the north-east border, south of Anthi, black and grey schistose slates ap-
pear between the jasper and the metamorphics, the foliation and apparent
bedding in the schistose slates and gneiss being parallel, with a high dip
to the south-west, Both recks are highly charged with vein-quartz, sug-
gesting local crushing or faulting.

These disturbed rocks were originally regarded as Bijawars and classed
with the horizontal jasper beds. Mr. Bose has, however, described patches
of slates, quartzites, and limestones, which are said to pass insensibly into
the metamorphics, but are quite distinct from the horizontal jasper rock.?
In this case the latter alone can be regarded as Bijdwar, and this appears
to be the more probable interpretation.

Proceeding eastwards up the Narbad4 valley from Handi, no rocks are
exposed under the Vindhyan scarp, on the northern side for a distance of
120 miles to where the Bij4wars form low hills in the Narsinghpur district
The cherty limestone and breccia are the only beds seen heve but this
may be because the lower rocks are covered by alluvium
does not appear again on this side of the valley,

Along the south side of the river, on the edge of the Gondwéna forma-
tions of the Sdtpura hills, there are more frequent outcrops of the trans-
ition rocks. The most westerly are near the Moran river about 30 miles
east of Hard4, where some narrow ribs of the cherty ifniestowe rotrude
through the Deccan trap, which covers all the rocks :

to the west of this
point. On this south side of the valley, the cherty limestone, generally

1 Memoirs, VY, 303, (1869). 1 * Memoirs, XXI, 13 (1884).

The gneiss
Chap. III.) BIJAWAR SYSTEM IN THE SON VALLEY. 55
much contorted and brecciated, is also the rock most frequently seen, but
other beds do occur, as in the Badri hill 15 miles east of Sohdgpur, where
a considerable thickness of trappoid and earthy rocks is exposed, the latter
being so little altered as to be easily mistaken for the Talchir shales of the
contiguous Gondwana area. In many places on this south side of the
valley gneissic rocks of doubtful character occur close to the Bijdwars,
and the relation between the two series is certainly not simple superposi-
tion, both being found at the same level in closely adjoining positions.

At the head of the Narbada valley in the north of the Jabalpur district,
there is a continuous exposure of Bijdwar rocks between the Vindhyan
and Gondwana areas. All the leading characters of the system already
noticed are represented here, with a greater development of the argillace-
ous element. Fine earthy slates of reddish tints are the lowest strata seen.
Their upper beds are associated with the quartzite, which underlies the
limestone and is intercalated with it, and the limestone itself is not so
constantly cherty as has been described elsewhere. Ribboned jasper
beds, passing locally into bluish quartzite, among which rich haematite
beds are well developed above the limestone, and both jasper and quartz-
ite are frequently brecciated. Locally conglomeratic, earthy schists are
also freely associated with this band.

Above the iron- band there is again a considerable thickness of earthy

-schists. Bedded trap occurs throughout the series.

These rocks are not on the whole greatly disturbed. Low undulating
dips prevail, although locally there is much contortion. The highly inclined
planes, so general in the schists, are of cleavage, not stratification. The
thickness of the whole series exposed cannot be great, probably it is under
1,200 feet, and there is scarcely any presumption that the conformable
slates beneath the limestone attain any great thickness.

Notwithstanding these conditions the rocks are in an advanced state of
metamorphism. ‘The limestone is generally crystalline. The schists are
often highly micaceous, hornblendic, and garnetiferous, and the iron-ore is
mostly the micaceous form of hematite. The section in the Narbada. at the
well known marble rocks, 10 miles south-west of Jabalpur, exhibits the
high degree of alteration atid local disturbance to which the Bijawars have
been subjected in this region.

Immediately to the east of the flat watershed between the Son and the
Narbadé the band of transition rocks is entirely concealed by an extensive
spread of laterite and alluvium, and beyond this we get into the region
of the lower Vindhyans, which stretch to the south of the scarp of the
upper Vindhyan plateau until they nearly come into contact with the Gond-
wdna deposits. After crossing the Son, however, the band of transition
rocks again expands gradually to a width of 25 miles in the south of the
56 GEOLOGY OF INDIA—TRANSITION SYSTEMS. [Chap. III.

Mirzépar district, and it is here we encounter the question whether one or
two formations occur within this basin of transition rocks.

The northern half (about 10 miles wide) of the transition band, at a
little west of the Rer river, is formed of regular Bijdwar rocks, such as
we have hitherto seen them—quartzites, hornstones, banded jasper and
hzematite, limestones and slates or schists, with an abundance of intercalated
trap. The whole band strikes against, and under, the lower Vindhyan
strata, where the Son takes a southerly bend opposite Agori. The south-
ern half of the transition band, 15 miles wide, as exposed in the Rer,
is entirely composed of fine slates, with intrusive trap only, the dykes
being mostly transverse to the bedding. Both groups are so intensely
crushed together that no decisive section of the junction has been found
in the low jungle-covered hills. Mr. Mallet mentions an instance at Ubra,
at the north end of the section in the Rer, where a quartzite of the
northern set seems to cap a ridge of the slates, but the case is not clear
and the question of the relation is quite open, except that it certainly is
not one of horizontal transition, as the two contrasting deposits are in full
force and character in close juxtaposition to each other.

The western extension of the section into the Rewd country has
been but imperfectly examined. ‘The slates have already disappeared at
the Gopat, and the northern band of true Bij4wars is in contact with the
gneiss. In this region, where the Son takes a bend into the area of
the transition rocks, there is a good instance of local metamorphism,
the transition rocks along the lower Vindhyan boundary, distinctly re-
cognisable as Bij4wars throughout the whole length between the Gepat
and the Son at Murai, being in a gneissose condition and intrusive granitic
rocks occur in them, The character of the contact of these beds with the
main gneiss to the south is, however, of the kind described by Mr. Hacket
north of Jabalpur, abrupt rather than transitional, but it is certain that they
themselves are locally gneissic, and have been effected by granitic intrusions.

lf it were certain that this character of the contact of the Bijdwars
with the southern gneiss is constant and not due to faulting, and also
that the gneiss of the Rer and the Gopat are the same, we could at once
affirm the distinctness of two transition systems in this ground, for the junc-
tion of the slates of the southern band with the main gneiss is perfectly
transitional—a gradual alternating passage from the strong gneiss, through
gneissose and other crystalline schists, into the fine clay-slate, as is well
seen in the section in the Rer. But while doubts exist upon these two
conditions, it must remain possible that these slates of the Rer are only
a bottom member of the Bijdwar system,

East of the Rerand the Kanhar several large inliers of gneiss and of
granitoid rocks, of more or less intrusive character, occur within the slate
area, and gneiss is the only rock seen below the Vindhyans at the Koel,
Chap. III.] TRANSITION ROCKS OF BEHAR. 57

This encroachment of the crystallines upon the zone of the transition
rocks is extended in Behar, where, for some miles north of the Grand Trunk
Road west of Gay4, gneiss reaches quite across the strike of the slates.

Several hills isolated on the alluvial plains in this neighbourhood are of
thorough granite.

Immediately east of Gaya, transition rocks appear again on the pro-
longation of those in the Son valley, and having the same strike. They
form several groups of hills in east Behar, known as the Maher, Rdjdgriha
(Rajgir), Shaikhpira, Kharakpur, and Gidhaur hills, which stand clear
of the main gneissic area and more or less isolated in the alluvial plains,
and those of Mahdbar and Bhiaura on the northern margin of the gneissic
upland. The aspect of all these hills at once shows that they must be
formed of very different rocks from the Bijdwars of the west, and sug-
gests also that all these Behar rocks belong to one system. They
generally present scarped faces formed of massive quartzites on every
side, the associated schists or slates appearing obscurely in the valleys.
All the peculiar Bijdwar rocks are wanting. There is no limestone, horn-
stone, jaspideous ironstone, or bedded trap. The only similar rocks in the
west are the slates of the Rer section, and there the quartzites, which
form such a prominent part of the transition series of Behar, are absent.

We have a somewhat detailed description of the Mahdbar and Bhiaura
hills by Mr, Mallet,! and the relation of the gneiss and transition rock-series
is shown to be very peculiar. The transition series here consists of three
divisions, an upper, composed exclusively of strong quartzites as seen in
Mahdbar hill, a thick middle band, in which fine mica-schists largely pre-
dominate and a basal member, in which quartzites again occur, sometimes
in great force, as in the Bhiaura ridge, though they may be altogether
wanting at no great distance, The frequent presence of these quartzites
here is of great service, by removing the doubts that so often arise, as to
whether planes of lamination in schistose rocks, of uniform composition,
are due to bedding or to cleavage.

It wouid be difficult to draw a more irregularly intricate line than tke
transition and gneiss boundary on Mr. Mallet’s map. Near the Bhiaura
and Mahabar ridges there is some approach to an average east and
west strike of the boundary, but, as the plane of junction between the two
series rises to the south, its line of outcrop meanders about in the most
devious manner. This is not due, as might easily occur, to the irregular
denudation of two deposits in flat parallel superposition. Here, asa rule,
the lower (older) rock forms the prominences, between which the schists
are deeply buried, yet the bedding in both rocks is found to follow

1 An abstract of it is published in Records, VII, 36, (1874).
58 GEC1.OGY OF INDIA—TRANSITION SYSTEMS. — (Chap. III.

the intricate twistings thus produced, the actual junction being generally
inclined at a high angle.

The boundary can always be fixed with precision, even in the absence
of the bottom or Bhiaura quartzites, on account of the strong contrast be-
tween the fine mica-schists and the coarse gneiss, yet the transition rock
seems to have fully partaken in the metamorphic action, for it is a tho-
roughly crystalline, garnetiferous mica-schist up to the base of the Mahdbar
quartzites. Variations are found in the gneissic series at the contact.
On the north side of the Bhiaura ridge the bottom quartzites lie steeply
against the dome-gneiss, elsewhere schistose gneiss occurs at the boundary.
The dykes and massive outbursts of pegmatitic granite of this region are
principally exhibited in the transition series.

A very close connection is thus established in this position, by conform-
ity of stratification and by a common metamorphism, between the transition
rocks of Behar and the gneiss in contact with them, and it is probable that
a large part of the gneiss of Bengal is of the same age as that at the
boundary of the transition series, There is, for instance, a very distinct
outlier of the Mahdbar schists and Bhiaura quartzites on the plateau just
north of the Grand Trunk Road at Barhf, 30 miles to the south of the
boundary in Behar.

There can scarcely be a doubt that the rocks of the RAjdgriha and other
detached hills of Behar are of the same formation as those of Mahabar,
and so the contrast of their mineral condition is interesting. The latter
have undergone general crystalline metamorphism, the former have only
very locally suffered this change, being for the most part still in slaty con-
dition, Yet it would seem that they too are closely surrounded by crys-
talline rocks, for whenever rock is exposed, through the alluvium near
these hills, it proves to be granite. At one spot near Ghansura, on the
north side of the Rajagriha range, there is a contact showing distinct in-
trusion of granite into the soft earthy schists. It is an ordinary ternary
granite, not like the pegmatitic granite of the Mahdbar region. In the
immediate neighbourhood of Gayd many forms of special metamorphism
and of contact-action are well exhibited.

The amount of disturbance is rather greater in the detached hills,
where the rocks are less metamorphic than in the Mahdbar region, and
the very peculiar confused form of contortion, noticed in the surface
of junction where the transition rocks rise against the gneiss to the south,
is well exhibited throughout the formation, but in larger proportions
in the top beds of theseries. The Mahdbar ridge itself is a typical instance
of this structure. It is a long, narrow, synclinal ellipse, the quartzites dipping
at a high angle all round, towards the centre, and curving continuously at
each end ofthe axis. The Rdjdgriha range contains a pair of such ellipses
compressed together, the quartzites being for the most part quite vertical
Chap. IIT.) TRANSITION ROCKS OF BEHAR. 59

along the sides. The Kharakpur hills, which form the largest of these
groups, are a congeries of these discontinuous flexures, little or no regu-
jarity being observed in the direction of the axes of contortion.

We have still to notice the rock underlying the quartzite in the smal]
ridge of Shaikhpdra, and in the little hills a few miles to the east at
Luckeeserai, the junction station for the chord and loop lines of the
East Indian Railway. There can be little doubt that the quartzite of these
localities is the bottom rock of the Behar transition series, the Bhiaura
quartzite. In the Shaikhpira ridge it rests steeply against a rock having
the texture of a thoroughly crystallised coarse granite, but completely
decomposed. The relative position of the two rocks is precisely that of
the Bhiaura quartzite and the dome-gneiss. Along a steady outcrop of
some two miles long no feature of special intrusion was observed, and there
is no extra metamorphism at this junction. The only contact-action that
occurs is of secondary origin, in the formation of layers and vein-like
strings of a sharply cellular quartz-rock much used for making hand-mills.

This section is noticed in connection with a more decided one at Luckee-
serai, only a few miles to the east on the same strike, where the quartzite
again rests against an amorphous mass of pseudo-crystalline granitoid
rock, of much less sharply defined texture than at Shaikhpira, in which
strings of pebbles can be detected. This is underlaid by strong beds
of coarse conglomerate, having the same dip as the overlying quartzite.
The pebbles and boulders in this conglomerate are mostly sub-angular,
and are exclusively of varieties of quartzite like those of the over-
lying formation, none being of crystalline rocks. They often appear elon-
gated in the direction of the foliation, and adhere firmly to the matrix,
which is a quartzose, sub-gneissose schist. Just east of Dhardrah station
some masses of this rock protrude through the alluvium close to the base of
the Kharakpur hills. Another outcrop of conglomeratic schist was observed
under the east end of the Gidhaur range and dipping towards it. :

These Lukeeserai beds remind one forcibly of the pseudo-gneiss ob-
served conformably underlying the Bijdwars in the section of the Ken river
in Bundelkhand, and the suggestion revives, however slightly, the question
of the possible correspondence of the transition groups in the two areas.

There is another rock frequently found with the undulating gneissic
rocks of Behar, and elsewhere in this zone or protruding from the alluvium
near the hills, that suggests the same connection, It is a jaspideous quartz-
ite, often brecciated, and not unlike the bottom Bijdwar rock of Bunde!-
khand and the Dhar forest. It commonly has the same moderate dip as
the rocks with which it occurs, but, when vertical or crushed, it is readily
mistaken for fault-rock or vein-stone.

Suggestions of an opposite tendency can, however, be pointed out from
observaticns recorded in preceding paragrapls. It was stated that the
60 GEOLOGY OF INDIA—TRANSITION SYSTEMS. [Chap. IIL.

contrasting groups of transition rocks in the northern and southern portions
of the section in south Mirzdpur cannot be, in any degree, representative
of each other by. horizontal transition, and the presumption would be
strongly in favour of the southern beds—the slates of the Rer—being the
older of the two. If the Behar rocks had to be affiliated to either of these
exclusively, it would certainly be to the latter group.

It has been already explained that the gneissic formations in lower
Assam and the hills to the south are more closely allied to those of the
peninsular region of India than to the metamorphic formations of the
Himdlaya. This relation holds also for the transition rocks, which are
largely developed on the south side of the hills, where the sub-metamorphic
beds are for the most part covered by the horizontal cretaceous rocks
of the plateau, but are exposed in the deep ravines that penetrate to the
_ very axis of the range. The lateral extension of these transition rocks

has not been ascertained. Onthe central cross-section in the Khasi country
they stretch for 30 miles from near the south margin of the plateau to
beyond the watershed north of Shillong, the culminating ridge with its sum-
mit 6,450 feet high being formed of the quartzites of the transition series,
which have hence been described as the Shillong series,?

These Shillong beds have a general resemblance to those in Behar. They
consist of a strong band of quartzites overlying a mass of earthy schists.
Great masses of granite and of basic trap rock also occur intrusively. The
former may well represent the similar rock seen to be intrusive into the
slaty schists of Rdjagriha, and the latter resembles certain trappoid rocks
in Behar. Thus altogether the affinity is sufficiently marked to introduce
the notice of the Shillong area in sequence with that of Behar. In the
lofty and deeply eroded ground of the Assam hills the sections are much
more favourable for study than on the alluvium-smothered plains, and some
very puzzling observations have been recorded regarding the relations
of the hypogene rocks to the Shillong series.

The lithology of these Shillong rocks varies much, according to local
conditions of metamorphism. In places the quartzites, generally very firm
and more or less schistose, are quite friable and might be called sand-
stones, but this state is probably due to decomposition, for the texture
always reveals the effects of chemical change. It is coarser grained
than is common in the Behar quartzites, and at the base, immediately over
the slates or schists, there usually occurs a conglomerate, often of
considerable thickness, made up chiefly of quartz pebbles, but with some
rounded fragments of coloured quartzites. Still, so far as has been made
out, the quartzite is conformable to the schists, but in troubled ground
it is difficult to make sure of such a point, The schistose beds also exhibit

1 H.B. Mcdlicott, Memoizs, VII, 197, (1860).
Chap, IT1.] SHILLONG SERIES. 61

much variety of texture, from ordinary clay slate to well foliated schists
and gneiss. These changes are simultaneous in both quartzites and
schists, and it is noteworthy that the increase of metamorphism is towards
the south, away from the area of the old gneiss,

The relation of the transition rocks to this gneiss has not been made
out. On the only section of which we have critical observations, nearly
due north and south through Shillong, the boundary with the gneiss occurs
in the low jungle-covered hills, where observation is almost impossible.
The dividing line between the two series crosses the high range to the
west of our section, and it is there that the junction should be examined,
The observation already noted, that the metamorphism increases to the
south, would suggest that the junction of the schists with the main gneiss
to the north may be lithologically abrupt. At the southern boundary there
is a steep plane of contact between the highly altered transition rocks and
the great accumulation of bedded eruptive rock, known as the Sylhet trap,
supposed to correspond with that of the R4jmahdl hills and, therefore, to be
of jurassic age. The cretaceous sandstones lie evenly and unconformably
on both formations.

In the midst of the transition area there is an extensive exhibition of
eruptive rock, of very different character from the Sylhet trap. It is a
dense, massive, basic diorite or greenstone. ‘The high road between Sura-
rim and Mauphlong crosses this rock continuously for five miles in the
gorges of the Kdlapdni and Bogap4ni rivers. The direction of the road is
oblique to the strike of the rocks, but at right angles to its outcrop the
greenstone is fully a mile wide. It nowhere betrays any bedded structure,
and its intrusive character is not so marked as might be expected with so
extensive adisplay of igneous rock. ‘There is, however, sufficient evidence
of intrusion for this greenstone, as a well defined dyke passes from the
main mass into the quartzite of the ridge, about half a mile south of
Mauphlong. Elsewhere one may walk for miles along the junction of the
two rocks without finding any signs of penetration of one by the other.

The relation of the granite, or at least of the larger masses of the crys-
talline rock, to the transition rocks, is also very puzzling. Two such
masses adjoin the high road across the Khési hills. One is the Myllim (Molim)
area just! south of the Shillong ridge, and close to the road between Mauph-
long and,Shillong. The other area is much more difficult of approach, the
granite being only exposed in the deep gorges under the sandstones of the
plateau, as on both sides of Surarim. ;

The rock isa thorough granite, It commonly affects a spheroidal struc-
ture, and it is composed of pale pink orthoclase, often in large crystals, a
small proportion of very pale greenish oligoclase, a little dark-green or
brown mica, and an abundance of disseminated hyaline quartz. Thee can
be no question that these great granitic masses are of later origin than the
62 GEOLOGY OF INDIA—TRANSITION SYSTEMS. [(Chap, IIT.

transition series, for the total want of symmetry in the arrangement of the
surrounding sedimentary rocks forbids the supposition that they could have
been deposited round the granite, yet the absence of any apparent con-
nection between the form of the intrusive masses and the disturbance
of the transition rocks is very difficult to understand. The quartzites (the
upper member of the transition series) are generally found at the boundary,
but their dip and strike are quite independent of the granite, as if their
contortions had been fully established before the granite was introduced,
and remained quite unaffected by it. The facts seem totally to preclude
the notions of fracture and compression commonly associated with the
word intrusion. The supposition of the mass being faulted into position
also lacks any corroborative evidence, the boundary lines are all rounded
and show no symptoms of fissuring, It is as if a great hole had been
burned out of the old stratified rocks and the crystalline mass let in, or as
if the transition rocks had been converted into granite up toa certam
boundary, without affecting the area beyond that line. Yet the junction
is quite sharp, the quartzites not being more altered at the very contact
with the granite than away from it. In keeping with all these negative
characters is the fact that no dykes or veins of granite have been observed
issuing from the great mass of Myllim, nor even in its neighbourhoed,. This
is the more remarkable, because dykes and veins of similar granite are not
uncommon in the southern part of the area, where the general metamor-
phism of the transition series is so much greater as to suggest that the
focus of hypogene activity lay in that direction, beyond the present southern
limit of these formations, It is also in agreement with the facts and
suggestions recorded to note that the granite is younger than the old
dioritic Khdsi trap; several small dykes of granite are seen ramifyirg
through the diorite in the bed of the torrent east of Surarim.

The gneissic uplands of Hazéribdgh and Chutié N4gpur, about 120 miles
wide, separate the transition rocks of Behar from those which occupy
parts of M4nbhim and Singbhim in south-west Bengal and stretch far to
the west, the whole transition area being about 150 miles long from east to
west, and 80 miles wide!.

Although the total thickness of this series must be great, no distinct-
ive zones are marked in it. From top to base it seems to be an indiscri-
minate alternation of quartzite, quartzite sandstone, slate and shales, horn-
blendic, micaceous, talcose, and chloritic schists, the latter passing ‘into
potstone, and some bedded trap, Well-preserved ripple marks are found in
the slates and shales, and some of the latter are so little metamorphosed
that they might be mistaken for Talchirs, but for the quartz veins that
penetrate them.

1 Memoirs, XVAI, 73, 124, (1881),
Chap, IID] TRANSITION ROCKS OF BENGAL. 643

Some large inliers of gneiss occur within this basin of transition rocks.
Around some of these inliers the boundary is in its original condition,
as at Chdibds4, where shales and sandstones rest flatly and quite un-
changed upon the coarse gneiss of the principal inlier, and the uncon-
formity of the two series is further proved in this case by the fact that
the underlying gneiss is profusely traversed by trap dykes, which do not
penetrate the overlying deposits. The boundary between the transition
rocks and the main gneiss of Bengal on the north is said to be a fault, on
account of the more or less continuous presence along it of a rib of vein-
stone. This boundary occurs, however, at the base of a long descending
section of the transition rocks and the beds along the line of junction
are such as elsewhere appear as bottom-beds of the transition series,
There are besides outliers of the slate series beyond the supposed faulted
boundary to the east, about Sipur, and an inlier of gneiss a short distance
inside it at Borobhum. We can at least conclude that the junction here, whe-
ther faulted or not, is abrupt, that is to say, without any gradation of strati-
graphical or mineralogical characters. In this part of the basin the maxi-
mum of disturbance and of metamorphism seems to occur away from the
boundaries. Further to the west, however, the junction of the slate and
gneissic series is described as transitional, and granite veins penetrate the
slates without much affecting them.

The most striking feature of this area is a mass of dioritic trap running
continuously, but with varying width, nearly east and west, through the
‘centre of the transition basin. Dalmé hill, 3,050 feet high, is formed of
this rock, and here the outcrop is nearly 3 miles wide. ‘The trap is de-
scribed as a great dyke, but its composition is described as complex and
obscurely bedded. A section north of R4mgatrh is given as—?

1. Indurated chloritic schist. 5. Indurated chloritic schist.
2. Porphyritic trap, 6. Brecciated trap-
3. Indurated chloritic schist. 7. Indurated chloritic schist,
4. Compact and amygdaloidal trap. 8. Brecciated trap.

Several other cases-of similar variation were observed. The supposed
dyke is found along the axis of a greatly compressed synclinal fold, and
has evidently been subject to much crushing, the description is not
such as one would expect in the case of a truly intrusive rock, and it is at
least possible that it is composed of contemporaneous volcanic rocks
whose structure has been obscured by disturbance.

As in Southern India, these transition rocks carry metalliferous lodes
of gold, silver, copper, lead, etc., but so far none of these have proved
remunerative—except to promoters of joint stock companies and a limited
number of speculators in mining scrip.

In Chutid Ndgpur a few exposures of quartzites and schist have been

1 Memoirs, XVII, 80, (1881).
64 GEOLOGY OF INDIA—TRANSITION SYSTEMS, (Chap. Ill,

separated from the gneiss, and it is probable that more will be found
when the gneissic area, intervening between the patches of Gondwdna
rocks, is more closely examined. Some of these patches are indicated on
the map, but they do not exhibit sufficient peculiarity to require detailed
notice,

West of Rdipur, in the Central Provinces, is another stretch of trans-
ition rocks, known as the Chilpi ghdt or Saletekri beds,’ which have been
but little examined. They consist of quartzites, dark green and buff slates,
and shaly beds, coarse conglomerates and numerous beds of trap. They are
for the most part much disturbed, but sometimes lie at easy dips. The
succession of the beds has not been worked out, and little more is known
of them than their approximate boundary.

To the south-east Dr. Ball found, on the eastern side of the plateau
south of Tarnot,? some much disturbed shales, with subsidiary quartzites
underlying the horizontal quartzites of the plateau. These very probably
belong to the same series as the Chilpi ghat beds, and on the accompanying
map have been coloured as such,

Turning to the north-west corner of the Peninsula the beds of the G walior
system’ are found, about 120 miles from the Bundelkhand outcrop of the
Bij4war system, resting upon the gneiss in precisely the same mechanical
relation, horizontally or with a gentle slope. The denuded outcrops of the
quartz reefs traversing the gneiss are in both cases covered by the bottom
deposits of the overlying transition groups, but a slight difference is
noticeable at the actual contact. The bottom layer of the Bijdwars is
commonly more or less adherent to the gneiss, the result of the partial
metamorphism that the Bijawars, even in Bundelkhand, have undergone,
whereas in the Gwalior rocks the bottom contact-layer is still unaffected
in contact with the gneiss.

A general list of the rocks of the Gwalior formation would not suggest
any separation from the Bijdwars. Each contains sandstone or quartzite,
limestone, jasper, iron bands and bedded traps. The arrangement of
these different strata is, however, markedly different in the two cases,
and the general facies of these two series does not suggest to the observer
that they are representative, Still, the Gwaliors are more nearly allied by
their mineral characters to the Bijdwars than to the lower Vindhyans, and
on this account, on account of their relations to the slaty series of tke
Ardvallis, and of the great unconformity which subsists between them and
the upper Vindhyans, they are here classed with the transition rather
than the older palzozoic systems.

1 The only published description will be | seriesby Mr. C. A. Hacket in Records, Ill, 33,

found in Records, XVII, 187, (1885). (1870) ; it is also mentioned in Mr. Hacket’s
2 Records, X, 175, (1877). paper on the north-east Ardvalli region,
3 There is a short notice of the Gwalior Records, X, 84, (1877). .
Chap. III. ] THE GWALIOR SYSTEM. 65

The area occupied by the Gwalior system is only 50 miles long, from
east to west, and about 15 miles wide. It takesits name from the city
of Gwalior, which stands upon it, surrounding the famous fort built upon
a scarped outlier of Vindhyan sandstone, The composition of the Gwalior
formation is very mixed, and admits of only a twofold and very unequal
subdivision, There is constantly at the base a sandstone, or cemi-
quartzite, a fine-grained stone, pale grey in colour, regularly and_ thinly
bedded, often conglomeratic at the base, called the Pdr sandstone, from
atown 12 miles south-west of Gwalior. It varies from 20 to 200 feet in
thickness, and is overlaid by about 2,000 feet of strata, consisting mainly ot
thin, flaggy, silicious, ferruginous shales, copiously interbanded with
hornstone and jasper, frequently of a brilliant red colour. Limestone, more
or less cherty, occurs on two principal horizons in these shales, but not
continuously, and there are two principal zones of a dense basic dioritic
trap. Alithese upper beds amounting to about 2,000 feet in thickness, and
constituting the bulk and the characteristic portion of the Gwalior forma-
tion, have been distinguished from the Pdr sandstone as the Mordr group,
the name being taken from the military station close to Gwalior.

With the exception of some very local slips and crushing, the
Gwalior rocks are undisturbed, having a steady, low, northerly inclination
of only three to five degrees, The features of the area correspond
with this arrangement of. the rocks. There is a continuous broad
plateau-range on the south, from 300 to 500 feet high, formed largely
of the P4r sandstone. On the west it is connected at right angles with
the Vindhyan scarp, which lies at a slightly lower level, and it stretches
thence eastwards to the Sind river, forming a steep scarp overlooking
the gneissic area of lower Bundelkhand to the south. ‘There are two
other ranges on the north, parallel to the Par scarp, but they are much
broken by cross-drainage, the two longitudinal valleys between the three
ridges being due to greater decomposition and erosion alorg the two
outcrops of bedded trap, It is only at the west end, near the Vindhyan
plateau, that these trappean bands are well exposed.

The general. easterly direction of the Pd4r scarp is very steady up to
the Sind river, but the line is much serrated by bays and headlands, in which
the nature of the junction with the gneiss is well exhibited. At the most
southerly points of the range the gneiss reaches to within a few feet of the
summit, and is capped by only a few feet of sandstone, but the surface of
the gneiss gradually slopes downwards in the valleys cut back to the north,
and the thickness of the overlying Gwalior beds increases. This slope of
the junction is largely due to the original form of the basin in which the
beds of the Gwalior series were deposited, for close to the edge of the
scarp the thickness of the Pdr sandstone is small, near Par only about 20
fect, while on the north side of the range, wherever sections are exposed,
as at Badhano, ten miles south-east of Mordr, the thickness is not less

F
66 GEOLOGY OF INDIA—TRANSITION SYSTEMS. (Chap. Ii,

than 200 feet. No trace of an unconformity between the Par and Morar
groups having been detected, this thickening must be due to the form
of the basin of deposition, and would seem to show that the present south-
ern limit of the Gwalior series represents very closely the original limit of
deposition of the Pdr sandstone.

On the top of the Pdr sandstone there occurs locally a compact
calcareo-silicious bed that is worth noticing, because the peculiar coralloid
forms it exposes by weathering were thought by Dr. Stoliczka to be of
organic origin. This rock is best seen just south of Bara, 25 miles east
by south of Morar,

The lower zone of bedded trap is about 400 feet from the base of the
Morér group. There are two or more flows with intervening shales
well seen on both sides of the Indore road, at from 6 to to miles south-
west of Gwalior. The thickness of these flows is very various, From
7o feet they thin out to nothing, but they are probably nowhere absent
on this horizon, obscure outcrops of them having been observed at several
places in the valley formed along their strike to the east. At some spots
there is an appearance of the trap having burst up through the under-
lying shales, Thus, in the stream near the Trunk Road north-west of
Bela, there is a low section showing the shales and trap in vertical con-
tact, but otherwise the interstratification is unbroken.

In connection with this lower zone of trap there occurs:a rock that will
again come under notice in the Cuddapah system, and also in the lower
Vindhyans. It is a compact porcellanic rock, as sharply and regularly
bedded as the associated jaspideous shales. Occasionally it is obscurely
porphyritic, having small indeterminate crystals scattered through it. An
analysis of a specimen from the Gwalior beds gave a composition ap-
proaching to that of orthoclase felspar.? But there is no presumption
that this porcellanic rock, or hornstone, which has more than once been
described as volcanic, has any connection with volcanic activity, and its
association here with trappean beds of highly basic composition is probably
quite fortuitous.

The upper zone of trap is on a much larger scale. The whole plain
of Mordr is underlaid by it, at least on the north side, and, if allowance
js made for the small dip, the flow can hardly be less than 500 feet thick.
It is admirably exposed in the undercliff of the Vindhyan scarp in the fort
hill and the promontories to the westward. Inasmall plateau about three
miles to north-north-east of the fort it is overlaid horizontally by typical,
rusty, jaspideous shales of the Gwalior series. Several detached hills
in the plain lying east by north of Morar are formed entirely of this
massive trap.

Limestone occurs principally on two horizons, in and above the lower
{rappean zone and, in the northern hills, above the great trap flow. In

! Records, 11, 35, (1870). | ? Records, III, 37, (1870).
Chap. IT. J THE GWALIOR SYSTEM. 67

both positions it is very uncertain and discontinuous. Within a space of
100 yards a mass of limestone, more than 50 feet thick, is found to be
totally replaced by ochreous shales.

The iron ore, which is largely mined in the Gwalior system, is quite
different from that found in the Bij4wars. The latter is a massive
concretionary hematite irregularly concentrated in ferruginous earthy
sandstones. The Gwalior ore is a fine wafer-like shale, composed of thin
flakes of haematite with still thinner films of clay. It is a decomposed con-
dition of the jaspideous shales, from which the amorphous silica has been
dissolved out, leaving the iron ingredient in a very favourable state for
mining and smelting. The conditions for this change seem only to have
obtained near the base of the group, as the mines are in the shales a
few feet over the Pdr sandstone.

To the east and north the Gwalior system is covered by the great
alluvial plains. On the west it passes under the upper Vindhyans, and two
inliers, exposed by the removal of these covering rocks, are crossed by
the Trunk Road. The only specific identification of the Gwalior beds
beyond this area is at the nearest point on the opposite side of this
northern extension of the Vindhyan basin, 7o miles to the north-west of
Gwalior. At Hindaun there is a narrow ridge of banded jasper and fer-
ruginous shales, which Mr. Hacket considers to be indubitable Gwaliors!.
The Gwaliors at Hindaun are more or less vertical, and in contact with
them, but not conformably, are some quartzite sandstone and red and
black slaty shales, with irregular bands of limestone, which will be again
referred to when dealing with the Delhi system.

‘The antiquity of these rocks is shown not only by this section at
Hindaun, but by the very extensive denudation they bad undergone pre-

 

 

Fig. 4. Section showing the relation between the Gwalior and Vindhyan systems at
the junction of the Pdr and Vindhyan scarps. Ks, Kdimur sandstone. Ke.
K4imur conglomerate. G. Gwalior beds. C. Bundelkhand gneiss.
vious to the deposition of the upper Vindhyan sandstones, At the western
end of the Pdr scarp, the Kdimur sandstones and conglomerate are de-
posited against a scarped face of Pdr sandstone and rest on the gneiss at
a lower level than the base of the Gwaliors close by *.

The Ardvalli system was formerly taken to comprise all the transition

1 Records, Ill, 40, (1870); X, 90, (1877). | ? Memoirs, VIL, 58, (1869).
F2
68 GEOLOGY OF INDIA—TRANSITION SYSTEMS. — [ Chap, It.

beds of the Ar4valli range, including those belonging to what we will now
distinguish as the Delhi system’. Little can be said of the petrography or
of the relations of the Ardvalli system, as limited by the exclusion of these
newer beds, and even the validity of the separation is in doubt. Mr.
Hacket was inclined to believe® that the schists of the Ardvallis were
merely the metamorphosed equivalents of the lower portion of his Delhi
system, but the disproportion, in thickness as well as in metamorphism,
between the beds of the great schist area of the Ardvallis and the much
smaller thickness of slates and limestone, which alone can be included
with certainty inthe Delhisystem, renders this doubtful. Moreover he has
recorded some observations, noticed below, which point to an unconform-
able break between the two systems.

Accepting the validity of a distinct Ardvalli system of transition age, it.
may be described as consisting of quartzites and limestones, often con-
taining coccolite, hornblendic and mica schists, abounding in crystals of
andalusite, staurotide and garnet, and felspathic schists and gneisses.

The contact of the schists and gneiss shows a gradual transition’, bothin
the centre of the range, and where they are in contact with the granitoid
gneiss of Dhariawad and near Chitor, which has been regarded as belonging
to the Bundelkhand or older gneiss series. This transition is in part appa-
rent, and due to the true gneiss not having everywhere been distinguished
from gneissose forms of intrusive granite, but is mot improbably to some
extent real.

The relation of the Ardvalli schists to the Delhi system is somewhat
doubtful; where the lower beds of the latter have undergone metamor-
phism they are difficult to distinguish, and it is probable that in those sec-
tions which appear to show a passage between the two, the break occurs
between beds which it is difficult to distinguish from each other. This is
rendered probable by the unmistakable unconformity which is shown by
some sections, such as the one near Nithahar, where the Alwar quartzites
rest upon the edges of nearly vertical argillaceous and quartz schists, and
in the hills south of the Basi railway station, where a thick band of coarse
conglomerate occurs at the base of the Alwar quartzites immediately above
what were regarded as beds belonging to the Raialo group.

East and south-east of Udaipur, in the heart of the range, conglomer-
ates, containing numerous boulders and pebbles of quartzite in a schistose
quartzite matrix, occur close to ridges of quartzite, which were regarded as
of Alwar age’. The position of these conglomerates is not very well
established. The nature of the boulders they contain would lead one to
suppose that they were of later date than the guartzites of the ridge close
by, but their position would indicate that they came between the quartzites
and the adjoining older schists.

1 Manual, ist ed., p. 49. (1881). MSS. Reports, passim,
7C. A. Hacket, MS. Report, 1886. * Records X, 86, (1877); XII, 4, (1879).
C. A. Hacket, Records, XIV, 282, °C. A. Hacket, MS. Report, 1883.
Chap. Til. ] THE ARAVALLI AND DELHI SYSTEMS. 69

Thesesections appear to leave but little room for doubting that an
unconformable break exists, and for this reason it appears advisable to
recognise the existence of a distinct system of schists and slaty beds,
underlying and older than the Delhi system, though it is impossible to
attempt any subdivision or detailed description of them.

Beds of compact silicious rock and jasper, slightly resembling those of
the Gwalior system, are recorded as occurring in the Raialo group in
Shaikhd4watf, near Chenpura, north-east of the Basi railway station, and
near Muhammadpur, south of Kherly railway station, At the time the
observations were made the term Raialo covered all the beds below the
Alwar quattzite, except the gneiss, and it is consequently uncertain
whether these beds belong to the Delhi system or not, but the occurrence
of jasper pebbles in the conglomerates of the Alwar group makes it prob-
able that they should be referred to the Ardvalli system, as here restricted.

In the central part of the Ardvalli range the Ardvalli schists are profusely
penetrated by granite veins, and have in consequence undergone great
metamorphism, But west of Udaipur there is an area where granite
is wanting, and the beds are almost as unaltered as the slates and lime-
stones below the Alwar quartzite, west and south of Nimach. This was
considered to be an indication that the last-mentioned beds were repre-
sented in a more metamorphosed condition by the Ardvalli schists, but it is
equally possible, and on the whole more probable, that they are an outlier
of the newer beds or, belonging to the older system, have locally escaped
metamorphism.

The rocks of the Delhi system extend, in a number of isolated out-
crops of varying size, from Delhi to beyond Nimach, a distance of about
340 miles from north-north-east to south-south-west, and for a width
of about 150 miles in a direction transverse to this. The name applied to
the system by Mr. Hacket in 18811 has proved an unfortunate one, as it
is but illexposed near Delhi, and we must look to the hills near Bidna, and
those of Mandsaur and the neighbourhood of Nimach, for the typical sec-
tions. It is, however, the name which has been used on the mapsand in the
publications of the Survey, and a change would lead to greater incon-
venience than its retention.

The beds comprised in the system consist of a lower group of slates and
limestones, and an upper, very much thicker, group of quartzites, knowa
as the Alwar quartzites. The lower group was, in the first instance,
named the Raialo, but as this name has subsequently been applied to all
the beds below the Alwar quartzites, including those which there seems
good reason for separating as an independent system, its use will be

- abandoned here.

4 Records, XIV, 281, (1881).
yk) GEOLOGY OF INDIA—TRANSITION SYSTEMS, [Chap. Til.

The Alwar quartzites may be described generally as well-bedded
quart :ites, of light grey colour and fine grain, in which ripple markings and
sun-cracks on the surface of the beds are common. Coarse grained beds
are of frequent occurrence, and slaty bands are met with, arkose is often
found near the base of the group where it rests on gneiss or granite, and
the earlier part of the period during which it was formed appears to have
witnessed an outburst of volcanic activity.

In the Bidna hills the Alwar quartzites have been divided into five
groups as follows :—!

. Wer quartzites and conglomerates.
Damdama quartzites and conglomerates.

. B:4na white quartzite and conglomerate.

. Badalgarh quartzite and shale.

Nithahar quartzites and bedded traps,

These groups are all separated by slight unconformities of denudation
and overlap, but the distinctions appear to be quite local, for, even in the
Bidna hills, they are distinguishable on some sections, while in other
outcrops it has been impossible to recognise them. All the groups vary
very much in thickness, and are completely overlapped near Nithahar by
the Wer quartzites, which rest directly on the schists,

West and south of Nimach? a very similar succession of beds is seen,
consisting at the base of a conglomeratic sandstone, overlaid by about 200
feet of shales and limestone, and capped by a varying thickness of quartz-
ite. The beds are very little disturbed and the exposures are completely
isolated, but, northwards of Sadri, there are exposures of highly disturbed
quartzite which, forming a series of ridges running northwards, can be cor-
related with the beds of Mandsaur and Sadri on the south, and those of the
Bidna and Alwar hills and Delhi on the north. The identification depends
partly on the observed relation of the quartzites to the older rocks, partly
on the similarity of lithological character, a perfectly justifiable method over
such short intervals as we are dealing with, and is helped out by the frequent
occurrence of beds of contemporaneous trap in the lower part of the series,
though these have not as yet been subjected to a critical examination.

The relation of this system to the gneiss west of Nimach, and to the
schists of the Ardvalli system, as now restricted, is one of complete un-
conformity, there being usually a conglomerate at or near the base of the
section, in which pebbles of the underlying gneiss are stated to occur near
Daulapdni. In the ridges north of Sadri the same unconformity has been
observed, and a similar unconformable contact, accompanied by a basal
conglomerate, has been observed near Nithahar where the quartzites rest
on vertical schists, near ‘Talra south of Alwar where they rest on granitoid
gneiss and contain gneiss pebbles, and at Marot, north of the S&mbhar
salt lake, where the bottom beds of the quartzite, conglomeratic with rolled -

Po

myo

Records, X, 86, (1877). 1 2 Records, XIV, 293, (1881).
Chap. III.) THE DELHI SYSTEM. 7

and subangular fragments of quartz ard felspar, rest on micaschists pene-
trated with granite intrusions!,

In the Nimach area the lower slates and limestones of this system rest
unconformably on the gneiss, but no case of unconformity with the, Ardvalli
beds appears to have been recorded. The complete overlap of the slates
by the quartzites which are conformable to them is, however, in itself an
ample proof of the unconformity of the Delhi system to the underlying
rocks.

The central sections of the Ardvalli exhibit an apparent passage of the
quartzites into the gneiss, one section in particular being mentioned, where
alternating mica sch'sts and gneissose beds are overlaid by schists and
quartzites, then a two-foot band of gneiss, capped by the Alwar quartz-
ites*. In these cases it is probable that a gneissose granite, intrusive
alongthe bedding planes, has been described as a gneiss, or the so-called
gneiss may be merely a metamorphosed arkose, in neither of which cases
would there be any proof of a transition between the true gneiss and the
Alwar quartzites.

In the hills near Hindaun quartzite sandstone, associated with red and
black slaty shales and irregular bands of limestone, occurs in close proximity
to the jasper beds which are considered to belong to the Gwalior system.’
No actual contact is seen, but the distribution of the two types of rock
leaves no room for doubting that the quartzites and slaty shales are the
newer of thetwo. In spite of their likeness to the beds of the Delhi system,
and the absence of anything at all resembling them among the Vindhyans,
they were referred to the latter system*. This reference appears to have
been due to the supposed difficulty of finding time for the deposition of all
the transition beds of the Ardvalli range, and their subsequent disturbance,
between the close of the Gwalior period and the commencement of the
Vindhyaa, the beds of both these systems being almost undisturbed, The
difficulty regarding the time required for the deposition of the transition
beds vanishes if we recognise two distinct systems in the Ardvallis, the
older of which may be contemporaneous with, or older than, the Gwaliors,
while the newer is younger; and as regards the disturbance, we will find
when we come to deal with the extra-peninsular mountain ranges, that
an intense and extensive disturbance of the strata has taken place during
the tertiary period, and even within the latter half of it. The time required
for the disturbance of the Ardvallis may consequently be reduced to a
very short period, geologically speaking, and if the suggestion,® that the
Vindhyans bear much the same relation to the Ardvalli range that the
deposits of the Indo-Gangetic plain do to the Himalayas, is correct, the

1 Records, XIV, 294, 206, 298, (1881). : Manual, 1st edition, p. 52,
2 Records, XIV, 207, (1881). Infra, p. 104.
3 Records, X, 90, (1877) ; Supra, p. 67.
52 GEOLOGY OF INDIA -TRANSITION SYSTEMS.  [ Chap. III.

disturbance would not be anterior to, but contemporaneous with, the
deposition of the Vindhyan beds.

Under these circumstances the correlation subsequently adopted,} which
classed the quartzites of Hindaun with those of the Alwar group, seems the
most probable one, and we have the Delhi system established as newer than
the Gwaliors. This conclusion is supported by the occurrence of pebbles
of jasper, closely resembling that of the Gwalior system, in the lower
beds of the Alwar quartzites, north of Dhaulap4ni and in the Bidna hills.

Before leaving this system we must notice a peculiar form which the
quartzite locally assumes at Kalidna, near Dadri, in Jind. The rock is here
extensively quarried for millstones, and in some of the quarries it has become
locally converted into what is known as itacolumite, or flexible sandstone.
The quartzite in its natural form is glassy, and the individual grains of
sand have become coated with an outgrowth of secondary quartz, giving
them an irregular outline when seen in section. Generally the rock appears
to withstand weathering extremely well, and is as hard and glassy a few
inches from the surface, as in the depths of the quarries; locally, however,
decomposition has been able to penetrate into the rock, and it has wea-
thered into a mass of very irregular-shaped aggregates of quartz grains,
held together by the interlocking of their irregularities, but capable of a
certain amount of freedom of movement over each other. There is
nothing to show why this peculiar form of weathering should have taken
place in some places, and not in others. It is not confined to particular
beds, nor is it continuous for many feet along the strike in the same bed?.

Far to the north west of the termination of the Ardvallis, after a wide
interval of plains traversed by the Sutlej and the Ravi, some hills occur on
the sides of the Chenab at Chiniot and Kirdna. These hills are only 40
miles distant from the Salt-range, but the rocks are totally different from
any that occur there, and correspond well with those seen in the Ardvalli
range. They consist of strong quartzites with associated clay slates, form-
ing steep ridges with a north-east to south-west strike, The highest
summit is stated by Dr. Fleming to be 957 feet above the plain, The
rocks seem, from the uncertain observations given of them, to be ina
less metamorphic state than those nearest them to the south east, a fact
which agrees with their remoteness from what is presumably the centre
of disturbance of the region. The oldest rocks of the Salt-range are pro-
bably, from their contrasting petrological conditions, very much younger
than the strata of Kirdna, and, as the former are at least Cambrian, we thus
obtain a small hint of the age of these transition deposits.

1C.A. Hacket, Records, XLV, 288, (1881). H. B. Medlicott, Records, VII, - y
2C. A. Hacket, Records, XIV, 285, (1881); | R. D. Oldham, Recor a XXL, ee
Chap. IIT. ] CHAMPANER BEDS. 73

East of Baroda, at the south-west extremity of the Ardvalli region, there
is an outcrop of rocks which must be referred to one of the transition
formations. It extends some twenty miles east from the Pawagarh hill, for
eight miles south from Champdner, and to a considerable but unknown dis-
tance to the north. The beds of this exposure, while resembling those of
the Bij4war system in general character and state of metamorphism, do not
contain any of its characteristic rocks, while the most remarkable rock
of the exposure is wanting in the Bijdwars. For this reason it is not
possible to refer them to the latter system. So far as can be judged from
the description, thay are more like the rocks of the Delhi system, but it is
impossible to definitely refer them to it, owing to the long stretch of
unexplored ground that separates the two. Under these circumstances it
will be best to treat them under the name of Champéner, from the capital
of the old Mahomedan kingdom of Gujarat, which stands upon their
margin.

The principal constituent of the Champdaner beds is a quartzite or
quartzite sandstone, the other beds being conglomerates, slates, and lime-
stones, with occasional ferruginous bands. The conglomerates are the
most distinctive beds of the Champdner area ; the matrix is a coarse, gritty
sandstone, containing pebbles and boulders, often a foot in diameter, and
occasionally ranging to three feet, consisting of granite, quartzite, talcose
slate, and crystalline limestone, but none of typically Bijawar rocks.
Cleavage, which is well developed in all the beds which are susceptible of
it, is occasionally seen in the pebbles of the conglomerate, but is rarely
‘distinguishable in the matrix.

The passage from the Champ4ner beds to the gneiss appears to be
gradual, so much so that it is frequently almost impossible to determine
where the boundary should be drawn. Within the tract occupied by the
metamorphics, quartzites are found, and a true conglomerate, containing
rolled fragments of quartzite and-very similar to that of the Champdner
beds, is found among the gneiss west of Jambughora. This area has not
as yet been subjected to a close examination, and it is impossible to say
whether the apparent transition is a real one, or the result of the intense
disturbance which both the metamorphics and the Champaner beds have
undergone.}

In the south and south-eastern portion of the country west of the
Ar4valli range there is a series of very ancient eruptive rocks, named after
the Maldni district of the Jodhpur state. They consist principally of very
silicious felsites, so hard that they are not scratched by quartz, and have

1 The description in the text is based on that | siders that the Champdner beds, by their
of Dr. W. T. Blanford, Memoirs, V1, 202,| mineral character and degree of metamor-

(1869) ;— Mr. R. B. Foote, in a letter received | phism should be referred to the Cuddapah
as this work is going through the press con- ; rather than the transition systems.
74 GEOLOGY OF INDIA—TRANSITION SYSTEMS. [ Ghap, III.

frequently the appearance and texture of jasper. They vary greatly in
colour, from black or dark-brown to pink, blue, or white, the dark-coloured
rock being always hard and undecomposed, whilst the light-coloured varie-
ties are softer and appear to be altered. The most constant character
is the presence of small crystals of felspar, usually of a pink or red colour,
in addition to which small grains of transparent silica are frequently
disseminated throughout the rock’, Their extremely silicious nature may
be due to alteration, but their porphyritic character, and the occasiona]
occurrence of ash beds, sufficiently attest their volcanic origin,

In places diorite was found associated with these rocks, and in some
of the hills west of Balmer coarsely crystalline granitoid syenite and peg-
matite are intercalated in large masses with the porphyritic felsites. True
granite may occur, but in the few hills examined mica was absent, al-
though the character of the rock was distinctly granitic. The presence of
similar granitoid rocks elsewhere is rendered probable by the occurrence
of pebbles and boulders in some of the later formations.

The Maldni rocks must be very ancient, but no idea can be formed of
their geological position, as they are nowhere associated with rocks of
known age, except where underlying beds of comparatively recent date,
and nothing resembling them appears hitherto to, have been detected
elsewhere in India,

They have been regarded? as of lower Vindhyan age, since they occur
undisturbed in close proximity to the highly disturbed slates and schists
of the Ardvallis. They have not as yet been found in actual contact with
the older rocks, but small hills of both are found standing up from the
recent alluvium and irregularly interspersed with each other. On the
other hand, nothing at all resembling the Mal4ni felsites has as yet been
found in the lower Vindhyans east of the Ardvallis. The general type is
that of a much more ancient rock, and felsites, closely resembling those of
Jodhpur except that they are not porphyritic, are found in the Toshém
hill*, dipping at high angles with the older beds of the Ardvalli range.
Besides this, the unconformity between them and the overlying sand-
stones, which are regarded as upper Vindhyans, is most marked, con-
trasting with the very much less pronounced unconformity between the lower
and upper Vindhyans of the typical area. The correlation of the sand-
stones with the upper Vindhyans is, however, conjectural, and if they
belong to the uppermost members of the system, this unconformity would
not be inconsistent with a lowermost Vindhyan age for the Maldnis, The
age of these last must remain doubtful for the present, but they appear
to belong to the transition rather than the Vindbyan rocks.

1 W. T. Blanford, Records, X,1 877). | 3
? Records, XIV, 303, (1881), aT a Metiahan, Records, XIX, 164, (BE).
Chap, III.] CORRELATION OF THE TRANSITION SYSTEMS. 95

Reference has just been made as to the doubtful propricty of classing
the Maldni beds with the transition systems and a similar doubt may
almost be expressed with regard to the Gwalior and Delhisystems. The
former of these finds its nearest analogue, as regards both mineral com-
position and degree of induration, in the Cuddapah system of Southern
India, and Mr. Foote’s suggestion that the equivalents of the Champdner
beds must be looked for among the Cuddapah rather than the Dhdrwdr
deposits of Southern India has been referred to.! As will be noticed in the
next chapter, there are grounds for questioning whether the Cuddapah
system should not be classed with these, among the newer of the transition
systems, rather than with the Vindhyans, However this may be, there
seem good reasons for accepting the Delhi and Gwalior systems as the
newest of those described above. Next after them would come the Bijawar
and Behar systems, the latter being the older of the two, and finally the
transitior rocks of south-west Bengal and the Dhdrwdrs of Southern
India, the last cf these being marked out as the oldest by the greater
degree of disturbance and metamorphism it has undergone, as well as by
the manner in which the eroded edges of its upturned and metamor-
phosed strata are covered by the nearly horizontal basement beds of the
Cuddapah system.

There can be little doubt that rocks corresponding to the transition
systems will be found extensively developed in the extra-peninsular
mountain ranges, but as yet these have not been sufficiently explored to
allow of their separation as distinct rock series, except in a few isolated
localities.

In Hundes and Spiti Mr. Griesbach has separated, under che name of
Vaikrita?, a series of beds-which overlie the granitic gneiss. It is described
as of great thickness, varying much in lithological composition, composed
principally of micaceous schists, talcose rocks, phyllites and gneiss. The
beds are now found occupying the cores of highly compressed synclinal
folds, the crests of the intervening anticlinals having been denuded away
till there is now an apparently continuous succession of strata across the

folds.

Somewhat similar schistose beds occupy large areas in the central part
of the range, and appear to extend far towards its southern margin in
Nepdl®.

In the Darjiling district Mr. Mallet has described a series of beds, said
to be transitional with the gneiss, under the name of the Daling series.

1 Supra, p. 73. foot-note. phosed, Memoirs, XXIII, 41, (1891).
2 Said to be the Sanskrit for metamor- 3 Records, VIII, 93, (1875).
76 GEOLOGY OF INDIA—TRANSITION SYSTEMS. [ (Chap. IIT.

They consist of light green, slightly greasy, slates, sometimes interbanded
with adark greenish grey kind, passing insensibly into ordinary clay slates
and more or less earthy or silvery according to the degree of alteration
they have undergone. There are also bands of quartzite and quartz flags,
occasionally some hornblende schist, sometimes slightly calcareous and
passing into an impure dolomite containing crystals of actinolite. This
is, however, a rare and exceptional rock, the most prominent lithological
distinctions between these and the succeeding Baxa series being the almost
complete absence of lime and the rarity of the brilliantly-coloured alterna-
tions of slates.!

The distribution of these beds is peculiar, and led to an erroneous idea
of their position being formed in the first instance. They occur along the
outer (southern) edge of the gneissic masses of Dérjiling and Daling, in
the valley of the combined Tfst4 and Ranjit rivers. They separate these two
areas of gneiss and extend on the northern side up each of the valleys,
dipping inwards towards the gneiss on all sides, and the junction is de-
scribed as transitional, except for a portion of the boundary north of Darjil-
ing, which is faulted. On the south the apparent passage is somewhat
rapid, but on the inner sections the Dalings are more metamorphosed, and
the distinction between them and the gneiss more difficult to draw. The
form of the outcrop and the direction of the dips combine to convey the
impression that the Darjiling gneiss lies in the centre of a synclinal and is
newer than the Daling series.

At the time the description was written a belief in the possibility of
regional metamorphism, that is to say, of ordinary sedimentary rocks
being converted, within a moderate distance, into true schists and gneisses,
was still held by many geologists, but the whole tendency of recent
investigations has been adverse to this opinion, and the opinion now
prevalent is that of two contiguous series of beds the one which exhibits
the greatest degree of metamorphism is primd facie the older. Added
to this, the apparent dip of newer beds under older is a common feature
of Himdlayan sections, and when we find that the apparent relation of the
gneiss to the Dalings is the same as of these to the Damudas, and again
of the Damudas to the tertiaries, it is impossible to escape the belief that
the true sequence is the reverse of the apparent one,

| Memoirs, XI, 4o, (1874),
CHAPTER IV.

OLDER PALHOZOIC (CUDDAPAH AND VINDHYAN)
SYSTEMS OF THE PENINSULA.

Older palzozoic rocks —Reason for adopting the name—Soutwern {yp1a—Cuddapah system
—Cuddapah area—Kalddgi area— Karntl series—Cuddapah area— Bhima area— Godavari
valley—Pakhal series Pengang4 beds— Chhatfsgarh—Sullavai series—CenTRat INDIA—
Lower Vindhyans—Upper Vindhyans—Relation of upper Vindhyans to the ArAvalli
range—Vindhyans west of the Ardvalli—Source of the diamond—Relative ages of the
rock systems described.

In dealing with the newer group of systems, intervening between the
gneiss and the lowermost fossiliferous beds of the Peninsula, we are met
by the same difficulty as with the transition systems,—the absence of
any fossil evidence by which we can judge of the true position of the beds.
In this case the absence of fossils is the more extraordinary as many of the
strata appear well adapted for the preservation of organic remains and
have undergone no disturbance which could account for their subsequent
obliteration.

The selection of a general name for the beds described in this chapter
is a difficult task. Omitting purely local names, they have been classed as
upper transition, azoic, or Vindhyan, but none of these are completely satis-
factory and the best course to pursue will be to take into consideration the
strongly marked unconformity that exists between the newest of them and
overlying beds, of upper palzozoic age, together with their general litho-
logical character, and class them as older paleozoic. This much we know,
that they must be considerably older than permian, but it is as impossible
to decide whether some of the oldest may not be precambrian, as to
determine whether they may not to some extent be contemporaneous
with part. of those classed with the transition systems,

The older palzozoic strata, as defined here, are principally developed
in two separate areas, one in the Madras presidency, the other in
central India. There are besides a number of exposures in the Godavari
and Mahdnadi valleys in which the beds are not so well exposed and have
been less studied than in the two principal areas.

There can be no doubt that the oldest rocks of this group of systems
are those that have been described as the Cuddapah system in Madras,
and they will consequently stand first for notice here.
78 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. {Chap. IV,

The rocks of the Cuddapah system occupy a large area about the
middle of the east side of the Peninsula, where the coast line bends from
a northerly to a north-easterly direction, This feature is probably con-
nected with the form of the Cuddapah ‘basin, which is of a roughly
crescent shape, convex to the west. The north-east horn of the crescent
is known as the Palndd, and reaches to Jaggayyapet, a few miles nort
of the Kistna river; the southern termination at Tirupati (Tripetty) hill
is 30 miles north-west of Madras, or only 18 if measured to the outlier at
Ndgari Nose. The town of Cuddapah stands in a south-central position
near the Pennerriver. Karntl is on the northern edge and, further south,
Gooty is just outside the western border, at its centre. The length of the
basin is about 210 miles and its width 95, the area being nearly 13,500
square miles,

The eastern edge of the basin constitutes a well-defined segment of
that vaguely expressed general feature known as the Eastern Ghats.
The actual face of the highlands is locally known as the Yellakonda
ridge. It is a flanking member of the Nallamalai range, which is formed
by a belt of contortion of the Cuddapah rocks along this side of their basin.
Between the hills and the sea there is a zone of low country, formed of
metamorphic rocks and alluvium, about 50 miles wide, constituting the
plains of the Carnatic, or Payan Ghdét (country below the Ghats), in the
Guntdr, Nellore, and North Arcot districts. The elevation of this ground
at the base of the hills is under 200 feet, the crest of the Yellakonda rising
to about 1,000, and the summits of the Nallamalai to 3,500. The centre of
the Cuddapah basin is occupied by the broad valley of thé Kundair, the
rocks rising again to form a steep range along the western margin of
the basin, 2,000 feet above the sea and overlooking the gneissic upland
of Mysore and Bellary, the elevation of which near the range varies
from 800 to 1,800 feet. The Madras railway enters the basin at Gooty
and leaves it at the southern point of the crescent, while the Kistna
river adopts a very similar course in the northern limb. The watershed
‘of the basin lies far to the north, and the Penner receives most of the
drainage. i

More than athird of the area, within the boundaries indicated, is taken
up by the overlying Karnil series, which occupies all the low ground of
the Kundair valley in the middle of the basin and another large space
in the Palnad.

The Cuddapah formation has been divided into the following groups! :—

sage (Srishalam).
- «Slates (Kolamnala).
Quartzites (Irlakonda).

Slates (Cumbum).
* (Quartzites (Bairenkonda).

Kistna group, 2,000 feet

Nallamalai group, 3,400 feet ‘i ‘

1 Memoirs, VIII, 126, (1872).
Chap. IV.] CUDDAPAH SYSTEM. 79

( Slates (Pullampet).
"U Quartzites (N agari).

Pépagh group asofeet «sf ote

The groups are all.more or less unconformable to each other, and all in
turn overlap the others and rest directly on the gneiss, but there is so
marked a unity of character running through all that it is necessary to
regard them as a single system.

Cheyair group, 10,500 feet . .

The distribution of these groups relieves us in some measure of the
enormous aggregate thickness of 20,000 feet given in the list. Although
the succession may be taken strictly in order of time, it is scarcely to be
supposed that there was ever at one spot a continuous superposition of
these strata to the extent of their aggregate thickness, Even within the
present rock-basin, which must be taken as only a part of the area of deposi-
tion, the groups are local and discontinuous, each in turn overlapping the
one below it and resting on the gneiss. In each case, however, there is
more or less of denudation-unconformity, as well as overlap, so that the
groups are much more than mere horizons of variation in deposition.

The original characters of deposition, and the induced characters of
disturbance, are closely related to the actual boundaries of the field. All
round the western boundary the junction is natural, and the deposits
rest as originally laid down upon the gneiss, the strata having undergone
comparatively little disturbance. On the east side of the basin, on the
contrary, there has been much contortion of the strata, the boundary is
represented as faulted and the beds often inverted. The lower groups
are found to the south-west, and are gradually overlapped to the north
and east.

In each of the groups of the Cuddapah series sandstores or quartzites
prevail at the base and earthy deposits forming shales or slates above,
limestones often occurring with the latter, The Pdpaghni group is only
found between the Tungabhadra and the Cheyair, being overlapped in both
directions by the Cheyair beds. It takes its name from the river, in the
gorge of which the best sections are seen. Its bottom member, the Gul-
cheru quartzite, rests upon an uneven surface of the gneiss, and rises up
to the west to form steep cliffs, over an undercliff of the crystalline rock.
Although the contact is quite sharp the two rocks are often connected
together into an adhering mass. A considerable thickness at the base is
coarsely conglomeratic, the pebbles consisting of the brecciated veinstones
and banded jasper-rocks, which form prominent outcrops in the adjoining
metamorphic area, but no pebbles of gneiss or granite were found except
at one spot.!_ These bottom beds are described as shore deposits.

In the Vempalli subdivision of the PA4paghni group limestone Is

1 Memoirs, VIII, 158, (1872).
80 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. (Chap. IV.

largely associated with the shales, and intrusive sheets of trap are
also of frequent occurrence. In contact with, or near, the trap the lime-
stone often contains bands of serpentine and steatite, as may be seen
close to Karndl, where the Vempalli band has overlapped the bottom sand-
stones, and rests directly on the gneiss.

The Cheyair group is well exposed on the Cheyair river. It is divided
into two areas by the Karnul formation stretching southwards, west of
Cuddapah, into contact with the Pd4paghni rocks. The constitution and
telation of the Cheyair group in the two positions are somewhat different.
Inthe north-west area, traversed by the Penner, the bottom band of sand-
stones and conglomerates is comparatively unimportant. It is there described
as the Pdlivendala (Pulavaindla) subdivision, from a town 4o miles west by
south of Cucdapah. North of the Kistna it overlaps the Vempallis, and
rests upon their denuded surface in the Penner ground, the conglomerates
and breccias being largely made up of the characteristic chert-bands of the
Vempalli limestone. Here, too, intrusive sheets of trap occur in the Pdli-
vendala band. The corresponding beds in the southern area are described
as the Ndgari quartzites, from the well-known hill near Madras. They
form for the most part the bottom-rock of the Cuddapahs resting on the
gneiss in this region. The conglomerates are here made up of pebbles
of quartz and quartzites (which are themselves sometimes conglomeratic),
and occasionally of red-banded jasper, being thus more like the Gulcheru
beds of the Penner area.

The upper dand of the Cheyair group in the Penner area is described as
the TA4dputri (Todapurti) beds, named from a principal village of the district.
They comprise a great series, in which slaty shales predominate, with
limestones, eruptive rocks both intrusive and contemporaneous, ferru-
ginous chert, and jasper beds. Although not greatly disturbed, the shales
are to some extent affected by cleavage and are hence qualified as slaty.
Limestone occurs in two principal bands. It is a finely crystalline grey
rock, with much segregated chert, which often assumes very fantastic
shapes, especially in the upper part of the beds and near trap-flows. Of
these eruptive rocks there are many strong outcrops, in two principal
bands, a main one near the base of the group, and another two-thirds up.
The only rocks that can be certainly classed as ‘eruptive are coarse-grained,
dark, basic diomrites, someties compact and of grey or pale-green colours,
They are shown to be contemporaneous by their outcrop being continuous
for long distances between well-marked bands of aqueous deposits, but
the intervening deposits frequently cease, and the flows locally coalesce ;
moreover, they are distinctly confluent with intrusive dykes, as is well seen
in the small bay below the southern flanks of the Opalpdd plateau, 20 miles
east of Gooty Perhaps the strongest argument for the contemporaneity of
the bedded traps on this horizon is the fact that no intrusive igneous rock
Chap. IV.) NALLAMALAI GROUP. 81

is known to occur higher in the formation, or in the Karndls, and this
could hardly be the case if the massive bands in the Tadputri zone were
intruded after the completion of the sedimentary series.

In this group there are, associated with the traps, porcellanic beds re-
sembling those of the Gwalior system. They have been regarded as of
volcanic origin, but there is a great difficulty in supposing so highly silicious
an ash could be produced by the same series of eruptions as gave birth to
the unmistakeable igneous rocks in the section, They do not in any way
resemble any known product of volcanic activity and their associations with
the lava flows is probably fortuitous.

In the Cheyair area the Pullampet slates and limestones represent the
TAdputri beds of the Penner. The traps and porcellanic beds are absent.
The limestones are again silicious, and sometimes they are brecciated in a
very unaccountable manner, without any disturbance of the strata. Some
beds present a rugged humpy surface, suggestive of a coralline formation,
but no organic structure has been detected.

The Nallamalai occupies a larger area than the other groups, principally
on the east side of the basin, and takes its name from the range. The
Bairenkonda summit, 3,500 feet above the sea, gives its name to the bottom
band of quartzites. In the Pdlkonda range, east of Cuddapah, these
quartzites rest with slight unconformity upon the Cheyair group. In the
Penner area the strong quartzites of the Gondicotta hills, overlying the
TAdputri shales, are on the same horizon. Here the beds have a gentle
north-easterly slope and pass under the Karndl formation, but when they
rise again to the east, in the Nallamalai, contortion is the rule, often to so
exireme a degree as to produce folded flexures and inversion. In the
synclinal troughs of these contortions the upper member of the group,
called the Cumbum slates, is found, the underlying quartziies rising up to
form the ridges.

The Cumbum slates are by far the thickest member of the group, and
cover the greater part of the area, They are not very uniform in compo-
sition. There are several subordinate bands of quartzite, which it is not
easy in broken ground to distinguish from the underlying Bairenkonda
rock, and the slates themselves present many vatieties, from fine, silvery,
talcose beds to coarse, earthy clay slates, of many shades of colour. Occa-
sionally they are foliated and schistose, and not easily distinguished from
the schistose heds of the adjoining gneissic area, when the two happen to
come in contact. As arule, however, quartzites are found at the junction,
Strong bands of limestone are frequent in the Cumbum slates. It is generally
compact or finely crystalline, micaceous or talcose, of a slate grey colour,
with pyrpletinges. The oldlead mines near Nandidlampet, 16 miles north
of Cuddapah, occur in a dark silicious variety of this rock.

At the north end of the Nallamalai, just south of the Karndl and Guntdr

G
&2 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [Chap. IV.

road, there isa great dome-shaped mountain known as Eshwarakupam. It
is composed of lower Cuddapah rocks dipping away from the hill on all
sides, and surrounded by Nallamalai beds. A great thickness of strata is
exposed, but it is not easy to identify them specifically with the groups
already described,

The plateau through which the Kistna has cut its gorge, known as the
Kistna Nallamalai, is formed of beds higher than the Nallamalai group and
unconformable to it. These beds are therefore distinguished as the Kistna
' group. They comprise three well-marked divisions ;. the Irlakonda quart-
zites, forming the plateau of that name on the west, where they are
1,200 feet thick; the Shrishalam quartzites, forming a higher plateau
to the north and east, called after a well-known shrine on the Kistna;
and the intermediate shales, which are called Kolamnala, after a stream
that traverses them. To the north the group spreads out over a flat
surface of gneiss, and to the east it passes under the Karndl beds of
the Palndd, in which region the rocks are, again, intensely disturbed on
the east. This group is supposed to be also represented further south in
the Nallamalai, but the evidence is not decisive.

In the south Mardth4 country, on the southern border of the great area
occupied by the Deccan trap, and in great part separating the trap-region
from the gneissic area of Mysore, there is a basin of somewhat similar
rocks named after the town of Kalddgi!, which lies near its eastern end. Its
peculiar position is ina manner accidental, for it is certain that the whole
of this basin was once overspread by the trap, which still stretches con-
tinuously along the crest of the SahyAdri for some distance to the south, and
elsewhere outliers of trap are found resting on the gneiss. ‘The strata
rest with total unconformity on the crystallines, quite unaffected by meta-
morphism, and are considered to belong to the Cuddapah system on the
strength of a general résemblance in lithological character, although the
patticular sub-divisions of the Cuddapah area cannot be recognised.

From the Kistna, below its confluence with the Gatparba, the Kalddgi
rocks stretch continuously westward for more than roo miles and then
disappear under the trap forming the crest of the Sahyddri. In this direc-
tion several inliers are exposed by the local removal of the basaltic cover-
ing, the largest of which, at the foot of the Phonda Ghat, in the Konkan, is
probably continuous with the main basin. On the north there is a large
inlier at Jamkhandi. In all of these inliers, however, only the lower beds
occur, so it is probable that the formation does not extend far beneath the
trap. On the south borders of the basin there are numerous outliers of
the bottom auartzites resting on the gneiss, both on the uplands of the

1 R. B, Foote, Mewoirs, XII, 70, (1876).
Chap. IV.] KALADGI SERIES, 83

Deccan and in the Konkan. The Vengurla rocks and other small islands
off the coast all consist of the very hard rocks belonging to the quartzite
series. The former continuity of all these patches of rock cannot by any
means be asserted, for it is evident that the deposits took place upon a
very uneven surface of the crystallines, of which there are extensive inliers
within the main basin, as at Gok&k.
The series is divisible as follows! :—
Upper Kalddgié.
Thickness,

6. Shales, limestones and hzmatite-schists - ‘ - 2,000 feet
5. Quartzites, local conglomerates, and breccias + 1,200—1,800

Lower Kalédgt.
4. Limestones, clays, and shales, . < + 5,000—6,000
3. Sandstones and shales . ‘ . 7
2. Silicious limestones, hornstores, or cherty breccias} 3,000—5,000
1. Quartzites, conglomerates, and sandstones .

The bottom conglomeratic rocks are made up of the debris of the ad-
Joining crystallines, and vary with the composition of the latter. They
generally slope up towards the boundary of the area and form a scarp over
a basement of gneiss. The cherty breccias form the most peculiar and
conspicuous member of this part of the series. Mr. Foote suggests, with.
much probability, that they are formed by the decomposition and crushing
of the highly silicious limestones that occur on the same horizon, A
large proportion of the total area, forming a continuous margin to the basin,
very wide on the south, and including all the outliers, is formed’ of the lower
members (Nos. 1, 2, 3) of the series, and in this position the rocks are very
little disturbed, and scarcely at all altered.

The limestones and shales forming the fourth division of the Kald4dgi
series are only found in a special basin of depression and contortion on
the north-east side of the area. They generally occupy low ground and
are much concealed, but may be fairly seen about the town of Kalddgi,
exhibiting much disturbance. _Several varieties of the rock are very homo-
geneous in texture and variously tinted, making pretty marble.

The only remnants of the upper Kalddgi group are found in the axes of
synclinal flexures within this special basin, their preservation being evi-
dently due to their being thus let in and encased by the folding of the
whole series, so that the maximum of disturbance and of metamorphism is.
exhibited in these remains of the topmost beds of the formation. The
principal of these elliptical synclinal areas of the upper groups are all within
a short distance of Kalddgi. The direction of the axes of disturbance is very
steady between west by north and west-north-west. This is also the
direction of the major axis of the basin itself, in which all the special

Memoirs, Xi1, 71, (1876).
Gz
84 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN  [Chap. IV,

contortion seems to have been concentrated on the north side, along what is
now the lower valley of the Gatparba.

Only four cases of intrusive rock have been observed in the Kalddgi
area, and all in the region of disturbance, in the highest beds ; three in the
Lokapur basin, and one in the Arakere synclinal valley. They are of com-
pact, green diorite, unlike the older diorites of the gneissic area.

The rocks of the Karndl series lie almost entirely within the basin
of the Cuddapah system, where they are found in two separate areas, the
larger of which occupies the whole of the Kundair valley and stretches to
beyond the Kistna, while the other lies in the district known as the Palndd.
The series has here a total thickness of only 1,200 feet, less than that of
the smallest group of the Cuddapahs, and might be regarded as a member
of that system, a view which has been urged’ on the ground that the
unconformity between the two is not much greater than those between the
different sub-divisions of the older system, and that on the east side of the
basin the Karntl series has felt the full effects of the disturbance which they
have undergone.

Such was not the opinion of the actual observers, who described the un-
conformity between the Cuddapah system and the Karnil beds as sufficient to

Nby W- |
SS a ot $
6 OK SOS SS BR
= >
te ‘Co: td

Fig. 5. Sketch, section illustrating the relation of Cuddapah and Karnil rocks, after King.

 

 

justify the separation of the latter.?_ This claim is strengthened by the oc-
currence of an outcrop of beds, referred to the Karnil series, under the edge
of the Deccan trap plateau in the valley of the Bhimd river. Although the
westernmost point of this basin is separated by only 8 miles from the
boundary of the Kalddgi area of Cuddapah rocks, and though each basin is
over 100 miles in length, yet no representative of any of the rocks found in
the one has been recognised in the other. If the reference of the rocks
of one area to the Cuddapah system and the other to the Karniul series, on
the strength of general petrographical resemblances, is to be trusted, this
indicates a change in the areas of deposition of the two periods sufficient
to justify the separation of the rocks formed in each.

1 Manual, 1st ed., p. 70. least, shows the interpretation put by the actual

2 The nature of this unconformity is exhibite | observer on the observations made by him.
ed by the sketch section, fig. 5,which, at the | ~W. King, Memoirs, VIII, 125, (1872).
Chap. IV.] KARNOL SERIES. 85

The Karnidl series, which is mainly a limestone formation with subor-
dinate bands of sandstone and shale, has been divided into the following
groups! :—
¢ Shales (Nandial).

Bandar 2503p * 5 7 7 *U Limestones (Koil Kuntla).
: Pinnacled quartzites.
Pa eee : eS : | Plateau quartzites.

Jamalamadiigt group e. te t6> -{ ae eS .

Banaganpalli group . ‘ : . . Sandstones,

The Banaganpalli group consists of sandstones, generally coarse, often

earthy, occasionally felspathic or ferruginous, and usually of dark shades of
ted, grey, and brown colours. Pebble beds are frequent, the pebbles being

small and numerous, composed of quartzite and various coloured cherts,
Jaspers and hardened shales, evidently derived from the cherty shales
of the Cheyair group, on which the Banaganpalli beds rest.

The Banaganpalli beds are of interest as beirg the principal source of
the diamond in the Cuddapah area. There are many places on or near
the Karnil group where diamonds have been worked for in surface
gravels, but at Banaganpalli these workings are carried on in the solid
rock. Shallow pits, not more than 15 feet deep, are sunk in the rock;
which is hard and quartzitic at the surface, but turns soft and easily worked
underground, where short galleries are driven in the diamond layer,
at, or close to, the base of the group, ‘The diamonds occur in some of the
more clayey and pebbly layers. Dr. King has recorded the opinion that
they are innate inthis rock, an opinion hased principally on the perfection
of many of the crystals. In the case of so hard a mineral the argument is
not conclusive, and the nature of the rock in which they are described as
occurring certainly suggests that they are of detrital origin. It is rather
mysterious why the rock-workings should be so crowded as they are over
certain spots, whilst large adjoining areas of apparently the very same
deposits are left quite untouched. If this irregular distribution of the
mines be only due to a delusion of the diamond-seekers there is still a very
large field awaiting exploration.

The Jamalamadtgi group takes its name from a large village on the
west side of the Kundair valley. It is composed at the top of buff, white,
and purplish non-calcareous shales, well seen near the village of Auk
(Owk). They have a maximum thickness of 50 feet, and pass down
gradually into a finely crystalline or compact limestone, generally blue-
grey, sometimes nearly black, and occasionally of pale buff and fawn
colours. A very inferior lithographic stone used to be obtained from these

4 Memoirs, VIII, 30, (1872). | 2 Memoirs, VAI, 96, (1872.
86 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [OQhap, VI.

beds, and the rock is now much used for building, large quarries having
been opened near the railway at the village of Narji, by which name the
stone is known. West of Banaganpalli the Narji limestone is about 4oo
feet thick, but thins out both to the south and north. In the Rdichdr
Dodb, about Karnal, it rests on the metamorphic rocks, where it becomes
cherty and brecciated in a peculiar manner and is described as a shore
deposit.

Between the open Kundair valley and the western ranges, or Yel-
lamala, there are in the Karnidl district some low flat hills, such as the
plateaux of Upalpd4dd and Undutla, These low plateaux are composed
of a sandstone or quartzite, locally intercalated in the Karnil lime-
stones and known as the Pdniam group, after the town of that name.
The greatest thickness of the quartzites is only 100 feet, and the group
disappears altogether to the north and south, nor has any sign of it been
observed on the eastern edge of the basin. An upper portion, formed of
firm white sandstone, has been distinguished as the ‘ pinnacled quartzites’
from its mode of weathering, the lower beds, or ‘ plateau quartzites,’ are
coarser, more earthy and ferruginous, of various rusty tints.

In a basin of slightly disturbed strata the uppermost group must cover
the largest area, and so the Kundair beds occupy the whole valley of the
Kundair. There is a thickness of 500 to 600 feet. The upper two-thirds
of purple calcareous shales and earthy limestones, distinguished as the
.Nandial shales after a large village of that name, pass insensibly down
into purer, compact and crypto-crystalline, flaggy limestones, known as
the Koil Kuntla band, from atown 10 miles south-east of Banaganpalli, in
which small papillz resembling casts of Cypris, and numerous discoidal
markings of half an inch to two inches in diameter are found,.' The town
of Cuddapah and all the large villages in the centre of the valley are on
the Nandi4l shales. In this position the rock is soft and crumbling, but
to the south and east, on the margin of the mountain region, these upper-
most beds of the whole sedimentary basin are quite slaty, being cleaved
and contorted proportionally with the underlying formations. The litholo-
gical character of this group, as insome of the Cuddapah groups, changes
to the north-west, and in the proximity of the metamorphics the Koil
Kuntla beds are described as shore-deposits, which never extended much
beyond their present boundary.

In the Palndd there is a large exposure of limestones which are be-
lieved to be of Karntl age, and even the sub divisions have been in a manner
specifically recognised in the south-west part of the ground.? The limestone
is underlaid by a diamond-bearing sandstone, which has consequent ly
been supposed to represent the Banaganpalli rock. In the Palndd country,
however, there is great difficulty in distinguishing this rock from a closely

Memoirs, VII, 46, (1872). | 2 Memoirs, VII, 197 (1872).
Chap. IV.] BHIMA SERIES. 87
associated sandstone, clearly belonging to the Cuddapalis, but of the Kistna
group at the very top of the Cuddapah series and several thousand feet
higher, stratigraphically, than are the beds of the Cheyair group underlying
the diamond sandstone of Banaganpalli; such at least is the position made
out for the bottom sandstone on the south-west of the Palndd, towards the
expanding rock-basin. On this side, too, some slight unconformity has
been pointed out between the Palndd limestone and successive masses of
the sandstone, and it has been remarked that the diamond workings here
are confined to the rock close under the limestone, so as to suggest the
limitat ion of diamonds to the horizon of the Banaganpalli group. All round
the north-east corner of the basin, however, this sandstone, there known
as the Jaggayyapet quartzite, rests directly upon the gneiss.

The leading structural character of the Cuddapah basin is maintained in
the Palndd. On the west side the strata are comparatively undisturbed,
while on the east border they are cleaved, foliated, and contorted, and
appear to be overlaid by a natural ascending sequence of shales, lime-
stones and quartzites, above what have been described as the Palndd
limestones, and so these upper rocks would be newer members of the
Karnil formation. According to another, more probable, view this sequence
is deceptive, being due to total inversion of the strata, the top quartzite
being really a Cuddapah rock.

On the north-western border of the Cuddapah basin the Karndl deposits
are described as overlapping the formations upon which, for the most part,
they rest, and lying upon the gneiss for a short distance up the Kistna
valley. Seventy-five miles further in this north-westerly direction there
is another area of rocks, having a strong likeness to the Karnal deposits,
and resting throughout their entire south-east border, for a distance of
more than 100 miles, immediately upoa the gneiss, while along their
entire north-western border they are covered by the Deccan trap. The
width of the basin thus exposed is exceedingly variable, both bound-
aries being very irregular in outline. It is greatest, about 25 miles, where
the Bhima river crosses the outcrop nearly at its middle, and from this
circumstance the name of the river has been taken for the local designa-
tion of the rock basin.?

The Bhimd series is mainly a limestone forma ion which has been divid-
ed as follows in the central portion of the basin :—*

Upper.
PP Thickness

(g) Red calcareous shales s . 5 : . » 30 feet.
(f) Fiaggy limestone beds ‘ . ‘: -
(e) Buft shales . ° . . . . ° ° 18 ”

1 Memoirs, X11, 139, (1876). | 9 Memoirs, XI, 160, (1875).
88 GEOLOGY OF INDIA—CUDDAPAH AND vINDHYAN [Chap, IV.

Upper—contd,
Thickness,
(d) Quartzites (sandstone) . 200 feet.
(c?) Blue thick-bedded splintery li dinestone, txeneaiodis in aa 200 55
(c}) Thin-bedded limestone, with chert 7 20 95
(c) Blue and grey splintery limestone, dceatanaly Bice.
ciated F 7 * ‘ 3 é ‘ é 200
Lower.
(4) Purple, red, drab, and dark-green shales, with calcareous
flags at top . . . . a j 100 5
(a) Quartzites (sandstones) and conglomerates . ‘ . 60 »

It is principally in the south-western part of the area that the bottom
sandy beds are developed to any extent. ‘The pebbles of the conglo-
meratic bands are derived from the adjacent metamorphics, upon a very
uneven surface, of which the Bhima deposits were laid down, as is shown
by the very winding outline of the boundary and by the occurrence of
. gneissic inliers, some of which are found near the trap of the north-west-
ern edge of the area. There is thus no presumption that the sedimentary
basin extends far beneath the eruptive rock.

At Bachimali, the extreme easterly point of the southern expansion of
the Bhima basin, there is a basement pebble-bed much resembling the
diamond layers of the lower Kistna valley. It is much broken up by
small pits, as if at one time it had been searched for diamonds, but there
seems to be no local tradition of any having been found.

The upper quartzite is quite a local intercalation, so that in some sec-
tions the series is almost exclusively made up of limestone. This is a very
fine-grained rock for the most part, with a texture approaching that of
lithographic stone. The colours are very various. Grey prevails, but
drab and pink tints are common. The rock generally occurs in flaggy
beds, and is much used for building, the pale cream coloured variety
being preferred, although the grey stone is the more durable.

The formation has undergone very little disturbance, and the inclina-
tion of the strata very rarely exceeds from 2° to 5°. Ata few places near
the boundary some crushing and faulting has taken place, as at Gogi,
where the lowest beds are vertical.

There are some patches of a singular limestone-breccia resting on the
gneiss within the confines of the Bhima basin, as west and north of the
village of Yeddihali in the Agani valley. The brecciation has clearly been
caused 7” situ, and Mr. Foote conjectures that these patches may be rem-
nants of a former spread of the Kalddgi rocks.

With the exception of a doubtful fragment of silicified wood (or bone)
found by Mr. Foote in the basement conglomerate close to the village of
Kasakanahal, just within the Agani valley, no traces of organic remains
Chap. IV.] PAKHAL SERIES. 89

were obtained from any of the Bhfmd rocks. Mr. Foote speaks of the
limestones as a pelagic formation, and remarks that there is a good deal
to suggest that they were once continuous with the like rocks of the
Karnil area, and that they have been separated only by denudation.

North of the Kistna some outliers of highly disturbed quartzites and inter-
bedded limestones have been recognised as belonging to the Cuddapah
system}, and north of these a large spread of Cuddapahs, originally
described ? as the Pakhal series, extends up the Godavari valley.

In the hills east of the Goddvari at Albaka the series is described as
consisting of two: well-marked members, namely, a more or less slaty
division, with many strong bands of altered arenaceous beds, and at
least two bands of limestone, best exposed near Pakhal; and an upper
division which is more generally arenaceous in its composition. The.
thicknesses of the divisions, where best developed, may be reckoned as—

2. Albaka division . : : ‘ ‘ . ‘ 2,500 feet.
1. Pakhal division . . ‘ ri 5 . ; 5,000 ,,

Near Pakhal the lower division can be divided into sub-groups as
follows :—~

5. A slaty band with thick seams of quartzite sandstone . 3,600 feet.

4. Grey and fawn-coloured silicious limestones é . 300 ,,
3. Clay slates and quartziles ; . ‘ : 7 500 4,
2. Silicious limestone . ‘ . 3 ‘ . 150 .,
1. Quartzites, with a few slates . ‘ i . - JOO 45

These sub-groups are not constant and overlap each other, The lowest
beds are frequently conglomeratic and at times pass into strong conglo-
merates, which rest unconformably on the gneiss, The upper division,
more fully represented east of the Godavari in the hills which run parallel
to the river and touch it at Albaka, is described as essentially a sandstone
and quartzite formation. :

The Pakhal outcrop extends in a north-west direction to the Maner
river, a short way beyond which it is abruptly cut off by a fault. The
rocks become gradually less indurated in this direction, the shaly beds
finer in grain, and the limestones less prominent. In the Maner valley
the shales, fine in texture, green and purple in colour, and much banded
with thin calcareous seams, closely resemble those of the Cheyair group.

On the east side of the Godavari there is another large area extend-
ing from about 10 miles south-east cf Albaka for over 100 miles to the
north-west. In this exposure the upper arenaceous beds are largely de-
veloped, but there appears to be a slight unconformity in the cliff sec tions

' Records, XVIII, 20, (1885). | 2W. King, Memoirs, XVIII, 209, (1880).
go GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [Chap, IV,

on the west of the range of hills, and it is not impossible that part of what
has been here classed, and coloured on the map, with the Pakhal series be-
longs in reality to the Sullavai service which succeeds it.

In the upper Prdnhfta valley, west of the great Wardha valley coal-
field, a series of limestones and shales of ancient date, known as the
Pengangd beds, have been recognised by Dr. King as identical with the
Pakhal beds of the Maner valley. They fall naturally into a lower lime-
stone and an upper shaly group.

The limestone group consists of pale or dark grey or buff-coloured,
seldom tred_ limestone, well-bedded,, with occasional layers of ribbon
jasper. It is overlaid by the shale group, composed of fine-grained earthy
shales, usually some {shade of red in colour, with occasional beds of
flaggy limestone. The shales have often a nodular structure and weather
into small thin discoidal fragments like tke Talchir shales, for which an
isolated outcrop might easily be mistaken, It is extraordinary that no
fossils have been found in these beds, whose texture is eminently fitted
for the preservation of organic remains, while there has been no subse-
quent disturbance or metamorphism to account for their obliteration.

The Penganga shales and limestones are usually found resting directly
on the gneiss without anyjrepresentative of the lowest quartzose conglo-
meratic zone of the Pakhals. North-west of Edlabad, however, there are
said to be sandstones which appear to dip under the limestone, and a
quartzite is recorded as occupying a similar position, inthe hills north
of Aksapur.?

These Pengangd beds were regarded by the earlier observers, and
have always been referred to in the Survey publications, as of Vindhyan
age. Further on we will return to the discussion of the validity of this cor-
relation, but in the meanwhile it may be noticed that the evidence in favour
of identifying the Pakhal and Pengangd beds with the Cuddapahs is as
strong as it can be in the case of unfossiliferous rocks, where there is an
absence of absolute continuity of outcrop. They were unhesitatingly
identified by Dr. King, who examined both areas. The general lithologi-
cal resemblance is described as very close, and, though the particular sub-
divisions cannot be recognised in the different areas, the Pakhal beds were
regarded as answering to the Kistna and Nallamalai groups. They
exhibit much the same degree of induration and a similarity in their rela-
tion to the gneiss, and to the rocks of the transition period, and the case
is much strengthened by the occurrence of a series of small outliers, in the
space intervening between the northern limit of the main Cuddapah area
and the southern extremity of the Pakhal outcrop.

1 W, T. Blanford. MS. Report, 1866. | 3 Memoirs, XVIII, 212, (1881).
2 Memoirs, XV IIL, 224, (1881). \
Chap. IV.] SULLAVAI SERIES. 91

In the degree and nature of the disturbance they have been subjected
to, the Pakhal beds resemble the Cuddapahs. Lying at low and gentle
dips throughout the western portion of the area they occupy, they are com-
pressed and folded at high angles about their eastern limit in the neigh-
bourhood of the Singareni coal-field.

The Cuddapah beds have been recognised in Bastér territory, where the
Indrawati river, at Chitarkot, falls over quartzites that rest horizontally on
the gneiss. They are overlaid by limestones and red shales, which over-

lap on to the gneiss and are overlaid by a yet higher series of quartzite
sandstones. The country has not been examined in detail, and it is not yet

certain whether the latter belong to the Cuddapahs or to the overlying
unconformable Sullavai series.?

Further north Dr, Ball recognised the same beds on the plateau
south of Tarnot,? and they spread out and occupy a large area of the
Mahdnadif valley in Chhatisgarh, where they have been regarded by
observers working from the north as Vindhyan. They are described as
consisting of a lower group, composed principally of quartzitic sandstones
at times pebbly or even conglomeratic at the base, and an upper group
of limestones and shales.3 The shales are nearly always of a red purple
colour, very rarely green or dirty grey, the limestones are fawn-coloured,
grey or even black, sometimes pink or pale reddish purple, thick-bedded,
compact, splintery or shaly in composition, graduating into shale, often
seamed with chert bands. These beds, which were recognised in 1866
by Dr. Blanford as very similar to the Pengangd beds, occupy the
centre of the basin and lie with easy rolling dips, the underlying sand-
stone band forming a margin round the outcrop, where it is often turned
up at high angles. ~

The evidence for classing these with the Cuddapahs is not at present as
good as inthe case of the Pakhals, seeing that the intervening ground has
been only cursorily visited, but the general resemblance is close, and cer-
tainly much closer than to the rocks of the overlying Sullavai series, which
must now be noticed.

Besides the beds of Cuddapah age Dr. King recognised, in the Goddvari
valley, a series, unconformable to the Pakhals, which he called the Sullavai
series. They consist typically of a massive quartzite sandstone and con-
glomerate with a few slaty beds, overlaid by generally salmon or chocolate
coloured sandstones, and capped by a strong series of thin and thick-
bedded, very pebbly and gravelly, quartzites or indurated sandstones, which
weather in the peculiar and picturesque manner characteristic of the pin-
nacled quartzites of the Karndl series in its typical area.

1 Memoirs, XVIII, 224, (1881). | ® Records, XVIII, 173, (1885).
2 Records, X, 174, (1877). * Memoirs, XVIII, 227, (1881).
92 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [Chan, IV,

The thickness of the series where best developed in the Dewalmari
hills and in the western outcrop near Kdpra are stated to be—?

Kapra, Dewalmari hills,
3, KApra quartzites and conglomerates. 100 feet. ' Joo feet.
2, Venkatput salmon and chocolate beds . 200 ,, 300 4,
1. Enchardm quartzites . . . 7 go00 ,, 600 ,,

The unconformity of the Sullavais on the Pakhals is indicated by both
overstep and overlap, yet on the whole there is a remarkable parallelism
of dip between the two near Sullavai itself. Some sections are, however,
recorded where the Sullavai sandstones rest almost horizontally on the
eroded edges of the nearly vertical Paknal slates and quartzites.

These Sullavai beds were recognised in the hills near Dewalmari and
the observation is important, as the sandstone of these hills had already *
been identified with that of the exposures, regarded as of Vindhyan age,
east of the Wardha valley coal-field. According to Dr Blanford these are
white and purplish quartzite sandstones, breaking with a distinct conchoi-
dal fracture, and, in the great exposure extending from Chimir to Mul,
they are associated with a more or less felspathic coarse grit, which
decomposes into a very soft rock, easily mistaken for Damuda sandstone.

There is a much wider and more distinct barrier between the great nor-
thern Vindhyan basin and the Chhatisgarh, or upper Mahanadf area, than
between the latter and any of the affiliated rocks to the south, The ridge of
gneiss which, to the west, forms the well-raised base of the basaltic plateau
throughout the districts of Mandla, Seonf, Chhindwérd4, and Bettl, and to
the north-east forms the highlands of Chutid Ndgpur, is interrupted at this
point and the Gondwdna deposits stretch across from the Son to the
Mahdnadi valley. The watershed between the Son and ihe Mahdnadi
drainage is pretty high, and is occupied by Talchir rocks, probably
of no great thickness, so that the gneiss most probably forms a rock-
barrier from east to west, though of course it is open to question when this
was produced. It may well be of post-Gondwdnaage. To the north of this
barrier of gneiss the Cuddapahs and Karndls of the south appear to be
represented by a great series, principally composed of sandstones, long
known to geologists under the name of Vindhyan.

The name Vindhyan, one of the oldest introduced by the Geological
Survey, was used to designate the great sandstone formation of Bundels
khand and Malwa, and was adopted from the name currently applied by

1 Memoirs, XVI11, 231, (1881). | ? W. T. Blanford: MS, report, 1866,
Chap. IV.] VINDHYAN SYSTEM. 93

Anglo-Iudian geograpkers to the scarped range along the north side of the
Narbadd valley. The Vindhyan system ranks third in superficial extent
within the rock-area of the Peninsula, occupying in a single basin a larger
surface than the combined areas of any other formation except the gneiss
and the Deccan trap. The form of the basin is peculiar, There is a great
area, 250 miles long, between Chitor on the west and Sagar on the east,
and 225 miles broad from Indargarh on the north to Bdrwai (or Mortaka)
on the south, all presumably occupied by upper Vindhyans, although a
very large part of it is covered by the trap of the Mdlw4 plateau. From
Sagar a long arm, with a maximum width of 50 miles, stretches eastwards
for 340 miles to Sdsserdm in Behar. Another broader tract extends north-
wards from Sagar, and passes under the Gangetic alluvium between Agra
and Gwalior. ‘The gneissic mass of Bundelkhand lies between these pro-
longations. The exposed surface of the Vindhyan deposits is about 40,000
square miles and, with the area beneath the trap, the basin would occupy
about 65,00c.

Throughout the greater part of their border the Vindhyan sandstones
are unconformably related to transition or gneissic rocks, but in the
eastern branch of the area in Bundelkhand and the Son valley, and in the
neighbourhood of Chitor and Jhalra Patan, they rest, with little or no un-
conformity, upon deposits of very different character. These lower
beds were at first noticed under local names in the several areas, but the
convenience and fitness of having a common name for deposits so nearly
related was soon felt, and the term lower Vindhyan has been used in this
sense in spite of the very disproportionate importance of the two divisions
so established, and a doubt as to whether they are really members of the
same conformable system.

On the map prepared for this Manual it has been found necessary to
make one colour serve for the lower Vindhyans and the Cuddapahs, but as
the lower Vindhyans, in the sense here used, are confined to the margins
of the Vindhyan basin, this is not likely to lead to confusion. From Sdsse-
r4m, at the extreme east end of the area, the lower Vindbyans are con-
tinuous at the base of the Kdimur scarp for 240 miles, disappearing at
the Son-Narbadd watershed, where the upper Vindhyans sweep across
into contact with the transition rocks. The greatest width of the lower
Vindhyans across their outcrop in this their typical area is 16 miles, just
where the Son enters its main valley from the south. At some points
on the lower reaches of the river their outcrop is less than two miles
wide. Some small inliers, appearing through the alluvium in Behar ata
short distatice east and north of the termination of the Vindhyan plateau, are
most, if not all, of them of lower Vindhyan rocks, which also crop out
94 GEOLOGY CF INDIA—CUDDAPAH AND VINDHYAN. [(hap. IV

from beneath the upper Vindhyans in some of the valleys on the north side
of the plateau west of Sdsser4m., In this direction, however, the lower Vin-
dhyans soon disappear and at the lowest level, where the Ganges washes
the base of the plateau at Chandr (Chunar), only upper Vindhyans are
exposed, The concealment of the lower groups is probably only due to
depression in the main axis of the basin, for the very same rocks appear
again beneath the Kdimur sandstone as it rises towards the gneissic mass
ef Bundelkhand.

It may be considered certain that the Semri rocks, under the K4imur
scarp in south-eastern Bundelkhand, arethe same as the lower Vindhyans
of Son valley, but their appearance on the north is much more irregular in
évery way, a circumstance which is easily accounted for. From Chebu,
close to the Jumna, they are seen at intervals below the Vindhyan scarp
for 160° miles to beyond the Dhasd4n. The principal exposures are for 20
miles east of the Dhasdn, and for 12 west of the Ken (Cane). East of the
latter river the beds are totally concealed for long distances, where the upper
Vindhyans pass over them on to the gneiss, and the lower formation is
only visible in the gorges of the principalstreams. About Karwi again,
where the main scarp begins to trend eastwards, oblique to the general
strike of the basin, the lower Vindhyans are freely exposed, but at Bhita,
where the Jumna first touches the rocks of the plateau a few miles above
Allahabad, the upper Vindhyans are at the water level, the position being
more to the dip of the basin.

The third of the principal exposures of lower Vindhyans lies in the
extreme south-west corner of the main outcrop of the system and extends
from about ro miles north of Chitor ina southerly direction to the edge
of the Deccan trap, whence it turns eastwards and occupies a narrow
strip, with irregular boundaries, extending to about 25 miles south-east of
Jhalra Patan. Besides these three principal exposures there are some
small outliers of what are believed to be lower Vindbyans resting on the
disturbed Alwar quartzites on the eastern margin of the Ardvalli region,
south-west of Karauli.

The classification of the lower Vindhyan beds wants the definiteness
that is attainable in the upper Vindhyans. There are no well-marked zones
of sub-division, and all the members of the group are not to be found on
every section, the irregularity being partly due to thinning out and partly
to a lateral change of mineral character. The want of constancy is more
conspicuous in the lower members than in the upper, a direct result of the
mode of deposition. The first beds formed were deposited on an uneven
floor of the older rocks, and as the irregularities of this became smoothed
off, and the area of deposition enlarged by gradual subsidence, the con-
ditions of sedimentation became more uniform and gave rise to more uni-
form and constant ‘stratification. af the succeeding beds,
Chap. IV.] LOWER VINDHYANS. 65
The following sub-divisions have been recognised in the Son valley :
11. T.imestone. 5. Porcellanic shales.

to. Shales. i Rohtds group, 4. Trappoid beds,
g. Limestone. 3. Porcellanic shales.

8. Shales and sandstone. 2, Limestone.
7. Limestone. 1, Conglomeratic and calcareous sand-
6. Shaly sandstone. stone.

These lithological characters by no means indicate well-defined or con-
stant horizons in the series. They are all variable and pass into each other,
both vertically and horizontally, by interstratification and thinning out, or a
horizontal replacement of one form of sediment by another.

The lowest two groups, which, strange to say, appear to be in some
degree equivalent to each other, are only found in the Son valley, and
doubtfully in some outliers near Sdsseram. They exhibit great and capri-
cious variations of thickness, which can only be explained by their having
heen deposited on an uneven surface. The conglomeratic beds vary in
type from a coarse thick-bedded conglomerate, composed of slightly rolled
fragments of the underlying older rocks, 6 to 8 inches in diameter, to
a coarse sandstone, with pebbles of quartz. The limestone No. 2 is also
capricious in its distribution. It is in part a tolerably pure limestone, but
is for the most part hard and silicious.

The porcellanic and trappoid beds are almost sufficiently described by
their name, The porcellanic beds, mostly grey in colour, are very much
like the beds described under the same name in the Gwalior and Cuddapah
systems, The other beds interstratified with them were called trappoid
from their resemblance, in mineral constitution and mode of weathering,
to traps, but they are in fact composed of the debris of crystalline rocks
which has undergone a subsequent induration. Their distribution is in
accordance with their origin, as they are conspicuous where there is an
abundance of crystalline rock close to the boundary, and absent or very
slightly developed where slates are the chief rock exposed.

The divisions Nos. 6, 7, 8 form a sub-group of limestones, shales, and
sandstones with a band of limestone, thicker and more prominent than the
others, about its centre. Some of the sandstone is described as _lithologi-
cally similar to that of the Kdimur group, and owing to the dark, often
black, colour of the shales, they were once mined into in the hope of
obtaining coal.

The three uppermost members, Nos. 9, 10, 11, form another group
for which the name Rohtds has been suggested’, derived from the ancient
fort of Rohtdsgarh. Taken together, they are by far the most constant of
any of the groups of lower Vindhyans in the Son valley.

'F,R. Mallet, Memoirs, VU, 28, (1871). | 2 Manual, 1st ed., p. 78.
96 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [Chap. IV.

The beds of No. 6 are described as being almost universally ripple-
marked in the Mahanad{f tributary of the Son, besides which sun-cracks
and the marks of rain-drops are very common.

In Bundelkhand the Semri, or lower Vindhyan, beds were classified by
Mr. Medlicott as follows!:

5. Tirohdn limestone. | 3- Dulchipur sandstones,
4. Pulkoa schists. 2. Semri shales and limestone.

1. Semri sandstone.

These are not all represented on every section. In the easternmost
exposures only the Tirohdn (Tirhowan) limestone, and what was believed
to be a representative of the Semri sandstone, are seen. The Semri
shales die out near Shdhgarh and the Semri sandstone, thinning out about
the same place, can only be traced to the Dhasdn river. Coincident with
the decline in thickness of these two groups, the Dulchipur sandstones,
whose most’ easterly limit is near Chopra, increase in thickness and
importance and come into direct contact with the Semri sandstone by the
overlap, or thinning out, of the Semri shales. At the western end of the
exposure of the lower Vindhyans the Dulchipur sandstone is the only mem-
ber of the group represented.

Though none of the sub-divisions of the lower Vindhyans in Bundel-
khand can be identified with those in the Son valley, the general resem-
blance in lithological character, and more especially in their relations to
the upper Vindhyans, is such as to make their identity certain.

The two uppermost members very circumstantially represent the
Rohtds group of the Son—the thin, sharply bedded, fine grained limestone
of very variable composition, both in chemical and mechanical ingredients,
and the flaky silicious shales between which, even more capriciously than in
the Son area, occur the most complete vertical and horizontal transitions.
When the Bundelkhand ground was first described the equivalence of
these different rocks was not detected, and consequently it was supposed
that the shales had suffered denudation before the deposition of the lime-
stone, and the limestone again before the deposition of the Kdimur sand-
stone, which is found resting directly on both. In one form or the other, as
shale or limestone, this group is found from end to end of the outcrop,
being, like the Rohtds group, the only constant member of the series.

There is one character connected with the limestone in Bundelkhand
that does not occur in the Son region. It is almost constantly overlaid
by a silicious breccia, not detrital, but apparently composed of thin
layers of agate, chert, and jasper, shattered in places either by con-
cussivn or desiccation, and re-cemented by sintery or hyaline silica, free

> Memoirs Ul, 6, 1860),
Chap. IV.] LOWER VINDHYAN SERIES, 97

from sand or other detrital matter. This breccia, which is adherent to
the limestone and also fills cracks in its upper surface, is connected
rather with the Tirohdn limestone than with the overlying Kdimur
sandstone, which often has at its base a breccia conglomerate very differ-
ent in character from the Tirohdn breccia, This bed is sometimes 4o
feet thick, as on Panwdri hill, south-east of Tirohan.,

In the middle area, at and west of the Ken, the Semri sandstone and
the overlying shale and limestone band are well developed. The latter is
also fairly seen in the gorges of the Ranj and the Boghin, east of Panner,
but in the eastern area, about Karwi, the Tirohdn (Ruhtds) limestone,’
very free from its familiar shales, is with one exception the only member
of the series, The exception consists of a very peculiar bottom rock
covering the granitoid gneiss. Where found under the limestone this rock
might readily be referred to the Tirohan group, for it often has layers of
dense, fine limestone just like that rock, and is otherwise cherty, as is
often the case with the limestone, but it is largely a detrital rock com-
posed of quartz-sand, felspar-grains, and (characteristically) glauconite.
Cherty segregation in many forms,~ spongy, pisolitic, amygdaloidal or dis-
seminated,—gives a most peculiar aspect to the bed. This rock is trace-
able in the hills south-west of Karwi, the most north-westerly of which
about Akbarpur are altogether of metamorphic rock, and have a pointed
or rounded outline, the next have a thin cap of Kdimur sandstone,
but the sedimentary beds thicken steadily to the south-east, and at the
sacred hill of Chhattarkot the gneiss is only seen at the base on the north-
west side. At the high elevation of the junction there is only a remnant
of the cherty contact rock coating the gneiss under the Kdimur sandstone,
but in the Chhattarkot hill the contact rock occurs under the limestone, hold-
ing its position as a true bottom-rock. At a few places inthe eastern area
the flaggy sandstones of this band are well marked, as in the gullies to
the south-east of Chhattarkot hill, and they become more developed to the
west or north of Panna, on Bisramganj Ghat, where they are 50 feet thick.
In this way they are traceable into relation with the Semri sandstone, in
which also glauconite grains are of common occurrence.

This peculiar contact-rock of the east has been more specially noticed
because of a conjecture that it may possibly be an original nidus of the
diamond. A common form of it is a semi-vitreous sandstone, or pseudo
quartzite, of a greenish tinge, the result of the local solidification of
sandstone by diffused silica. Large pebbles of this rock are very abun-
dant in the conglomeratic diamond bed of the Rewd shales at the Panna
mines, and it is said they are broken up in the search for diamonds.! The
diamond-bearing beds of the upper Vindhyans are now at a much higher
level than any existing outcrop of the Semri beds, but it is very probable

-

- } Memoiss, WH, 71, (1869).
H
98 GEOLOGY OF INDIA~OLDER PALAOZOIC SYSTEMS. [Ohap, IV,

‘that this peculiar rock once extended over the ther elevated surface of the
gneissic area.

Both in the Son valley and in Bundelkhand there are indications that
the present limit of the lower Vindhyans is not very far from that of their
original extension. The irregularity of the lower groups in the Son valley,
together with the coarseness of texture of the lowest member of the group,
show that they were deposited on an uneven floor of deposition. The
thinning out of the subdivisions, except those forming the Rohtds group,
east of Bardhi, and the complete absence of Nos. 6, 7, and the sandstone of
No. 8, are evidently due to this area having escaped the sedimentation,
which went on elsewhere. Another observation of importance is that in
the outliers to the south the lowest conglomerate is much thicker than
in the main exposure, indicating an approach to the limits of deposition.

In Bundelkhand the original limitation of the lower Vindhyans is most
unmistakeably exhibited by the overlap of the Kd4imur group on to the
gneiss and Bijdwars. This is clearly enough seen on a large-scale geo-
logical map, where, in all the northern prominences and outliers, the upper
Vindbyans are in direct contact with the older rocks, while in the deep
eut valleys draining from the south the lower Vindhyans intervene. It is
confirmed by the record of sections, where the lower Vindhyans are seen
to be banked against a sloping surface of Bijdwars.

The lower Vindhyans of Chitor and Jhalra Pétan have not been so
fully described as those of the other two areas. They consist of shales,
limestone, and sandstone, the latter often conglomeratic near the base of
the series and sometimes containing boulders that range up to three feet
across, but do not appear to contain any of the volcanic or pseudo-
volcanic beds found in the Son valley. No unconformity with the upper
Vindhyans, or trace of one, is mentioned as occurring in this area, and the
beds do not exhibit that degree of compression which is seen further east,
At their western limit they rest unconformably on the gneiss and trans-
ition formations, but along the southern margin the boundary is formed
by the overlying Deccan trap. The lower Vindhyans are here exposed
ia an anticlinal, whose southern half is for the most part concealed, but
at Jhalra Patan and south of Rampurd there are outliers of the upper
Vindhyan sandstone, intervening between the lower Vindhyans and the
edge of the trap.}

The justice of classifying these beds with those next to be described ina
single system is open to question. There is most certainly an unconformity
between them and the so called upper Vindhyans. This might be inferred
from the complete overlap of the lower Vinilhyans by the Kdimur group
But there is better evidence than this in the very different facies of the two,

'C. A, Hacket, Records, X1V, 291, (1881). *
Ohap. IV.) RELATION OF LOWER TO UPPER VINDHYAN. 99

the lower being on the whole argillaceous and calcareous, and the upper
arenaceous and argillaceous, and more especially in the sudden and wide
spread change from the fine grained deposits of the Rohtds to the coarse
sandstone of the Kdimur group. Apart from this, the occurrence of debris
of lower Vindhyan beds 100 feet above the base of the Kdimur group
shows that the lower Vindhyans must have undergone some disturbance
and been then exposed to denudation.

Two deceptive features have, however, given grounds for exaggerating
the importance of the break between the upper and lower Vindhyans.
Before the equivalence of the upper shale and limestone of the Rohtds
group had been recognised, it had to be assumed that one or the other
had becn very extensively denuded before the deposition of the Kd4imur
beds, an assumption which involved much irregular superposition, although
none could be detected in actual sections. The other deception more
important, because it involves the introduction of disturbance uncon-
formity, is the apparent contortion of the lower Vindhyans before the
Kdimur period. This view rested upon the fact that the lower Vindhyars
are often found sharply twisted in close proximity to the perfectly undis.
turbed Kdimurs in the Son valley. The upper Vindhyans themselves
have, no doubt, undergone considerable flexure in this zone, as may be
seen in the Son area, on the west at Bilheri and on the east at the
Ghaggar. But these broad undulations were not at first thought sufficient
to include the frequent flexures seen in the lower rocks which, though
sharp, tever seem to carry the beds much out of an average horizon.
This opinion had, however, to give way to the fact of invariable com-
plete parallelism of the layers of the two formations whenever a contact
could be observed, even in proximity to those contortions. It is important
to dwell upon this observation, because some unconformities of this
class, reported and insisted on elsewhere, rest upon no other evidence
than that found to be fallacious in this case. It may even be suggested
that such appearances might possibly be produced independently of any
general disturbance of associated thick and thin, or hard or soft, deposits
merely by pressure from an adjoining elevated mass upon yielding un-
derlying beds, as occurs in the familiar case of the ‘creep’ in coal mines,

The classification of the strata composing the upper or true Vindhyans
is as follows :—?

Upper . . 13. Sandstone.
BHANDER (BUNDAIR) . 12, Shales (Sirbu).
Lower . , Jit. Sandstone.

1o. Limestone.
g. Shales (Ganurgarh).

1F,R. Mallet, Memoirs, V11, 50, (1871). | 2 Memoirs, Il, 56, (1860); VII, 27, (1871).
H 2
too GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [Chap. IV.

( Upper « 8 Sandstone.
Riwi « z : ( 7. Shales (Jhiri).
t lower +4 6. Sandstone.
( 5. Shales (Panna).
( Upper 4 Sandstone,
Kéimur Kvmore) « ¢ 3 Conglomerate.
(| Lower _¢ 2, Shales (Bijaigarh).
€ 1, Sandstone.

The general composition of the Vindhyan rocks is as uniform as their
general arrangement. Although chiefly made up of sandstones, which are
the coarser type of detrital deposits, the fineness of the rock throughout
the formation is remarkable. With the exception of the Kdimur conglo-
merate, which is constantly present as a bottom bed all round the bound-
ary in Bundelkhand, pebble beds are of rare oceurrence. The Kdimur
conglomerate is everywhere conspicuous through the prominence in it of
bright red jasper pebbles, presumably derived from the jasper bands so
abundant in the Gwalior formation. Where the Vindhyans rest upon the
Gwalior beds, the rock is rather a breccia than a conglomerate, the
included fragments being quite angular. The amount of this debris
throughout such a length of outcrop, to such a distance from the nearest
known area of Gwalior deposits, suggésts the extensive removal of these
peculiar rocks from the position now occupied by the gneiss.

There are general characteristics peculiar to each of the great sand-
stones. The Kdimur rock is fine grained, greyish, yellowish or reddish
white, sometimes speckled brown. False bedding is frequent and massive
beds are abundant, but on the whole the bedding is of moderate thickness,
sometimes flaggy and shaly. The Rewd4 sandstone is somewhat coarser,
and generally presents a mixture of massive strata and false-bedded flags,
The Bhander sandstone is softer than that of the lower bands, very tine
grained and generally distinguishable as of deep red with white specks,
or of pale tints with or without red streaks. The beds are generally
thinner, and not more than 6 to 18 inches in thickness, but massive
beds also occur, as is exemplified by the great monoliths cut from the
quarries at Ripbds near Bhartpur. Ripple marking is common through-
out the greater part of the Vindhyans, and occurs in great profusion and
variety in the upper Bhanders.

The different shale bands of the upper Vindhyans do not present any
constant distinctive characters. Thin, sharply bedded, flaggy, silicious or
sandy, sometimes micaceous shales, of greenish and rusty tints, form the
prevailing type throughout, Purely argillaceous shales are rare,

In the main Vindhyan basin diamonds are only known to occur in the
upper Vindhyans. Here, as elsewhere, the great majority of the diggings
are alluvial, but the principal workings, upon which most labour is
spent, are in a bed at the very base of the Rewd shales. Notwithstanding
Chap. IV.) UPPER VINDHYAN SERIES. ot

the immense range of this group, it is only known to be productive
within a small area of the Panna state, on the borders of the Bundelkhand
gneiss, and the surface diggings are confined to the same neighbourhood.
Here again, as already noticed of the Banaganpalli mines in Southern
India, the diamond layer is conglomeratic and the inference would seem
to be that the diamond occurs as a pebble with the others. The obser-
vation recorded’ that a particular kind of those pebbles at the Panna
mines is broken up and searched for diamonds, and that these particular
pebbles are derived from a peculiar bottom bed of the lower Vindhyan
series, would of course point to this latter rock as the original nidus of
the gem. But the observation in question needs confirmation.

The search for diamonds in Panna is not, however, confined to posi-
tions in which the gems could be derived from any existing outcrop of the
Rewd shales. There are numerous pits (all apparently surface diggings)
in the gorges and on the slope of the upper Rewd sandstone south of
Panna, and at a much higher elevation than any present outcrop of the
bottom shales or of the lower Vindhyans.

The Bhander limestone is the most variable rock of the series. Some-
times there is a considerable thickness, as much as 260 feet, of firm stone ;
elsewhere there is very much less, the carbonate of lime being apparentiy
disseminated amongst the calcareous shales associated with the limestone
and partly taking its place. The limestone is generally earthy and
compact, of grey, yellow or reddish tints, sometimes purer and either
compact ot crystalline. It was in this rock, at Nagode, that fossils were
thought to have been found long ago by Captain Franklin; they were sup-
posed to be Gryphza, and the rock was on this account assigned to the
lias, It is not known what became of the specimens, and repeated search
at the same locality has failed to verify the discovery. It is highly
probable the objects discovered were not organic at all, and quite certain
that the specific determination of them was fanciful.

The mutual relation of these sandstones, shales, and limestones is
most intimate throughout the upper Vindbyan series. The passage
upward, from shale into limestone, or into the great bands of sandstone, is
always more or less gradual, by interstratification, while the change into
shale at the top of the great sandstone beds is as generally abrupt. Both
the chief and minor subdivisions are wonderfully persistent over the
whole of the great basin, all being found in both the eastern and northera
areas into which the main area is divided by the Deccan trap. The lower
Bhander and lower Rewd sandstones are very attenuated in certain
directions, but there is an equivalent increase in the thickness of the
enclosing shales. In certain positions also the great bands of shales thin

1 Supra, 9. 97; Memoirs, 11, 71, (1860).
102 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [ Chap, IV,

out altogether, and the main sandstones coalesce. These reciprocal
variations in the distribution of the coarser and finer deposits have distinct
relation to position with reference to the border of the area, the shales
being in force towards the middle of the basin, and being replaced by
sandstones near the margin, showing that this border is approximately an
original limit, and that the actual basin corresponds pretty closely with
the basin of deposition. There are local exceptions to this condition, and
it is in the direction in which these exceptions occur, on the Ardvalli
side, that the only recognisable distant outliers of the upper Vindhyans
have been observed.

A formation so constituted, and for the most part but little affected by
disturbance, can result in but one form of surface. Accordingly the upper
Vindhyan area presents a three-fold plateau, each step formed of one of the
main groups, with minor plateaux, terraces or ledges corresponding to the
various subdivisions, The thick sandstones form vertical scarps over a
talus of the underlying shales. There is, moreover, a basin shaped lie of
the beds, apparently to a great extent original, whereby the surfaces are
rendered more or less concave, and the edges of the successive scarps of
sandstones scarcely higher than the outer one, composed of the Kéimur
rock, From this arrangement it follows that the upper group occupies by
far the larger part of the area, even the middle step of the plateau, the
edge of which is determined by the Rewd sandstone, being chiefly oceupied
by the lower Bhander shales.

Over almost the whole of the area occupied by the upper Vindhyan
beds they lie little disturbed and almost horizontal, and any violent effects
of disturbance are restricted to the south-south-east and the north-west
margins of the basin. Two local exceptions to this rule may be noticed.
In the Panwari ridge, south of Tirohdn, the Rohtds limestone is capped quite
horizontally by Kdimur sandstone. The hill is more or Iess detached from
the main plateau, and in the broken ground intervening, the sandstone is
found dislocated and dipping in the most irregular fashion, quite inex-
plicable by any ordinary mode of disturbance. The displacement is pro-
bably due to the underground solution and removal of the Rohtds limestone
and the consequent subsidence of the sandstone.

The other special instance of disturbance is not local in the same sense
as the last, as it is probably only a symptom of much more that is con-
cealed. It has been said that over the wide expanse of Vindhyan rocks
between Gwalior and Nimach, the Bhander and Rewa beds lie quite flatly,
and it has been presumed that to a considerable extent they stretch in this
manner under the trap of Mdlwd. Close to Jhalra Patan, however, at the
northern edge of the basaltic plateau, a sharp axis disturbance passes. from
the south-east, beneath the trap, to the north-west, throwing up the‘Via-
dhyan strata in an anticlinal flexure, with dips of 70° on each side. Along
Chap. TV, ] RELATION OF UPPER VINDHYANS TO THE ARAVALLIS. 103

this steep outcrop the standstone weathers into long narrow ridges. This
feature gradually dies out to the north-west. It is a hint that the disturb-
ance, which so violeatly affects the Vindhyans of the Dhar forest, extends
far to the north under the traps of Malwa.

The disturbance of the strata along the south-south-east border of the
Vindhyan basin, to as tar west as Hoshangdbdd, is plainly a recurrence,
on the same lines, of the compression which. had produced the contortion
and cleavage in the adjoining transition and gneissic rocks, It seems to
have taken different forms in different parts of the ground. Along the
whole Scn valley, there is little or no faulting in the zone of disturbance,
but at the Son-Narbadd watershed one or more faults occur at and close
to the boundary, the east-north-east strike being remarkably steady
throughout. Down the Narbadd valley towards Heshangdbdd, the dips in
the Vindhyans become unsteady. At Hoshangdbdd, and again in the
Dhar ferest; there is a decided predominance of a north-west, south-
easterly strike, and as the east-north-east strike remains constant in
the contiguous transition and metamorphic rocks, it may be inferred that
the former strike is the later of the two. Itis that to which the features of
the Vindhyans south of Nimach and at Jhalra Patan conform.

The north-westera boundary of the Vindhyans is in the main a fault of
great throw, along which the almost horizontal Bhander sandstone is brought
into contact with the highly disturbed Ardvalli beds. Beyond this
fault there are a few small, but important outliers, composed of the lower
members of the system. “The latgest of these occur south-west of Karauli,
where a natrow ridge of Alwar quartzites is faulted against undisturbed
sandstones of upper Vindhyan age on the south-east, and on the north-west
is overlaid by two alternations of sandstone and limestone, the lower being
regarded as lower Vindhyan (Rohtds group) and the upper as Kaimur.'
The beds have been compressed and are exposed in two narrow synclinals,
about 20 miles in length, but there ate small outliers of the lower beds to
the south-west of Nardoli, as far as the parallel of Ranthambhor.

From just north of Bindi, extending almost to Indargarh, a narrow
strip of Kdimur sandstone rests with little disturbance on the slates imme-
diatély west of the great boundary fault.

The throw of this fault must be at least 5,000 feet, and there is natural-
ly some difficulty in accounting for a single fault of so great a throw
having been formed subsequent to the deposition of the Vindhyans, and
among beds which have undergone so little subsequent disturbance as
they have. But we will find when treating of the Himdlayas that the
nature of the boundary between the Vindhyans and the disturbed Ardvalli

9 Records, XV, 288, (1884).
104 GEOLOGY OF INDIA—CUDDAPAH AND VINDHYAN. [ Chap. IV.

beds is very similar to what we may infer is the contact between the
undisturbed deposits of the Indo-Gangetic plain and the disturbed beds of
the Himdlayas. Moreover, along the foot of the Himalayas, there is a strip
of upper tertiary beds which have been disturbed, but to a less degree than
the older beds of the range, while the equivalents of these beds are believed
to occur under the alluvial plain, in perfectly conformable sequence with
the most recent alluvium. Now, if these suppositions are correct, as is
almost certainly the case, we can imagine that, after ages of denudation,
the upper tertiary rocks of the Siwdlik zone will be almost removed,
and the northern boundary, of what is now known as the Indo-Gangetic
alluvium, will theu exhibit very much the same features as the boundary of
the Vindhyans towards the Ardvalli range now does, The upper beds
will be in contact with highly disturbed rocks of much more ancient date
along a great line of fault. Beyond this will be a few outliers composed of
the lower beds of the series, the Siwdliks of the present classification, and
1o the north of these there will be a broad exposure of the wreck ofa
mountain range.

In the case of the Himdlayas the fault has been gradually formed pari
passu with the deposition of the Indo-Gangetic alluvium, which is contem-
poraneous in its origin with the principal elevation of the Himalayas and
formed of the debris of that range.

It is natural to suppose that the similar structure in the case of the
Ardvallis indicates a similarity of origin, and that the great Vindhyan
spread of Central India is formed of deposits which bore the same relation
to that range as the Indo-Gangetic alluvium does to the Himdlayas.

The suggestion is an important one since it would fix the period of the
formation of the Ardvalli range, or at any rate of its principal importance,
as contemporaneous with the deposition of the upper Vindhyan rocks that
were formed of its debris. It would account for the greater prevalence of
sandy beds near this margin of the deposit and would place the original
limit of deposition not very far beyond the present limit of the outcrops.

__ Allusion has been made? to some small outliers, believed to be of Vin-
dhyan age, which occur on the north and west Bundelkhand gneiss.
‘They differ much in character, and their peculiarities of composition may
help to explain their apparently anomalous position, Although the gneiss
reaches high up under the scarp of Kéimur conglomerate all round the
western border of the area which is described as a local edge of deposi-
tion, these small outliers occur at the level of the low country. If they
agreed in composition with the rocks of the main area, which are so strik-
ingly constant in this respect within that area, the fact might be at once
explained by a subsequent change of level, but such is not the case.

) Supra, p. 29.
Chap. IV. J VINDHYAN OUTLIERS IN BUNDELKHAND. 105

The most curious of these outliers form a very broken chain running
to south-south-east from close under the Pdr scarp at Ladera (7 miles east
of Antri) to Uchdr on the Sind river. Most of the exposures are quite
level with the plain, or only to be seen in the beds of streams. In a few
cases, as at Ladera and Pichor, they form narrow ridges up to 300 feet in
height. The rock is sandstone of the upper Vindhyan type, and at the
north end, close to the Par scarp, it contains large angular pieces of the
banded Mordr shales. Elsewhere it is quite free from coarse debris of any
kind. From many clear sections it is quite evident that these ribs of sand-
stone once filled a more or less continuous run of cracks, fissures, chasms,
or small valleys inthe gneiss. On both sides of the Pichor ridge the gneiss
reaches well up on the sides of the sandstone mass, with vertical surfaces
of contact. In the low ground, at the point of the ridges, and in the
small outliers, thin vein-like runs, of 3 feet wide and upwards, of the
sandstone are well seen, completely let into the gneiss, as it might be
filling an emptied trap-dyke, the rootlets of the wider chasm above.
Even in the larger masses no bedding is visible but sometimes, at
the edge of the mass, planes of pseudo-lamination and even ripple
marked surfaces occur parallel to the vertical wall of gneiss. The
lines of ripple were horizontal, and the steep face of the ripple turned
downwards in every case observed. Some of these features seem to
necessitate the supposition that the sandstone was let into this position
by disturbance, but all the other circumstances have suggested the explana-
tion given.

At Mahdrdjpur, 10 miles south of Antri and 14 miles east of the Vin-
dhyan scarp, there is a small group of hills, about three square miles in
extent, formed of fine sandstone overlying about fifly feet of flaggy shales,
both of Vindhyan type. The strata are greatly disturbed, but most irre-
gularly, as if compressed from every side. Although so much broken, the
rock is quite free from vein quartz, which is also a general character of the
Vindhyans as compared with the Gwalior strata.

The small hills of Sondr, 10 miles south-east of Narwar, and of Mohdr
16 miles farther in the same direction, present the same characters of
composition and disturbance as at Mahdrdjpur. At Mohdr a trace of the
Kdimur conglomerate occurs in the sandstone above the shales, which
cover a considerable area round the base of the hill, and may be looked
upon as ower Kdimur. .

A consideration of all the peculiar circumstances of these outliers
would seem to suggest that they may represent small local basins of the
upper Vindhyans. It seems that the process of denudation all round the
Vindhyan area has been to decompose, and remove the chemically-consti-
tuted metamorphic rocks which once formed high land around the sedi-
mentary basins, whether this relation were original or due to subsequent
106 GROLOGY OF INDIA—OLDER PALAOZOIC sysTEMS. [(Qhap, IV,

warping of the surface, leaving the softer but undecomposable detrital
rock to project where once had been depressions of the surface.

To the west of the Ardvalli range there are numerous exposures of
horizontal sandstones, mostly hard and usually more or less red in colour,
with occasional strings of pebbles.1_ They are frequently false bedded and
show ripple marks on the surfaces of the slabs. With these there occur
exposures of a compact grey cherty limestone which is believed to overlie
the sandstone, though no actual contact sections have beer. observed.

Wherever these sandstones rest on older rocks, whether of the “Maldni
series or the tilted beds of the Ardvallis, the relation is one of complete un-
conformity, usually accompanied by a band of strong conglomerate at the
junction. On the east of the Ardvallis a similar unconformity is to be
observed, but the western beds are even more markedly superficial with
respect to the older rocks, which must already have been disturbed and
elevated into a mountain range at the time of their formation.

The outcrops of these beds are scattered over a large area in western
Rdjputana, but they are for the most part individually small and surround.
ed by sandhills and alluvium. It is consequently very difficult to make out
the true thickness, or even the superposition, of the beds in the different
exposures, and it is possible that, in the neighbourhoed of Pokaran, beds
which have been coloured as Vindhyan are really much newer. The
sandstones are darker in colour, somewhat softer, and contain more pebbles
than near Jodhpur and are distinctly underlaid by a boulder bed consisting
of a fine matrix through which numerous large blocks, many of which show
distinct signs of glaciation, are scattered. There is no record of such a
formation having been observed at the base of the Vindhyans further east,
but a similat glacial boulder bed near Bap was regarded as Talchir. If the
two are identical the Pokaran sandstones would belong to the Gondwana
system. They differ, however, in type, from any of the sandstones of
western RAjputdna which can be referred to that period, and the glacial
beds differ from those of Bap in being interbedded with dioritic trap and
ash beds, and in the absence of boulders of the supposed Vindhyan
limestone, which are extremely abundant in the boulder bed of Bap.?

In degree of induration, lithological character, and relation to the tilted
beds of the Ardvallis, these sandstones and limestones closely resemble the
Vindhyans to the east, and they are regarded, with a strong show of proba-
bility, as of that age But, on the hypothesis of the relation between the
Aravalli range and the upper Vindhyan sandstones which has been pro-
posed above, there can never have been any continuity between the out-
crops on either side, and it is incorrect to speak of the Jodhpur beds as
outliers of the Vindhyan basin. They must have been formed contempora-

1 Records, XIV, 209, (1381). 1 2 Records, XXI, 32, (1888).
‘Chap. IV.) SOURCE OF THE DIAMOND. 107

neously in an independent basin of deposition and the connection between
the two would be similar to that which subsists between the recent deposits
north of the Himalayan range and the Indo-Gangetic alluvium.

The rock systems which have been described in this chapter are remark-
able as being the source to which all the diamonds found in India— if we
except the reputed occurrence of diamonds near Simla—can be traced,
though the actual workings are more commonly in recent stream gravels.!
In Southern India the principal source of the diamond is the Banagan-
palli group of the Karnul series near the town of that name, but they are
also obtained at other places within, or just outside, the boundary of the
Cuddapah basin. In the Mahdnadi valley they are found near Sambalpur,
just outside the boundary of the Cuddapah area, but have not been found
in situ as yet. The third diamond district of India is near Panna,where
the original home of the diamond is in the lower part of the upper
Vindhyan series, though the possibility of its derivation from lower Vindhyan
beds has been indicated.

This repeated occurrence of the diamond at, or about, the same geological
horizon might be held to be corroborative of the general resemblatice,
which has led the rocks in which it is found to be classed together. But
the evidence is of small value, for the diamond is in every case derivative,
and its original source has not yet been found,

About six years ago it was supposed that an original source of the
diamond had been discovered in a volcanic neck near Wajra Karur, filled
with a substance which closely resembled the blue clay of the Kimberley
diamond fields. A more detailed examination has shown that the Wajra
Karur rock is the product of decomposition of a basic volcanic rock, not a
peridotite, like the Kimberley blue clay.” Previous to this the rock had been
thoroughly prospected for diamonds, but none were found in the matrix
though diamonds have certainly been found on the surface after rain, and,
as there are no outliers of Cuddapah or Karnil conglomerates within the
drainage area, it was supposed that they must have been washed out of
the decomposed volcanic debris filling the neck.’

The conclusion is not a necessary. one, for the rock filling the neck
has been more largely removed, being more easily weathered than the sur-
rounding gneiss, and the neck now forms a depression in the general
surface of the country into which the diamonds might have been washed.
A different origin for the diamond has been suggested by M. Chaper,* who
believed that he had obtained diamonds, sapphires, and rubies from tne

1 Details of the known diamond localities will | ° Records, MIX, 130 (1886).
be found in Vol. III of this Manual, pp.1—5,}| ‘Sur une pegmatite diamantifére de l’Hin-
and Records, XXII, 39-49 (1889). dostan, Bull. Soc. Gécl. de France, 3rd series,
P. Lake, Records, XXIII, 72 (1890). XIV, 330, (1886); Records, XXH, 39, (1889).
108 GEOLOGY OF INDIA -OLDER PALAOZOIC SYSTEMS. [Qhap, IV,

debris of a pegmatite vein in the gneiss. He did not, however, find any of
these gems in their matrix, nor did he apparently himself procure them
from the debris. The evidence produced is of the slightest and has not
materially helped forward the solution of the problem of the original
source of the diamond.

It remains now to consider briefly the relations of the rock systems
described in this chapter to each other. In the first edition of this manual
the Cuddapah system was classed, along with the Gwaliors, among the
upper transition deposits, but subsequently the rocks of the Godavari
valley, which had previously been regarded as Vindhyan were identified
with the Cuddapahs of Madras, and Mr. Medlicott, abandoning his
previous opinion, accepted the Cuddapahs as the equivalent of the lower
Vindhyans.! Adopting this conclusion we have in the southern area
two unconformable series or systems, of which the uppermost is quite
unimportant compared with the lower. In the central area much the
same is the case, but in the northern area the two series arte said to be
nearly conformable and the lower is certainly much less in thickness
than the upper. The first, and most obvious impulse would be to class
the lower Vindhyans, the ~Pakhals, and the Cuddapahs together, as
an older system, and the upper Vindhyans, the Sullavais, and the Karnils,
as anewer. The classification would be to some extent in accordance
with the lithological resemblances of the rock series, and may be correct
in the main, but the truth must be less simple than this.

It is impossible to suppose that the lower Vindhyans of Central India
can, in any proper sense of the word, be the equivalents of the great
Cuddapah system, and if the account of the relation between the upper
and lower Vindhyans given above is correct, it indicates a difference from
that which subsists, between the Pakhal and Sullavai beds of the Godavari
valley, or between the Cuddapahs and Karnils of Madras. If the Karnil
series represents any portion of the Vindhyan system, whether upper
or lower, it is difficult to escape the conclusion that the commencement
of the Cuddapah epoch must date further back than the oldest of the
typical lower Vindhyan deposits, and may be in part contemporaneous
with the newest of those described among the transition systems, a supposi-
tion strengthened by the resemblances which may be observed between the
banded jaspers, hornstones, and porcellanic beds of the lower Cuddapahs,
and the similar rocks of the Gwalior series. The resemblances are, how-
ever, not such as amount to identity, and it is certainly more convenient
for purposes of description to class these beds with these newer rather
than those older than them,

1 Records, XV, 2, (1882),
CHAPTER V.

OLDER PAL-ZOZOIC SYSTEMS OF THE EXTRA-PENINSULAR
AREA.

The Salterange—Centtal Himdlayas—Unfossiliferous slates of the outer Himdlayagy—The
Jaunsar system—The Deoban limestone—The Baxa series—Eastern Tibet and Burma,

In the last chapter an older paleozoic age was accepted for the less
disturbed unfossiliferous rocks of the Peninsula, on the ground of the great
discordance that exists between them and the next succeeding beds, which
are known to be of upper palzozoic age. In the extra-peninsular area we
have no need to content ourselves with such indirect evidence as this, for
the presence of older palzozoic rocks has been proved, in three distinct
areas, by the discovery of cambrian and silurian fossils.

The most important, because most fully studied, of these is the Salt-
range, where an extensive series of conformable strata, nearly 3,000 feet
in thickness, is divided into the following groups :=—

Salt pseudomotph zone : : ‘ ‘ » 450 feet.
Magnesian sandstone . . ‘ ‘ ‘ + 250 45
Neobolus beds! . i a . ‘ : + 100 4
Purple sandstone . ‘ . ‘ . © 450 a
Salt marl? 7 . 4 . ‘ 5 + 1,500 4,

‘The complete sequence is only seen in the eastern part of the range, for
the series is unconformably overlaid by the succeeding one, and the groups
are successively thinned out to the west by denudation, as one after the
other becomes the uppermost to be seen.

The salt marl censists typically of a fine grained rock, varying from
dull purple to bright red in colour, composed of very fine grained clayey

1 This differs from that given by Wynne, | and a lower purple sandstone. They are only
(Memoirs, XIV, 6g) as the thicker sections of | seen at one place in the Khisor hills, and
his include part of the overlying carboniferous | there does not seem to be sufficient evidence
beds. for thia addition to the series, especially in

3 Dr, Waagen (Pal, Ind., series xiil, 1V, 44) | view of the doubts cast on the sedimentary
has added, below this, a grey gypsum group | origin of the gypsum (infra, p. 111).
110 GEOLOGY OF INDIA —EXTRA-PENINSULAR OLDER PALOZOIC. [Chap, V,

matter, mixed with some disseminated gypsum and carbonates of both lime
and magnesia.!. The marl itself never exhibits the slightest signs of strati-
fication,® nor are any coarser grains of sediment found mixed with it, but
it acquires an appearance of stratification from containing beds of gypsum
and rock salt, a few layers of dolomite, some beds of sandy dolomite in the
lower part of the series, and near Kheura a six inch band of bituminous
shale and some irregular patches of an obscure, dioritic-loo king, dark purple
trap, said to be associated with paler purple volcanic ash.

The most interesting and important of these, both as regards thick-
ness and economic value, are the salt and gypsum deposits. The former
appears to be somewhat irregularly developed and, except where mining
operations are carried on, difficult to observe, but it is described as oc.
curring in regular beds which exhibit distinct lamination. The greatest
development is in the Mayo mines at Kheura, where there are over 550
feet of salt. Half of this thickness, or 275 feet, is made up of five beds
of nearly pure salt, which is mined and placed upon the market without
being refined, the other half, known as kadar, is too earthy and impure
to be used in its natural state, and has, consequently, no marketable
value.

The impurities of the salt are principally sulphate of lime and chlorides
of magnesium and calcium. In the Mayo mine a lenticular band, with a
maximum thickness of 6 feet, has been found, composed of a mixture of
ylvine (chloride of potassium) and kieserite (sulphate of magnesium, with
one equivalent of water), which also prevailed through a thickness of about
seven feet below the sylvine band. Glauberite (anhydrous sulphate of soda
and lime) has also been found. Epsom salts (sulphate of magnesia, with
seven equivalents of water) commonly crystallises in the passages of the
mines on the surface of the salt marl and the £a/av, and, as this salt would
result from the absorption of water by kieserite, it appears to indicate
that the magnesian salt is of common occurrence in the rock.

Besides being disseminated through the red marl, the gypsum occurs
abundantly in beds and irregular masses, overlying the salt as a rule, but
also occurring more doubtfully in lower situations. It is found everywhere
accompanying the red marl and, at Mari and Kdlabdgh on the Indus as well
as in the Khisor range, contains more or less numerous bipyramidal crys-
tals of quartz.® Sometimes layers of hard flaggy dolomite are found in
thick masses of the gypsum and at one or two places numerous large and
perfect casts of hopper shaped crystals of salt are found in the dolomite

layers.

1 Memoirs, XIV, 70, (1878). 5 A. B. Wynne, Memoirs, XIV, 73, (1878) ;
2C. S. Middlemiss, Records, XXIV, 28, | XVII, 239, (1880); T. H. Holland, Records,
1891). XXIV, 231, (1891).
Chap. V.] SALT MARL OF THE SALT RANGE, 111

Till quite recently there would have been little hesitation in regarding
this group of beds as sedimentary in their origin, and older than the sand-«
stones which overlie them on most sections. It is true there were always
difficulties in the way of this interpretation, among which were the absence
of stratification in the marl, the enormous masses of sea water whose eva-
poration wouid be involved in the formation of the great beds of salt, the
great age of the beds containing so soluble a mineral and, more especially,
the fact that much of the gypsum was known to be not fully hydrated,
and Mr. Middlemiss'! has recently declared himself in favour of the
theory, broached by more than one of the earlier observers,-that the
marl, with its gypsum and salt, is hypogene, and has come into its pre-
sent position by a process analogous to that of igneous intrusion,

The arguments in favour of this conclusion are two-fold and derived,
firstly, from the abnormal stratigraphical position in which the marl is some-
times found and, secondly, from the mineralogical character of the marl
and its peculiarities of structure and included minerals.

As regards the position of the salt marl, it is ordinarily below the
purple sandstone and the form of the boundary, imbaying up the deep
valleys, is such as to suggest that it is in normal infraposition, but,
according to Mr. Middlemiss, there is no transition between the two
beds, and in the sections quoted by him there are numerous fragments of
the sandstone scattered through the upper layer of the marl. Besides this,
the salt marl is found at higher horizons intruded along the cores of
flexures and along fold faults. None of these observations are, however,
of vital importance, as the features might easily be the result of pres-
sure acting on a soft material like the marl.

The second argument is derived from the condition of certain included
fragments of dolomite, These are described as pitted, corroded and honey-
combed, and showing, in one section at Kavhad, an ultimate passage into
the red marl. In connexion with this, another section, two miles north
of Burikhel, must be noticed, where the salt marl is immediately
overlaid by the glacial boulder bed forming the base of the upper
palzozoic system of the Salt range. The marl here contains numerous
large aggregates of a hard compact dolomite, whose peripheral portions
are pitted or honeycombed and the cavities filled by gypsum, only the
central portions being unaltered. Jn the boulder bed no trace of the marl
could be found, whether as included fragments or by a colouration of the
matrix, and none of the numerous dolomite pebbles,. resembling those
of the central portions of the lumps in the subjacent marl, exhibit any
pitting or corrosion. From this it was concluded that the corrosion of the
dolomite in the marl was of later date than the boulder beds. Here the

' Records, XXIV, 26, (1891).
112 GEOLOGY OF INDIA—EXTRA-PENINSULAR OLDER PAL £OZOIC. (Chap. V.

doubt arises whether the dolomite fragments in the boulder bed might
not have been derived from the magnesian sandstone group, rather than
from the dolomite in the marl, unless indeed it is claimed that the marl has
here replaced the dolomites of the magnesian sandstone group.

It will be seen that the evidence is not conclusive, but taken asa
whole it throws a considerable doubt on the hypothesis that the marl
in its present condition was formed superficially, and this doubt is streng-
thened by the manner in which inliers occur among the upper tertiary
sandstones, in situations where its presence is difficult to account for, if
it is of cambrian age, 7

In Kohat there is another great development of salt which has been
regarded as of tertiary age, and it would be natural to regard this, with its
associated gypsum and marl, as the equivalent of that in the Salt range.
It is said, however, that the Kohat salt exhibits a difference of colour so
constant and characteristic that it is always possible to say with certainty
from which area the salt came. Too much weight must not, consequently,
be attached to the other resemblances, but it would certainly be strange
if two separate developments of salt, each on so large a scale, had taken
place in closely adjoining areas at such widely different eae as the
lower cambrian and tertiary.

The discussion of the origin of the salt marl has been nee forward
greatly by a recent investigation by Mr. Holland,! who has shown that the
associated gypsum was certainly formed by the sion of water with anhy-
drite. Not only are the inclusions in the quartz crystals all anhydrite, but
the matrix proves to be a mixture of anhydrite and gypsum in varying
proportions according to the degrees of hydration it has undergone.
Anhydrite could not have been formed at the surface or at a lower temper-
ature than 125°C, (257° F)?, nor could the quartz crystals in it have been
exposed to a red heat, as at that temperature they would have been
attacked by the sulphate of lime. He suggests that the anhydrite is due
to the action of acid vapours on a pre-existing limestone or dolomite, and
in this we probably see the true explanation of the marl. It is not a
hypogene rock intrusive in its present position, nor is it a sedimentary
rock formed superficially as such with its associated gypsum and salt, but
is due to the alteration of pre-existing sediments, whose exact composition
is unknown, by the subterraneous action of acid vapours or solutions.3

The group next succeeding the salt marl is composed of even-grained
sandstones, of a dull purple colour, containing carbonates of lime and mag-
nesia. The lower 50 feet to 100 feet are more argillaceous and may
indicate a transition to the beds below, but with this exception bands of

! Records, XXIV, 231, (1891) ; XXV, 54, one “Seyler Poggendorf. Annalen,

(1892), CXXVII, 161, (1866).
§ Records, XX\V, 243, (1891).
Chap. V.] CAMBRIAN FOSSILS OF TIE SALT RANGE. 113

clay are rare or absent throughout the group. At the upper limit the
colour of the sandstones becomes paler and some buff bands are seen.
The sandstones show ripple marks but rarely, and the only trace of life
they contain are a few, almost equally rare, obscure fucoid markings,

The purple sandstone group is succeeded abruptly, but conformably, by
dark or blackish shales, associated with sandy and calcareous beds
only some 20 to 100 feet in thickness, but extremely important on account
of the fossils they have yielded. These have come from two horizons, one
near the base of the group froma shale band containing Neodolus, the
brachiopod which has given its name to the group, and the other close
to the top, immediately below the magnesian sandstone.

The fossils from the lower horizon comprise the following species? :—

Discinolepis gvanulata, Lakhmina linguloides.
Schizopholts rugosa. 55 squama,
Neobolus warthi. — Lingula kiurensis,

» wylnes. 7 wartht.

Fenestellu, sp.

The fossils collected by Mr. Middlemiss from the higher horizon have
not yet been examined, but some fossils collected by Dr. Warth, which
probably came from the same horizon, include*—

Conocephalites wartht, Ayolithes wynnet.
Olenus indicus. 3 kussakensis.
Orthis wartht.

Besides which a trilobite belonging to the genus O/ene//us has been found
by Mr. Middlemiss.

None of the species from this group have been found in other parts of
the world as yet, but Conocephalttes warthi is said to be very close to
Solenopleura cristata from the Paradoxides beds of Sweden. The general
facies of the fauna, however, leaves no room for doubt that the beds are
of cambrian age, and, consequently, the oldest in India whose age can be
determined with any approach to certainty.

The fossiliferous beds are succeeded by the magnesian sandstone
group, which forms conspicuous scarps along the eastern part of the range.
It consists of from to to 300 feet of hard, light cream coloured or white,
dolomitic sandstone or sandy dolomite, associated with which are beds
of light coloured sardstones, and sometimes oolitic or flaggy bands,
the latter cccasionally covered with. fucoid markings and separated by
greenish or dark coloured shales.

The topmost member of the system, as developed in the Salt range, is
composed of thin bedded and flaggy sandstones, with in terbedded shales or

' Pal. Indica, series xiii, lV, 92, (1891). | 2? Pal. Indica, series xiii, 1V, 104, (1891).
fs t
114 GEOLOGY OF INDIA—EXTRA-PENINSULAR OLDER PAL#OZOIC. [Chap, V.

clays, characterised by their bright red colour and by the numerous pseudo-
morphic casts of salt crystals which cover the bedding surfaces. These
casts are found indenting the lower surfaces of the separate flags, and there
can be no doubt that they are due to salt having crystallised out, by the eva-
poration of salt water, and having then been covered by fresh sediment,
which, adapting itself to the angularities of the crystals, preserved their
form in the casts now seen. As might be expected, these beds are unfos-
siliferous, the only organic remains found being obscure fucoid impressions
and worm tracks.

In the Central Himdlayas of Hundes and Spiti a system of conformable
strata, of older palzozoic age, was described by the late Dr. Stoliczka! as
the Babeh and Muth series. Mr. Griesbach, as the result of a more extend-
ed survey, has recently rearranged these beds as follows ? :—

f 8. White quartzite, with limestone.
* 07. Red crinoid limestone.
Devonian (?) ‘ - 6. Dark coral limestone.
5. Flesh-coloured and brown quartzites and shales.
4. Coral limestone.
3. Red quartz shales and slates.
2. Shales and silky phyllites, with a great thickness
Haimanta system 3 of quartzites.

1. Quartzite, generally purple and a great thickness
of conglomerate. ;

Carboniferous

Siulrian

The whole of fhis sequence of rock groups is described as perfectly
conformable throughout and separated by a slight unconformity from the
beds above ; it will therefore be treated as a single system for descriptive
purposes,

The three lowest groups are united by Mr. Griesbach under the name
of the Haimanta (snow clad), a convenient designation for the great thick-
ness of beds, intermediate between the silurian and the gneiss and schists
which are either entirely unfossiliferous or have only yielded organic
remains too badly preserved to be identifiable.

The Haimanta series is divided into three groups, the lowest of which
is characterised by a great develupment of conglomerates, composed of
rolled and subangular fragments, among which quartz and gneiss pre-
dominate. The matrix is a hard, often deep purple coloured, sometimes
partially schistose quaitz rock. The junction with the underlying schists
is said to be transitional and often not determinable with accuracy, but the
prevalence of coarse conglomerates points to an unconformity and the
apparent transition may be due to the compression and metamorphism
the beds have undergone.

' Memoirs, V, 17-24, (1865). | ? Memoirs, XXIII, 49-€4, (1891).
Chap. Vj HAIMANTA SERIES OF THE CENTRAL HIMALAYAS, Its

The conglomerate group is overlaid by a great thickness of greenish
grey phyllites, slates and thicker bedded quartzites, traversed by quartz
veins, towards the upper part of which reddish brown and pinkish quartz
shales are intercalated. They are said to resemble, lithologically, the
Simla slates and contain some very obscure fossil remains which have
been referred to Bellerophon, with some Crinoid stems and casts of
bivalves,

The uppermost member of the Haimantas is described as very constant
and conspicuous and, consequently, very useful horizon in unravelling
the structure of the hills. It is described as consisting of bright red and
pink quartz shales passing upwards into greenish grey shales and quartz-
ites with pink and shaly partings, Some thin bands of deep red lime-
stone are occasionally found in the lower part of the group,

The total thickness of the Haimantas is about 3,020 to 4,000 feet, of
which from 250 to 500 belong to the upper group.

The passage from the uppermost Haimanta be Is to the lowest, classed
as silurian, is gradual, and the two types of rock are interbedded at the
junction. The lower silurian beds, according to Mr. Griesbach, consist at
the base of dirty coloured greyish pink quartzite, with calcareous partings,
passing upwards into grey shaly quartzites, alternating with dark blue to
black coral limestone,—limestone being the prevailing rock of the group.

The lower beds of silurian age are comparatively thin, being only about
300 feet thick, and are succeeded by from 1,000 to 1,200 feet in the
Niti sections, and much more in Spiti, of dirty pink to flesh coloured quart-
zites, with greyish green intercalated shales, and some limestone bands in
tne lower part. In Spiti Dr. Stoliczka observed some contemporaneous
trap in these beds,’ but none was seen in the Hundes sections.?

The upper Silurian quartzites pass gradually into dark grey or black
limestone, generally concretionary, showing sections of Corals and Bra-
chtopoda in large numbers, on the weathered surface. These again pass
gradually into bluish grey and brownish red limestones, containing frag-
ments of Crinozds.

The uppermost member of the system is composed of white quartzite,
ranging from 350 feet to 800 feet in thickness It was classed by Dr.
Stoliczka with his Muth series of doubtfully silurian age,? but according to
Mr. Griesbach, it is overlaid in Spiti with partial interstratification, by
limestones of carboniferous age containing Athyris royssi, Productus,
etc.*

1 Memoirs, V, 20, (1865). worked out and its relations fully determined,
2 A large number of silurian fossils from | there does not appear to be any benefit in print-
General Strachey’s collections were described | ing a nominal list of the species that have been
by Messrs. Salter and Blanford in 1865, and the | described. None of the fossils found in Spiti
collections made by Griesbach, which are | by Dr. Stoliczka were specifically determinable,

stillin course of description, will doubtlessadd | % Memoirs,V, 23, (1865).
to the number. Until this fauna has been! 4 Memoirs, XXIII, 63, (1891).

12
116 GEOLOGY OF INDIAmEXTRA-PENINSULAR OLDER PALOZOIC. (Chap. V.

On the accompanying map a very large area, in Kashmir and the
north-west Himdlayas south-west of the snowy range, has been coloured
as silurian, although a very considerable, if not the greater, portion of this
area is probably occupied by post-silurian beds. This course has been
adopted owing to the impossibility of distinguishing generally between the
older and younger slates, where all have been equally disturbed and are,
as a rule, equally unfossiliferous. In Kashmir the whole area mapped by
Mr. Lydekker as belonging to his Panjdl system has been coloured as
silurian, though it has since been shown? that a considerable portion of
this system is probably of upper palzeozoic age, and the same colour has
been used in the Hazdra district for that series of slates with limestones
and quartzites which are known as the Attock slates.

The Attock slates, which derive their name from being particularly well
seen in the hills along the Indus south of Attock, consist of dark coloured
slates and limestones, some sandstones and trap, both intrusive and inter-
bedded. Nothing is known of their age except that, inthe Sirban moun-
tain near Abbottabad, they underlie, with a strongly marked unconformity,
beds which are older than the tiias.? They must consequently be, in part at
least, of older palaozoic age. On the other hand, it is certain that beds of
secondary age are folded up amongst them, for limestone containing
Dicerocardium has been found east of the road between Khdnpur and
Haripur, and Ammonites east-south-east of Haveliyan® ; quite lately, too,
some fossils of decidedly cretaceous appearance were found by Mr. Gries-
bach in the area mapped as Attock slates. It becomes obvious from this
that the area coloured as silurian contains both upper palzeozoic and mezos
zoic rocks, but, as it is impossible to separate these at present, it has been
thought best to give one uniform colour on the map to the area occupied
by the generally unfossiliferous, slaty series, except where sufficient in-
formation is available to justify the separation of a distinct area of post-
silurian rocks,

In Kashmir the Panjal system of Mr. Lydekker includes, besides slates
and associated beds of transition and older paleozoic age, a great series
of beds which there is good reason to believe are carboniferous and permi-
an. These form the principal portion of the typical sections of the Panjal
system, and as they will be dealt with further on,* there remains ‘nothing
definite to be said of the lower beds in this place. The Panjdl system
was regarded as perfectly conformable to the underlying gneisses, but it is
probable that unconformable breaks exist, which were not recognised in
the course of a rapid exploration.

Further to the south-east a system of strata has been distinguished,

1 Records, XX, 130, (1888) ; infra, p. 13%, § Records, XII, 121, (1878).
2 Memoirs, 1X, 335, (1872). 4 Infra, p. 134.
Chap. V.] THE JAUNSAR SERIES, 117

under the name of Jaunsar, which exhibits considerable resemblance to the
Haimantas and silurian of the central Himdlayas. The lowest beds, known
to belong to this system, are characterised by the prevalence of purple
quartzites with intercalated red slaty beds and, low down in them, a
boulder bed, composed of large fragments of quartzite, dispersed through
a fine pralied matrix,

In the upper Pabar valley the quartzites are found lying unconform-
ably on the gneiss. In eastern Sirmur they appear to be underlaid by a
great thickness of grey slate with a band about 300 feet thick of grey lime-
stone, but the true position of these beds has not been established with
certainty. In the Bangal valley of eastern Sirmur the purple quartzites are
overlaid by about 200 feet of a dark grey felsitic trap, covered by as much
more of mixed trap and ashes, which are of subaerial origin. In north2rn
Jaunsar these are replaced by a great thickness of slaty beds, mixed with
bedded lava flows and impure volcanic ashes, and a band of limestone, some
300 feet thick. Above the volcanic beds comes a great thickness of sub-
schistose slates.?

The purple quartzite beds exhibit considerable lithological resemblance
to those of the Babeh series (Haimanta) and the volcanic beds may well be
contemporaneous with those of silurian age mentioned by Dr. Stoliczka.

South of the snowy range the Jaunsar system has not been recognised
with certainty away from Jaunsar and eastern Sirmur, but, low down in the
Simla slates, as the beds which underlie the Blaini group in the Simla
district have been called, there are some purple quartzites which may
belong to the system, while there can be little doubt that the purple
quartzites and volcanic breccias described by Mr. Middlemiss in western
Garhwal* are of the same age.

+

The Jaunsar system is unconformably overlaid by a great limestone
series, which forms, and derives its name from, the peak of Deoban, north
of Chakrdta. It consists of a ‘pale to dark grey limestone, often more
or less dolomitic, with interbedded grey shales. The limestone is frequently
mephitic, in places contains numerous cherty concretions, and is occa-
sionally oolitic.’

The age of this limestone series is unknown, it appears to be un-
conformable to the Jaunsar system and is certainly overlaid, with marked
unconformity, by beds which there is reason to believe are of carboniferous
or permianage. Outside Jaunsar it has not been identified with certainty,
but is probably represented by the great limestone formations seen south-
east of the Chor, in the Shdli mountain, in Kulu, andin Garhwdl and Ku-
mdun. The correlation, which is based on general lithological resem-

1 Records, XXI, 131, (1888). | ? Records, XX, 34, (1887).
3 Records, XXI1, 133, (1888),
118 GEOLOGY OF INDIA—EXTRA-PENINSULAR OLDER PAL&OZOIC (Chap. V.

blance and relative position in the series, derives great support from the
discovery, in the limestone of Kilu!, Jaunsar*, and Kuméan’, of curious
structures, exhibiting considerable resemblance to each other, which may
be merely concretionary but are quite possibly of organic origin.

In the eastern Himdlayas Mr, Mallet distinguished a series of beds
named after the hill fort of Baxa,* composed of variegated slates,
quartzites and dolomite, the latter being the most prominent member
of the series and having a thickness of 1,500 feet. The dolomite is
described’ as generally massive with obscure bedding, but frequently shaly
and passing at times into a dark grey slate. It is saccharoid, light grey,
rarely while in colour, with nests of more coarsely crystallised calcite and
drusy cavities, lined with the same mineral, scattered through it. There
is also a strong band of quartzites and quartz schists, and carbonaceous
slates have been observed. This series was regarded as younger than the
Damudas, but, as has already been explained,' it is probable that the section
was ministerpreted, and it should be looked on as older, probably consi-
derably older than the Damudas, and consequently of lower palzozoic
horizon.

Further to the east devonian fossils were found by the Abbé Des
Mazures near Gouchou in eastern Tibet,S but no other older palazozoic
rocks are known along the eastern frontier till near Mandalay, where
Dr. Noetling has recently faund a shaly limestone containing Crinords,
Orthoceras and a species of Echinospherites.7. The latter of these alone
would be sufficient to stamp the beds as silurian. These older palaozoic
beds are no doubt largely developed in the country east of the Irawadi
valley, but no details have as yet been published.

* Geol, Mag., 3rd. dec., V, 257, (1£88). 5 Supra, p. 75.
? Records, XXI, 333, (1888). ®§ Comptes Rendus, LVI, 878, (1864).
3 Four. As. S.c., Beng., IM, 628, (¥834). * Recores, XXMUI, 79, (1890).

4 Memoirs, XI, 33, (1874).
CHAPTER VI.

CARBONIFEROUS AND TRIASSIC ROCKS OF EXTRA-PENIN®=
SULAR INDIA.

The Salt range—Central Him4layas—Carbonaceous system of the outer Himdlayas—Kashmfr
—Hazdra—Afghdnistan—T ‘enasserim— Unfossiliferous slaty series in southern Afghanistan
—Arakan— Manipur and Négé hills.

The upper palzozoic rock groups, of peninsular and extra-peninsular
India alike, bring forcibly before us the impracticability of a rigid ap-
plication of the European divisions of the geological scale. Almost
everywhere the palaozoic rocks pass upwards, without an unconform-
able break of any importance, into beds of mesozoic age, and it will be
found convenient to class the upper palzozoic and trias together in the
description, as it has been found necessary todo onthe map. But with
this exception the carboniferous and triassic rocks of the two areas pre-
sent so strong a contrast that it is necessary to treat them separately, In
the Peninsula they are represented by a great system of subaerially
formed river deposits known as the Gondwana system ; in the extra-penin-
sular area the rocks of the same age are of marine origin, and as their age
can be more satisfactorily established, and they form an important link in
the chain of argument by which the age of the Gondwdna system is
determined, they will stand first for description.

The Salt range—but now in its western half—is again our typical
area, where the series is best exposed, most fossiliferous, and has been
most completely studied. It has there received a considerable amount
of attention from the Geological Survey, as well as from independent ob.
servers, and the classification now adopted is more complete than the
simple division, into speckled sandstone and carboniferous limestone,
originally adopted by Mr. Wynne.! The first of these names may, how-
ever, remain to distinguish one of the principal divisions, and for the

other the name Productus limestone, proposed by Dr. Waagen, will be
used, as the term carboniferous is misleading.

1 Memoirs, XIV, 69, (1878). The descrip- | account in Pal. Indica, series xiii, 1V, (1890)
tion of these rocks is based partly on Mr. 91). The paleontological data are all taken
Wynre’s memoir, and partly on Dr. Waagen's | from Dr. Waagen’s account, op. ctZ., Vol I.
120 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. (Chap. VI.

The oldzst member of this system was distingnished by Mr. Wynne as
the speckled sandstone. It rests uncenformably on the older palzozoic
rocks, and is usually, if not invariably, characterised at its base by a boulder
clay, formed of a fine grained matrix of shale or sandy shale, usually olive-
green in colour and weathering into acicular fragments, through which are
scattered blocks of hard rock, ranging to several cubic feet in size,
almost invariably subangular and frequently showing faces that have been
smoothed, polished and striated in the manner characteristic of glacier
action. These fragments are principally composed of slates, quartzites,
and crystalline recks, whose nearest analogues are to be found 750 miles
to the south, in the syenites and porphyritic felsite of the Maldni series, but
mixed with them are numerous fragments of the older palzozoic beds of the
Salt range, more especially of the magnesian sandstone. The glacial origin
of these beds is so obvious, and the resemblance to a boulder clay is so
close, that it is difficult now to understand how there should ever have
been any doubt regarding its mode of origin, were it not for the difficulty
of accounting for the presence of glaciers and floating ice at sea level,
in so low a latitude, at a period when, on @ prior? grounds, a milder

climate might have been expected to prevail than at the present day. The
difficulty is inereased by the apparent derivation of the included fragments

from the south, but it is part of a much larger problem which cannot be fully
dealt with as yet, and we must be satisfied with accepting the fact of the
existence of glacial conditions at sea level.

Many of the glaciated pebbles of this bed show a peculiarity whieh is
rare in other glacial boulder clays. Instead of being smoothed and striated
on one or two surfaces only, they often have a number of flat su:faces meet-
ing at obtuse angles and each showing a different direction of striation.!
It is in fact evident that, after one face had been ground and smoothed, they
had been slightly shifted so as to offer a fresh surface and, by a repetition
of this process, many of the. pebbles have acquired a facetted appearance,
as if they had been ground by a lapidary.

In the trans-Indus extension of the Salt range the boulder beds are under-
Jaid by grey shales, in which three species, Ayolithes ortentalis, H.
3. and Cardiomorpha
indica, have been
found,? but east of the
Indus a more extensive
and most interesting
fauna has been obtained

’ from the beds which im-
Fig. 6.—Conularia warthi, Waag., nat. size. mediately overlie the

   

  
 

ANY
CRE
Sab} dh)

   

LCERLLIR TL
SSS aang aioe

     

'A.B. Wynne, Geol, Mag., 3rd dec., Ill, | (1888).
731, (188¢); H. Warth, Records, XXI, 34, 2 Pal. Indieu, szries xiii, 1V, #14, (1891).
Chap. VI.J CARBONIFEROUS OF THE SALT RANGE. 121

boulder clay. The first indication of this was the discovery by Dr. Warth
in 1885, of Comuularia near Pid, but as his fossils were obtained from trans-
ported pebbles, their importance in fixing the age of the beds they were
derived from was disputed. Subsequent research has shown that, though the
nodules in which they are preserved are certainly transported pebbles, they
must have been of almost contemporaneous origin with the bed in which they
are now formed. Moreover, some fossils from a sandstone band above the
conglomerate indicate a precisely similar homotaxis to those in the pebbles,
and-in discussing the relations of the fauna we may take it as a whole,

The fossils obtained from the conglomerate band are given in the fol-
lowing list, where the letter 4 indicates that the species is alsu found in.
the marine beds below the coal measures of New South Wales! :—

A. Pleurotomaria nuda. A. Aviculopecten cf. limeformis.
Bucania warthi. A. Martiniopsis darwint.
A. Conularia (evigata. A. Chonetes cracowensis,
A. Conularia tenutstriata. Diseina, sp,
Cinularia warthi. Discinisa warthi,
A. Sanguinolites cf. mitchelli. Serpulites undulatus.
a5 tentsont. ‘i wartht.
Nuzula, sp. tuba.

Pseszdomonotis subradialis.

From the associated sandstones the following four species were ob-
tained : 2—

A. Eurydesma_givbosum. A. Eurydesma cordatum.
A. 5s ellipticum. A. Ma@onia gracilis. -

The most noticeable point about this fauna is the very large proportion
of species, thirteen out of twenty-two, which are identical with forms
found in the marine carboniferous beds of New South Wales, a ‘proportion
which not only shows that the two are approximately contemporaneous in
origin, but that they must have been deposited in the same great marine
area, with free communication between the two localities. The Australian
beds were formerly regarded as equivalent to lower carboniferous of the
European sequence, but are now considered as upper carboniferous, if not
homotaxial with the permo-carboniferous of Europe.

The number of identical species would of itself be sufficient to place
the Salt range boulder beds on the same horizon as the marine carboni-
ferous beds of Australia, and it is a remarkable fact, whose importance
will appear hereafter, that these same marine carboniferous beds of New
South Wales have been found to contain large boulders of foreign rock,
exhibiting distinctly glacial smoothings and striations, imbedded in fine
grained silt, along with delicate Fenestellz and bivalves whose valves are
still united in the position in which they lived and died.3

} Pal, Indicu, series xiii, 1V, 60, 145, (1890-91). |? Pal. Indica, series xiii, IV, 147, (1891).
: 4 Records, XIX, 41, (1886).
GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS.

(Chap. VI.

(22

 

 

 

 

 

 

 

 

 

 

ol Op gi Eo Siecle aqe law it 9 seef evel pope * : . . uoj}s
-pues pajyoods saddq) 10 Jamo]
19M0'7] .
S | Cz} ox o1jbh |] ve] | BE fee FE fo yrpe (¥378y) UOISTAIp Jamo]
v j a
: Li] zg} 69 12)Se1 Sz! g {1 ler |bL [2 |b | a | °} © | (esata) votstarp arppry 2
§ : SIPPHAL a
a
3 ~~
< pe] Sx] 2S} CSizg tir} igs|s je | rl). + + (q8eq a
= ]2y) worstaip taddq =
3 5
& iol
4
(| gz] Zt| SS} zSigzi]| b | a1} ej 1 [€ lob |gt jee | C1] 1 | V | * (aeyspunyyy) suors %
“I4Ip 21PplA pus Jamo7y .
saddy,
€ | of) £ | 9S/06 or} 1} ig |€S jor |S | F t | (iqe{) spaq epodojeydan ft * aadd y,
dnoaZ napiy | | gb] "| Zt€9 JP x |} a [vin fo rf fot | & | | & | apy) spaq ysoudoy
w ., t| rleslallelalmalelola a v
PSRs scl2ziq ($|8 12 le |S 12/2 18.1218 |F
*SYOM SIHL Nl GALd0dV NOILVOIAISSVID, S| oy z 2 ? 3 g 7 e 3 z s 3 3 “NOIL' ssv19
2/3) 5) $i5 5/3 gis |2]e |° VOLUISSVIG SNIOVVAA "EC
a 5] 8] al" 5 el cana
g| 7] £| 3 : z
| 3| B\ 3 “| ¥

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘aouvd 9105 243 fo sataas snzonporg ay) fo S¢noas ay] ut spzissog fo uowngrassip ayz fo quawajojs 40]ng0 7
Chap. VI.] CARBONIFEROUS AND PERMIAN OF THE SALT RANGE, 123

Australia is not the only place where glacial boulder beds of upper
palaozoic age have been found, and we shall have frequent occasion to
refer to évidences of glacial action and will use such evidence for purposes
of correlation, when paleontological evidence is wanting, But as the real
importance of this horizon is in establishing the homotaxis of the Gond-.
wana system, the fuller consideration of the recorded observations in
other countries, and the discussion of the validity of this method of cor-
relation will be deferred till that rock system is dealt with.?

This group of glacial and fossiliferous beds is overlaid by about 400
feet of mostly light coloured, reddish or purplish sandstones, unfossi-
liferous except for some obscure fucoid markings and plant impressions,
known as the speckled sandstone group.* The sandstones are interbedded
with some red shaly bands, and lavender coloured and purple argillaceous
and gypseous bands, which are especially prominent near the top of the
group. Ripple marks and oblique lamination are frequent, and the
weathered surfaces of the sandstone is frequently studded with rounded
knobs due to a local concentration of the calcareous cement.

Above the speckled sandstone comes a great series of beds which
have long been known for the wealth of fossil remains they contain, and for
the presence in them of Ammonites associated with Brachiopods of
palzozoic type. The detailed classification of these beds is a matter of
some difficulty; the first column of the accompanying tabular statement
shows that adopted by Dr. Waagen, but if the distribution of the fossil
remains, as shown in the tabular statement, is examined, it will be seen
that this grouping ignores the two most prominent paleontological breaks,
that between the Katta and Virgal beds and the still more striking one
between the Jabi and Chidru beds, nor is it, so far as can be judged
from the published descriptions, what would be adopted on purely
lithological grounds. The most important modification required appears
to be the separation of the topmost, or Chidru, beds from the rest of
the series, a separation demanded not only by some slight indication of a
physical break and by the small proportion of species which are found in
lower groups, but more especially by the complete change in the type
of the fauna from one marked by the prevalence of Brachiopoda to one in
which their place has been taken by the Lamellibranchiata. Excluding
these beds, the other groups fall naturally into three divisions of two
groups each, as indicated by the brackets on the right hand side, a group-
ing that will be adopted in the following description.

The two groups of the lower Productus beds are more distinctly sepa-
rated than in the overlying divisions. The lower group consists mainly of

Vinfra, chap. VUE, | ? Memoirs, X1V, 90, (1878).
124 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. (Chap. VI.

soft sandstones, coaly near their base, full of fossils, contrasting in this
point strongly with the lithologically not very dissimilar beds of the speck-
led sandstone, in which no fossils have been found. The group becomes more
calcareous to the west, and in the trans-Indus extension of the Salt range
is said to consist of limestone. The upper group is not very sharply
defined from the lower, but is more calcareous ; where normally developed
it contains many beds of light coloured yellowish or whitish limestones,
intercalated with marly and sandy beds.

The first point to notice about the fauna of this group,! whose general
facies is decidedly upper carboniferous, is the entire absence of a single
species also found in the fossiliferous beds at the base of the speckled
sandstone, the only species presenting any degree of alliance being Bucanta
kattaensis, which might be a modified descendant of 2. warthz from the
lower beds. The change is complete,
and with it disappears all connection
with the Australian carboniferous fauna,
to be replaced by arelationship with
upper carboniferous and permian faunas
of Europe. Not counting allied species,
the lower group contains no less than
Fig. 7.- Productus semi-reticulatus, Mart, sixteen species identical with forms

Mower productue pede: found also in the carboniferous and
permian of Europe, or in that series of strata, intermediate between the
two, which the Russian geologists have distinguished under the name of
permo-carboniferous. The names and distribution of these species will
be easiest explained by the tabular statement below :—

 

 

 

 

 

 

 

 

Lower Upper’ Permo-

carboni- | carboni- | carboni- | Permian.

ferous. ferous. ferous.
Dielasma elongatum 4, i . : . isi sais * *
Athyris royssti . . . . . . * te *
Spiriferina cristata +, ‘ . i ' ie * x* %
Spirifer striatus . . . : 3 . * te as

4 marcout . . * *

»  moosakhailensis . ag * we

@latus . . . . " * * *
Martinia ci. glaira . . . . * iia sia
Orthis indica ° . . . % %

» pecosii, . . . . . . oC % sia
Reticularia lineata . . . . . * x x shy
Streptorhynchus pelargonatus ot % % *
Productus lineatus : . . . ‘ * * wee

5 cora. . . . ay % % eee
a semireticulatus: . . . . %* -%& x oe
Ks spiralis . . . . * x

 

 

——————————

1 Complete lists of the fossi's will be found in Pal. Jnd., series xiii, IV, 160, 182, (1890-91).
Chap. VI] PRODUCTUS BEDS OF THE SALT RANGE. 125

From this it will be seen that four of the five species which are found in
the mountain limestone range up into higher beds, and all of the five found
as high as the permian are also found in lower beds, while there are four-
teen species found in upper carboniferous and permo-carboniferous beds,
seven of which have never as yet been found in newer or older beds.
This fixes the homotaxis of these beds as upper carboniferous, or inter-
mediate between that and permian, a conclusion in accordance with the
general aspect of the fauna.

The upper group of the lower division only contains four species that
are found in the permo-carboniferous of the Ural mountains, of which only
one (Strophalosia horrescens) is not found in the lower group. It also
contains two species, Dielasma itattubense, and Spirifer marcoui, which
have also been found in South and North America respectively.

We have here evidence of a great change in the distribution of land
and water from the time when the lower beds of the speckled sandstone were
being deposited and the sea stretched continuously from the Salt range
to New South Wales. The sea no longer flowed over eastern Australia,
where fresh water sandstones and shales with beds of coal were being
formed, and the barrier which, at an earlier period, shut out the European
forms of life had been submerged, allowing the western fauna to in-
vade the Salt range permo-carboniferous sea.

The same sea appears to have extended eastwards into China, for, from
thin beds of limestone above the coal of Lo Ping, a fauna of upper carbon-
iferous type has been described, which contains eleven species also found
in the Salt range. Of these Reticularia lineata, Martinia glabra,
Orthis pecostt, Productus semireticulatus, Strophalosia horrescens, and

3 Richthofenta stnensis, are found in the lower
group of the lower Productus beds, while another
species, Productus compressus, is found in the
upper group of the same division. The other
five species are found only in the middle or
upper division.?

%

ep
Lo)

The middle division of the Productus beds
is the most important in thickness, and the
most conspicuous owing to its forming great
precipices in the western portion of the range.
It consists principally of compact grey lime-
stone, frequently dolomitic and full of fossils
which are usually difficult of extraction, the
very numerous specimens obtained having been
mostly derived from marly beds interstratified

 

Fig. 8.— Section of Richthofenia
laurenciana, lower and middle
Productus beds, .

1 Pal. Indica, series xiii, 1V, 163, (1891).
126 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [Chap. VI.

with the more compact limestone. Corals are common and, in places,
they are accumulated in great reefs in which no bedding can be traced.
The lower group of the middle division, which has a thickness of 200 to
300 feet, so far as can be gathered from the descriptions,is overlaid by from
10 to 50 feet of marly beds intercalated with thin limestone, which form
the upper group. Silica, which is always present in varying quantity in
the lower group, and sometimes forms large flinty concretions, is much
more prevalent in the upper group, and the fossils are beautifully sisicified,
especially at Musa Khel, and are generally very numerous.

The fauna of this division is a very extensive one, comprising 16g dis-
tinct species, and might e:sily be extended by systemntic collecting! Of
these twenty-two have been found in other parts of Asia and Europe
whose distribution in time is shown by the following tabular statement :—

 

Carboni- |Permo-cr- :
ferous, |boniferous. Permian.
Macrocheilus avellanoides  . 3 é 7 . x
Pseudomonotis gosforthensis . 7 3 - %
Cametophoria humbletonensis , ‘ : ‘ gs x
i superstes . ‘ ee . x
Spirigerella derbyi, . > x af Si
Athyris royssit . . . . x % ¥%
Spiriferina cristata . . 3 2 % x
Spirifer moosakhailensis . ‘ . ; % % -
Orthis cf. indica . ‘ ‘ . ‘ 5 al x
Productus lineatus x OT x
3 cora ‘ . . . : . . * * se
»  —aumboldti . . . . . * was *
5 abichi . . a . ‘i . sas %X
as tumidus ° : F . - é F . ie xX
Marginifi ra typica . . . . . : c se x i
Fenestella pervlegans . . : . . *% %
Synocladia virgulacea « 4 «4 «+ ¥
Thamniscus dubius A F < 7 %
Acanthocladia anceps . . . . x
Spirorbis helix a é a . ‘ is i x
Giinitzella columnaris . é . . . ‘ a x
Stenopura ovata. “i x . 5 o e erg %

 

 

 

 

 

 

 

It will be seen from the tabular statement that the fauna is distinctly
newer in type than that of the lower divisicn, and contains a larger propor-
tion of exclusively permian forms. It is also marked by the papencance
of certain species with decidedly mesozoic affinities, namely Nautilus
peregrinus, and Oxytoma atavum, which are allied to jurassic species,
Hemiptychina inflata has its nearest ally in the trias of Italy. Pecten is

1 A full list of fossils will be found in Pal, Indica, series xiii, IV, 60, 186, 19% (1891).
 

 

 

 

 

 

 

 

   
  
 

Productus subcostatus, Waag.

Lyttoma tenuis, Waa,

 

Atbyris royssii. Lev,
Lonsdaleia salinaria, Waag& Wentzel.

Spirigerella derbyi, Wang.

 

Camerophoria purdoni, Dav,

 

Spirifer moosakhailensis, Dav,

PERMO CARBONIFEROUS (Middle Productus Limestone) FOSSILS.

Calcutta Phototype Co,
 

Balizropnon joossiaaus, Kou A
Nenodiseza caroonarisa, Wang,

Productus gratiosvs, VWaag.

Bellerophon jonesianus, Kon.

=a

PERMIAN (Upper Productus Limestone) FOSSILS.

Entalis herculea, Kon,
Chap. VI] UPPER PRODUCTUS BEDS OF THE SALT RANGE. 127

seldom found below the trias, and the genus Oldhamina has its nearest
relation in the rhetic Pterophlotos. ‘These mesozoic forms preclude us
from assigning the group to an older datethan the permian, but the relation-
ships are not close or extensive enough to justify assuming a more recent
date than is indicated by the rest of the fauna.

The upper Productus beds, whose whole thickness does not exceed 100
feet, consist of light yellowish sandy dolomite impregnated with silica,
The fossils are abundant and always silicified, but usually only on the
inner and outer surfaces of the shell. Rusty coloured ferruginous dolo-
mites, which are occasionally found in the lower division, are not entirely
absent from the upper one, but they are very rare.

There is said to be a somewhat abrupt lithological change from the
divisions immediately underlying this, but there is no corresponding change
of the fauna, over one-third of the total number of speciés being also found
at a lower horizon. The fauna is a very extensive one, no less than 175
species having been recorded, and Dr. Waagen has expressed an opinion
that this number might be doubled by systematic séarch.?

The general facies of the fauna is permian, though only eighteen species
have been found elsewhere in beds that are believed to be of permian age,
halfin the permian of Europe and half in Armenia. But, mixed with these
palzozoic fossils, there are numerous forms which show more distinctly
mesozoic affinities than those found in the groups below.? Among these the
most remarkable are the Ammonitidg represented by Cyclolobus oldhami,
Arcestes antiquus, A, priscus, Xenodiscus carbonarius, X. plicatus,
Sagoceras hauertanum, When these were discovered, some twenty years
ago,® the finding of well characterised ammonites in beds containing numer-
ous brachiopoda of palzozoic types, and believed to be of upper carbonifer-
ous age or intermediate between that and the permian, was regarded as so
extraordinary that doubt was cast on the accuracy of the observation. Sub-
sequent researches have shown that, even if we exclude all the species having
mesozoic affinities, there is nothing in the fauna to warrant us in assigning
it an older date than the zechstein, or upper permian, of Europe, while the
presence of the mesozoic forms, and their superposition on distinctly per-
mian strata, leads us to regard the beds as even newer than this. The presence
of true ammonites in these beds has conséquently become less surprising
than it was at first, owing on the one hand,to the establishment of the true
age of the Salt range Ammonjtidz as uppermost permian, or even newer,
and on the other, to the”subsequent discovery of Cephalopoda clo.ely
allied to the Ammonitidz in upper paleozoic rocks.

3 Full lists will be found in Pal. Indica, | ® Pal. Indica, series xiii, IV, 213, 223. (1891).
series xiii, IV, 69, 210, 221, (1890-91). 3 Memozrs, 1X, 351, (1872).
128 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [ Chap. VI,

The topmost productus beds of Dr. Waagen, here separated as the
Chidru group, are only about 15 feet in thickness, composed of soft
light yellow sandstones, with coaly bands at the base in some of the sec-
tions. No actual unconformity between these beds and those below has
been detected, but a considerable interval of time probably intervened
between the deposition of the two, for only seventeen of the sixty-three
species are also found in the beds below, and there is a complete change
in the type of the fauna from one in which the Brachiopoda comprise more:
than half the total number of species and the Lamellibranchiata less than
one-tenth, to one in which the proportions are almost exactly reversed. The
fossils are all described as more or less rare, except Margaritina schwa-
geri, which is said to be very common,

The general type of the fauna is distinctly less palawozoic than
any of the preceding ones. Only four species Schizodus rotundatus,
Nucula trivialis, Pseudomonotis radialis, Athyris sub-expansa, are
identical with permian forms of other countries, while of the species
peculiar to this group twenty-four have paleozoic affinities, and no less
than twenty-two are allied to mesozoic forms. So far as the palzon-
tolugical evidence‘goes we are already wellon the way iato the secondary
era, even if the beds cannot be regarded as lowermost trias in age.?

The Chidru group closes the conformable sequence of beds containing
palzozoic fossils. No unconformity between them and the next overlying
beds has been established as yet, but one is suggested by a section,
recorded by Dr. Waagen, in. which the mesozoic beds followed immediately
on the upper Productus beds, with a basal conglomerate but without the
intervention of the Chidru group.* However this may be there must
certainly have been a considerable time interval between the two, for not
asingle species bridges the interval and is found both above and below
the separation of the Chidru group and the ceratite beds which overlie it.

The general aspect of the triassic ceratite beds is such that they might
easily be classed with the Productus beds, the succession varies much as to
details, but consists generally of a thin limestone with Ceratites at the base,
succeeded by a thick marly zone which yields readily to weathering, and
turns a light greenish colour. It is overlaid by grey sandstone, and
flaggy limestone with many ceratites, passing upwards into grey nodular
marls capped by hard limestones and calcareous sandstones. Some of the
bands of limestone contain glauconite, and beds of conglomerate occasion-
ally occur,*

Owing to these beds having been at first confounded with the under-

1A list of species is given in Pal, Indica, | * Pal. Indica, series xiii, IV » 230, (1891).
series xiii, IV, 60, 228, (1890-91). 3 Pal. Indica, series xiii, IV, 227, (1891),
4 Memoirs, XIV, of, (1878).
Clap. VI.J CARBONIEEROUS OF THE HIMALAYAS, 129

lying Preductus beds, and to the later collections not having as yet been
described, it is impossible to give a list of fossils. Ceratites abound and,
most, probably, all of the species described by de Kéninck? are from
triassic beds. Besides these, which are the characteristic fos-ils of the group
species of Orthoceras, Anoplophora, Cardinia, Gervilia, and Rhynconella
are found, among which the bivalves predominate. The most remarkable
form, however, is a species of Bellerophon, a genus not known to occur ia
tecks of later than palzeozvic age in Europe.

In the central Himalayas there is no marked unconformable break
between the lower and upper palzozoic rock systems, such as is found in
the Salt range, and the carboniferous follows with perfect conformity on
the underlying beds.

The oldest rock group which can be regarded as carboniferous is a
crinoidal limestone, usually ced in colour.? Mr. Griesbach’s collections have
not yet been examined, but some fossils brought by Mr. Hughes froma white
crinoid limestone in the Milam pass were ound by Dr, Waagen to
include :—*

Hemiptychina himalayensis, Spirifer glaber.
Notothyris sutvesicularis, Productus semireticulatus.
Athyris royssit. Lyttonia, sp.

The korizon of this tauna is regarded by Dr. Waagen as about that of the
upper portion of the lower Productus beds, The crinoid limestone is over-
laid by a series of fine grained, hard, white, quartzites, in thick beds with a
few shaly partings, which were originally included by Dr. Stoliczka in his
Muth series. In Spiti they are, according to Mr. Griesbach overlaid by, and
partially interstratified with, flaggy dark grey to blue limestones, which
contain Athkyris royssit, and Productus, sp., marking them as carboniferous
in age

In the-Sp‘ti valley General McMahon has recorded the occurrence, in
two places, of beds of fine grained slate, through which small rounded
quartz pebbles are scattered. The similarity of this bed to that of the
Blaini group, which will be described further on, as well as its structure,
are suggestive of the action of floating ice. The exact horizon of the bed
has not been determined, and it is not certain whether it should be classed
with the group just described, or that which overlies it.‘

The white quartzites, with their overlying limestones, are abruptly over-
laid by a group of shales. The junction is said to be unconformable, the
1 Quart. Four. Geol. Sec., X1X, 11, (1863). | * Records, XII, 63, (1879). See also Reeurds,

% Memoirs, XXIU, s9,-(1801). Supra p. 114. | &XI, 151, (1888).
3 Pal. Indica, series xiii, 1V, 167, (1891).
130 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [(Chap. VI.

unconformity being accepted on the strength of the sudden lithological
change, and of the shales resting in different sections on different horizons of
the carboniferous strata. These shales, apparently together with part of
the underlying quartzites, were distinguished by Dr. Stoliczka! as the
Kuling series and regarded by him as carboniferous. Mr. Griesbach has
divided them into a lower portion, composed of dark, generally black,
somewhat micaceous shales, often carbonaceous, with coaly traces here and
there, of permian age; and an upper portion of very similar shales,
difficult to distinguish lithologically, though they are somewhat less earthy
and micaceous, which he regarded as lower triassic in age.

Of the fossils from the Kuling series, described by Dr. Stoliczka, which
were probably for the most part derived from the lower part of the shales,
Productus semtreticulatus and P. purdoni are found in the Productus
beds of the Salt range, the former in the lower, the latter in the middle
division, but are both rare. Spirifer moosakhailensis is found in all three
divisions, but is abundant only in the middle one.

The shaly beds pass conformably upwards into a great series of lime-
stones, the Lilang and Pdra of Dr. Stoliczka, which have been subdivided

by Mr. Griesbach as follows *:

10. Liver coloured limestone with greenish g- ey
shales in Niti, strong limestones elsewhere.

Corbis, sp.
9. Friable shales and earthy beds in Niti, lime-
stones and shales in Spiti and Nilang.

Upper Spirifer lilangenis, Keuper.
8. Limestone. 77 opites, sp. . . . .
7. Earthy limestone and shales . : é
6. Black limestone, flags and dolomites. Dao-
Trias a ane eS
5. Black limestone flags and partings of splintery
black shales, :
4. Hard grey concretionary limestone in thick
beds. Ptychites geraidi. Muschetkatk,
powee 3. Earthy limestone. Rhynconella semiplecta.

shales,
; 4 SPs Buntsand-
1. Dark shales and limestone part- \ Otocevras stein

2, Lim: stones alternating with Posidonomya,
ings, woodwarai.

Speaking generally, the lower part of the central Himdlayan trias may
be described as a series of very dark, almost black, hard limestones with
partings of shales; the upper part varies more, being represented by a great
thickness of friable shales in the Niti and Milam sections, but in Spiti and
to the east it is a limestone formation. The total thickness of the trias is
about 4,000 fect in Niti, but probably exceeds that to the east; in Spiti

1 Memoirs, V, 24, (1865). | ? Memoirs, XXIII, 69, (1891).
   

Spirifer stracheyi. Salt.

Megalodon triqueter. Wulf. Megalodon triqueter. Wulf.

 

  

Ammonites diffssus. Hau,

 

Clydonites oldhami, Stol, Dicerocardium Himalayense, Stol.

 

Ammonites jollyanus. Opp. Ammonites thuilleri.

TRIASSIC FOSSILS. The Calcutta Phototype Cor
Chap. VLI TRIAS OF THE UWIMALAYAS, 131
Dr Stoliczka estimated the thickness at 1,000 to 2,000 feet. The under-
lying Productus shales are quite insigniticant in comparison to this, being
only 150 feet thick, The most noteworthy feature of the Himalayan
trias is the abundant cephalopod fauna of the lowest beds, a fauna
extremely abundant in specimens, though not so extensive as regards
species, and remarkable for its transitional character between a palzeozoic
and mesozvic facies.

In the central Himdlayas the trias is succeeded with perfect con-
formity by a thick series of limestones, formerly
regarded as rhatic and liassic, above which
comes an abrupt lithological change, not known
to be accompanied by unconformity, at the base of
the jurassic series.

According to Mr. Griesbach’s description the
oe 9.—Daonella (Halotia) combined rhztic and lias have a thickness of

omimeli, Wissm. \Trias). : : :
2,500 to 3,000 feet, of which the lias only com-
prises 100, and are divided as follows: 4

 

5. Black shales and dark earthy limestones with oolitic structure; lower lias fossils.
4. Grey crinvid limestones in’ irregular thin beds; Terebvatul.s horrida, Gervilia
inflata.
Lithoden.tron limestone in thick beds.
- A great thickness of limestones and dolomities, Megalodor, sp.
1. A great thickness of dolomites and flaggy dark limestones with thick-bedded
dolomites, which pass downwards into the upper trias.

&

In his original description of the rocks of Spiti, Dr. Stoliczka adopted
a classification different from the more recent one of Mr. Griesbach for the
beds above the carboniferous, It is given here in abstract,?—

Urrer TaGiine (Jics).— Dark earthy or dolomitic bituminous limestone; thickness
nearly 1,000 feet.

Lower Tacuine (lower lias or rhetic).—Dark grey, brown or black limestone,
often earthy or Lituminous, weathering into a rusty brown; more tl-an 1,000 feet thick.

P4ra Limesi ones ‘rhetic or upper trius).—Black, dolomitic, strongly bituminous
limestones ; 700 feet thick. ,

Liane SERIES (upper or- middle trias)—Dark limestones, calcareous shales and

--slates; the limestone compact or oolitic and quasi-concretionary in some of the lower

beds; 1,000 to 2,006 feet thick.

In the preceding pages a brief outline of the published descriptions of
‘the newer paleozoic and older mesozoic rocks of the central Himdlayas
has been given. No details have been entered into, nor has any attempt
been made to clear up the discrepancies which are apparent, as the fossils
that have been collected in this area are at present being examined and de-
scribed, and it has already become evident that this examination will result -
in a considerable modification of the correlations adopted upto now. The

1 Memoirs, XXIII, 73, (1891). | 3 Memoirs, V, pt. i, (:865).
K 2
132 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [O2ap. VL

most important results obtained so far, according to the distinguished
paleontologists, whose names it would be unjust to couple with conclusions
so provisional in their character, are, that there is a very strongly marked
palzontological break between the silurian and the carboniferous, that in
the carboniterous beds two successive faunas can be recognised, one allied
to that of Australia, the other to the Productus fauna of the Salt range,
and that the beds formerly regarded as rhztic and lias should be classed
with the trias, leaving a distinct paleontological break between the trias
and jura of the central Himdlayas. ‘The publication of these results, which
it is hoped will not be long delayed, will be a most important addition to our
knowledge of the stratigraphy of the central Himdlayas, and will doubtless
clear up many of the inconsistencies and uncertainties which now obscure it.!

The beds which have been just described are found in two great
basins of disturbance, one of which stretches along the north of Kumdun
and Garhwdl, the other occupies the valleys of the Spiti and Zanskar
rivers and stretches beyond them to the neighbourhood of Kartse. They
are found again in the Kashmir valley, but before proceeding to the de-
scription of the outcrops it will be best to turn to the Simla district.

The older rock systems of this area have already been described and
we now come to that which has been named the carbonaceous system.?
The beds of this system present certain marked peculiarities by which
they have been recognised with a greater or less degree of certainty over
a large area south of the snowy range, from the western borders of Nepdl
to the confines of Kashmir, but it is only in the neighbourheod of Simla
aud the protected hill states that they have been studied in any detail.

The lowest member of the system appears to be part of what has al-
ready been referred to under the name of Simla slates.2 This name was
applied. to a great series of slates, gritty slates and quartzites, in
which no break has yet been detected, though it is highly probable that one
or more will be established by detailed survey. Whether any of these
beds should be classed with the carbonaceous system is open to doubt,
but the uppermost ones appear to be perfectly conformable with a group
of beds, which have so marked an individuality that they are of tie great-
est importance in unravelling the complicated structure of the hills,
and in establishing the homotaxis of ' this system. The characteristic

1 As these pages were going through the | abundance of cephalopoda in: the lowest beds,
press « preliminary note on the Cephalopoda | where they are poorly represented in Europe
of the Himalayan trias by Dr. Mojsisovics has | Sitz .K. Akad. Wiss., Wien, CI, pt. i, (1892) ;
been published. The whole range of the Euro- | Records, XXV, 186, (1892).
pean trias has been definitely recognised, | * Records, XXI, 133, (1889),
and attention is drawn to the remarkable | ® Supra, p. 117.
Chap. VI] cARBONACEOUS SYSTEM OF THE OUTER HIMALAYAS, 133

member of this group, which Mr. Medlicott named Blaini? after the
stream of that name, flowing westwards from Solan, is a conglomeratic
slate, composed of rounded pebbles of quartz, ranging up to the size of
a hen’s egg, or, in other cases, angular and subangular fragments of slate
and quartzite of all sizesup to some feet across, which are scattered at
intervals througha fine grained matrix. It contains numerous fragments of
the volcanic beds of the Jaunsar system, where exposed in the neighbour-
hood of their outcrop in the Naira valley in eastern Sirmur, and here even
the matrix has much the appearance of a volcanic ash,? though as a rule
there is not the slightest trace of volcanic material either in the matrix or
in the included fragments. This remarkable rock has been observed
from east of Mussooree at intervals to beyond Simla, and is generally,
though not invariably, accompanied by a band, 20 or 3o feet thick, of thin
bedded, usually pink, dolomitic limestone, which lies on top of the zone in
which the beds of conglomeratic slates occur. The agency by which
blocks of stone were dropped over so large an area into a tranquil sea,
in which alone the matrix could have been deposited, must have been
no local one, and the only one that appears at all adequate is that of
floating ice. No smoothed and striated fragments have been found as yet,
though one is occasionally met with showing striation resembling those pro-
duced by glaciers, but the rock has invariably undergone much compres-
sion and disturbance, at times accompanied by a distortion of the shape of
the included fragments, which might account either for the obliteration of
distinctly glaciated surfaces, or for the production of those scratches
which have been observed. No certain conclusion can, consequently, be
drawn from the occasional presence, or general absence, of striation, but
the only alternative hypothesis, that the rock is in fact the volcanic breccia
which at times it resembles, appears to be excluded by the infrequence of
volcanic material in it, and the absence of any other associated volcanic
beds, while the included fragments are too mumerous to be accounted for by
the action of floating drift wood.

The boulder beds are overlaid by a series of shales or slates, charac-
terised by the greater or less prevalence of carbonaceous matter, which
were originally described as infra-Krol*, from the fact that they underlie the
limestone of the Krol mountain. The name has since been so commonly
used in the publications of the Geological Survey that, in spite of a certain
awkwardness, it cannot well be abandoned now.

The carbonaceous impregnation of these shales is very irregularly dis-
tributed, being often extremely conspicuous, especially where the rock

1 Memoirs, Ill, pt. ii, p. 30, (1864). Thename| correct. Records, X, 204, (1877).
was originally spelt Blini, being the spelling | * Records, XX, 156, (1887).
adopted in the Atlas of India. The spelling in | ® Memoirs, ILI, pt. ii, 29, (1864).
the text was subsequently introduced as more
134 GEOLOGY OF INDIA— CARBONIFEROUS AND TRIAS. [ Chap. VI

has undergone crushing, but at other times wanting, at any rate near the
surface. Not infrequently the blackest and most carbonaceous beds
weather almost white by the removal of the carbonaceous element.
Above these beds there is usua!ly a series of quartzites of very variable
thickness, varying from about 20 teet in the sections south of the Krol
mountain to some thousand feet in western Garhwdl. They are very
noticeable at Simla, forming the whole of the Boileauganj hill and the
lower part of Jutogh, where they have been called Boileauganj quartzites.

In western Garhwdl, between the Tons and Pdbar rivers these quartzites
contain fragments of undecomposed felspar, usually subangular but in some
of the beds large and angular, mixed with quartz, mica and fragments of
the accesory minerals of the Himdlayan gneissose granite and gneisses,
the whole forming a rock which, having become foliated by a subsequent
slight metamorphism, is sometimes difficult to distinguish at first glance
from the true gneissose granite."

The uppermost member of the system is another group of car-
bonaceous slates, associated with carbonaceous or graphitic limestones’,
which pass upwards, in western Garhwal, into blue limestones. In the
Krol mountain the uppermost beds are blue limestones with associated
shaly bands, mostly grey in colour, though there is one distinct zone of
red shales, but as no carbonaceous beds are associated with them, and
as the underlying quartzite exhibits remarkable variations in thickness,
it is uncertain whether these limestones of the Krol group are the
equivalents of those just referred to or belong to a later unconform-
able system.

The beds of the carbonaceous system contain, on most of the sec-
tions, interbedded basaltic lava flows and more o: less impure volcanic
ashes, either recognisable as such, or represented by hornblende schists,
where the rocks have become schistose. The range of the volcanic
beds varies on different sections, Their usual position is in the upper
band of carbonaceous shales, but they are also found among the quartz-
ites and in the upper part of the infra-Krol, though they never, so far as
is known, extend down as far as the Blaini group,

In Kashmir fossiliferous beds of upper paleozoic age are underlaid
by a great slaty series, the whole of which was grouped by Mr. Lydekker
in his Panjal system? and regarded as silurian and cambrian, partly on ac-
count of its underlying supposed carboniferous beds, partly on account of
the opinion regarding the age of the Blaini group which was prevalent
when he wrote. A large part of these slates are doubtless of older palzo-

1 Records, XX, 160, (1888). I * Records, XX, 148, (1868).
3 Memoirs, XXM, 209,-(:883).

e
Chap. VI

CARBONIFEROUS ROCKS OF KASHMIR. 135
zoic age, but part at least appear to be more properly classed with the
upper palzozoics.

The oldest of the beds with which we are at present concerned is a
conglomeratic slate, composed of subangular fragments and rounded
pebbles of slates and quartzites, imbedded in a matrix of fine grained slate.
The rock is in every way similar to the Blaini group of the Simla area,
and the same arguments in favour of a glacial origin are applicable in
both cases. Besides this, the Kashmir boulder bed occupies th esame
position relative to fossiliferous beds of carboniferous age as the glacial
boulder clay of the Salt range does relative to the Productus beds, and as
it is reasonable to suppose that the extreme cold which affected the one
area must have extended to the other, we may take it as certain that the
so called Panjdl conglomerate is also of glacial origin.

‘The boulder slate is overlaid by a series of quartzites ard black car-
bonaceous slates, in the upper portion of which there is an abundance of
contemporaneous volcanic rock, and above these there are thin bedded,
light blue and white fossiliferous limestones from which a number of
fossils have been collected by different observers. The following list,
quoted on the authority of Mr. Lydekker, includes all those known up to
now: —1?

 

CyePHaLopoDa— Rhynconeila barusiensts.
Orthoceras, sp. 5 kashmtriensis.
LaMBLLIBRANCHIATA— E. P. 5 pleurodon var. dav: e-
Avicula, sp. uxiana.
Aviculopecter, sp. Spirifer barustensis.
Solenopsis, Sp» » kashmiriensts,
Ss. » ketlhavit, (S. Raja).
BE ols A. P. »» moosakhailensis.
E P, Athyris subtilita, A.P. setig
A.E.P. 4, royssit, ge aeches
Camerothoria, sp. t shinies
2 a ” .
AE Ghadetes te pepe. E. A. P. Spiviferina octoplicata,
=. m me on : E. A. P. Streptorhynchus crenistria.
Pe E. A. Strophomena romboidalis var.
35 laevis. wuedoud:
Discina kashmiriensis. fess s ;
Orthis, sp. evebratula austeniana.
E. A. P. Productus cora. ; ” sacculus.
E. P. ” costatus. PoLtyzoa—
E. P. m2 i P. Fenestella (?) lepida.
ze er P. » sykests
E. AP. 5 longispinus. P, ‘i mavastomnh
E. A. ” scabriculus. A. Protoretepora ampla.
E, A. P. ” semireticulatus. E. Vincularia muitangularis.
E. » (?) spinulosus.
E. P. ‘ striatus. CRUSTAGEA—
Retsia, sp. E. Phillipsia (2?) seminifera,
: 1 Memetrs, XXL, 158, (.883).
136 GEOLOGY OF INDIA—CARBONIFZROUS AND TRIAS. [Uhap, VL

To these may be added Lytton/a, which, according to Mr. Lydekker, is
seen on the weathered surface of the limestones.!

In the foregoing list the letter E denotes that the species is also found
in the mountain limestone of Europe, A in the marine carbonife-ous of
New South Wales, P in the Productus beds of the Saltrange, S in the
Xuling series of Spiti. The determinations are all on the authority of
Dr. Feistmantel, as quoted by Mr. Lydekker. Dr. Waagen, who has
adopted a narrower definition of a species, finds only two, vzz. Athyrts
subitlita, Hall.=Spirtgerella derbyi, Waag., and Spirifer moosakhailen-
sis, Dav., identical with Salt range forms? Discina kashmiriensis, Davy.,
is said to be almost identical with D. warth:, Waag., from the beds at the
base of the speckled sandstone, and Dr. Waagen is inclined to place the Kash-
mir beds ata horizon intermediate between these and the Preductus beds,
representing in fact part of the unfossiliferous speckled sandstone. The
number of species identical with, or closely allied to, Australian forms is dis-
counted by the fact that three-quarters of them are also found in Europe,
but so far as it goes is more in accordance with Dr. Waagen’s correlation
than with a later date, and we may conclude that the carboniferous fossili-
ferous beds of Kashmir are somewhat older than the lower Productus beds
of the Salt range. :

The carboniferous of Kashmir is overlaid by a series of limestones,
exposed in synclinal folds of various sizes, at either end and north of the
valley. They are the equivalents of the triassic limestones of Spiti.
They are sparingly fossiliferous, and though a considerable number of
forms have been obtained from them, the only ones specifically identifiable
are Ammonites (Ptychites) gerardi, Megalodon gryphoides, and Spirifer
stracheyt. Besides these, stems of Crzno/ds, Orthoceras, doubtful Ceratites
and Gondatztes, and several genera of Gasteropods and Corals, all more
or less doubtful, have been obtained.#

It will be seen from the descriptions that there is a great. similarity
between the sections in Kashmir and the Simla area. In both, boulder-
bearing slates of presumably glacial origin are overlaid by a series of
slates and quartzites, characterised by a carbonaceous impregnation and by
the presence of contemporaneous volcanic beds, and in both the uppermost
member isa limestone, The resemblances are not mere lithological ones,
between rocks such as have always been in process of formation at every
period of the earth’s history. They are exhibited by rocks which owe their
origin to wide reaching causes which have only occasionally acted, and it
is difficult to resist the conclusion that they are evidence of the contempors
aneous origin of the two rock series, and not merely accidental.

3 Records, XVM, 37, (1884). § Memoirs, XXII, 158, (1883).
2 Pal, Indica, series xiii, 1V, 166, (1891).
Chap. VII AGE OF THE CARBONACEOUS SYSTEM. 137

The country intermediate between the Simla and Kashmir areas has not
as yet been examined in any detail, but we know from Colonel McMahon’s
descriptions that similar carbonaceous beds with associated volcanic
basaltic traps, underlaid by a conglomeratic slate resembling the Blaini of
the Simla area, are found in Chamba and near Dalhousie.!

These observations serve to link the Kashmir and Simla sections and
strengthen the conclusion that was based on lithological resemblances. It
has not as yet been corroborated by tne discovery of fossils in the south-
eastern area, though the rocks are in many places perfectly adapted
for the preservation of organic remains. Even in Kashmir territory, once
the drainage area of the valley is left, fossils become rare. As we go south-
eastwards they become more and more sporadic in their occurrence, and,
except in one small area in western Garhwédl,? not a single fossil of older
date than tertiary has yet been found south of the first snowy range.

The correlation of the beds of the carbonaceous system has long been a
stumbling block in the way of our knowledge of Himdlayan geology. When
Dr. Stoliczka visited Spiti in 1864, the rock systems below the Blaini had
not been identified, and the section along the road through Simla was
believed to represent pretty fully the sequence of stratified rocks in the
lower Himalayas. He attempted accordingly to distribute the beds of the
Simla section over the section seen in Spiti, and suggested that: the
Blaini ‘conglomerate’ was the equivalent of certain conglomerates in
the Muth series—Haimantas according to the classification adopted in
this work—and consequently at least as old as silurian.? It does not
appear that Dr. Stoliczka offered this correlation as more than a guess,
to which he probably himself attached small value. It seems certain that
he did not recognise the peculiar character of the Blaini rock, but regard-
ed it as an ordinary conglomerate, and he never saw the rock in the Spiti
river, which is an exact equivalent of the Blaini conglomeratic slate.
Yet the glamour of his genius has shed an importance over this
guess which it was never intended to possess, and time after time the
Blaini group has been unquestioningly referred to as silurian. The more
probably correct correlation was pointed out in 18884 and the evidence,
then was practically as strong as it now is. It comes as near certainty
as is possible in the absence of palzontological evidence, while there is
really no evidence worthy the name in favour of the older view. Yet such
is the vitality of error that the older palzozoic age of the Blaini has been
accepted without question in one of the latest publications on Himdlayan
geology, and the very writer who first drew attention to the probability

1 Records, XIV, 305, (1881); XV, 34, (1882);| 8 Memoirs, V, 141, (1855).
XVII, 34, (1884). 4 Records, XX, 142, (1888).
® See infra, p. 229.
138 GEOLOGY OF INDIA—-CARBONIFEROUS AND TRIAS. —[Chap. VI.

of its being carboniferous or permian, is quoted as supporting the vicw
which he combated.!

The other correlations of Dr. Stoliczka, of the quartzites of Boileauganj
with the Kuling, and of the Krol with the Lilang limestone of Spiti, are prob-
ably correct, and curiously enough an apparent confirmation was published,
about the same time as his memoir, in Professor Giimbel’s description? of a
specimen from the Schlagintweit collection, said to have been obtained at
Dharampur near Solan in the Simla district, containing three fossils,
Lima lineata and Natica gatllardoti, found also in the Muschelkalk of
Europe, and a new species, NV. simlaensis. Dharampur in the neighbour-
hood indicated is, however, a well known locality on tertiary rocks,
and specimen in question must have come from a totally distinct ground,
probably in Tibet.

In the Kdgd4n and Kishenganga valleys, north-west of Kashmir, there
are a number of small outcrops of carbonaceous slates, overlaid by white or
buff crystalline limestone, folded into the gneiss and schists.? They are
probably representatives of the upper palozoic and triassic rocks of Kash-
mir. They have not been closely examined or surveyed, and are mentioned
here merely as indications of the former extension of these rocks, and as
occupying a geographical position intermediate between the carbon-trias,
of Kashmir and of the Hazara district of the Punjab.

At the extreme north-western extremity of the Himdlayas, fossi-
liferous rocks are found south of the snowy ranges in the district of
Hazara. No fossiliferous beds cf carboniferous age have yet been identi-
fied in this corner of the Punjab, and the only indication of their pre-
sence, west of the Jehlam and north of the Salt range, is the dis-
covery, by Mr. Lydekker, of Productus humboldti in a loose block of
limestone near Hasan Abddl.t The age of the rocks underlying the
triassic group of the Sirban mountain near Abbottdbdd is uncertain, but
they are quite unconformable to the underlying Attock slates, and may be
catboniferous. They comprise two divisions; the lower consists of sand-
stones, shales, and silicious limestones, all red in colour, wite an argilla-
ceous breccia, full of fragments derived from the underlying rocks at its
base, but the published descriptions are insufficient to determine whether
this may ot may not represent the glacial boulder bed at the base of the
carboniferous in the Salt range and in Kashmir. The upper division is
composed of dolomites only, lighter in colour than the lower beds, often
highly silicious and of considerable thickness. These dolomites are over-

! Yemoirs, XXII, 54, (1891). 2 Memoizs, XXII, 205, (1883).
2 Sitgungster, K. Bair. Akad Wiss, Min-| * Manual, ist edition, p. 501.
chen, 1865, Bd. II, p. 364.
Chup. VI HAZARA DISTRICT. 139
laid by a group of heematitic rocks, quartz breccias, sandstones, and shales
which may belong to the trias.}

In western Hazara there is a great series of much contorted rocks to
which Mr, Wynne has given the name of Tandwal (lanol),? from the name
of the district they are found in. They comprise an enormous thickness of
grey and drab quartzites and quartzose beds, in rapid alternation with dark
earthy beds, flaggy, shaly or slightly schistose, associated with conglome-
ratic slates containing pebbles of quartz and quartzite, ranging up to
the size of a goose’s egg. In the synclinal folds are thick zones of various
coloured pseudo-brecciated, silicious, cherty or compact, grey, black and
buff dolomitic limestones, with which are occasionally associated intensely
black graphitic and sulphurous shales, or else purple and red sandstones
and slaty bands.

These rocks, whose general description accords fairly well with that of
the carbonaceous system of the Simla region, are regarded as the equi-
valent of the beds below the trias of the Sirban mountain, chiefly on ac-
count of their superposition and probable unconformity to the Attock slates;
and partly because of the occurrence cf red and purple slates and quartz-
ites at the base of each.

Triassic rocks attain a great development in Hazdra, being 1,500 to
2,000 feet thick in the Sirban mountain, and some 3,000 to 4,oco feet near
Khdnpur. Owing to the disturbance they have undergone they occupy
a number of small expcsures, too small to be shown on the map, in the areas
coloured as silurian and nummulitic, respectively. In the Sirban mountain
they consist chiefly of black or dark grey, distinctly bedded, limestone,
with thick zones of massive dolomite, some of which contain numerous
opaque lamina of quartz. Near Abbottébdd, where the series is complete,
dolomites form the lowest beds, and are followed by thin bedded, fossilifer-
ous limestones, containing Megalodon, Dicerveardium, Chemnitata, and
Gervilia. The dolomites are not always present, and the base of the series
may be formed by the limestones, which are succeeded by quartzites and
dolomites of considerable thickness, again overlaid by thin bedded lime-
stones and slaty shales containing Nerinea, Nevitopsis, Astarte, Opis,
Nucula, Leda, and Ostrea#

The other exposures of triassic rocks in Hazdra, while exhibiting some
variations, do not differ essentially from those of the Sirban mountain, and
it is not certain how far the differences which have been observed may be
only due to the obscuring effect of the intense disturbance they have
undergone.*

Carboniferous and triassic rocks can be traced along the southern slopes

1 Memoirs, X, 335, (1872). 3 Mencirs, 1X, 336, (1872).
2 Kecoras, X11, 122; (1879). 4 Records, XIi, 124, (1879).
140 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [OLap, VI.

of the Pfr Panjal and Dhdola Dhér ranges, but no fossils have so far been
found, except some obscure gasteropods in the Jehlam valley. They are of
the ordinary Kashmir type of quartzites and carbonaceous slates, underlaid
by the boulder. bearing slates and overlaid by limestones. There are some
inliers of massive grey limestone in the tertiary area, which are faulted up
on their south-west side. No fossils have been found in the limestone,
which is bedded, compact, dark grey to black in colour, resembling the
limestones of the Himalayas, and is probably of triassic age or older. They
have been coloured brown on the map, as that represents their most
prcbable age, but it must not be left out of sight that they may well belong
to the older, precarboniferous series of limestones of the Himalayas or
to a later post-triassic age.!

To the north of Kashmir a series of limestones, slates, and quartzites
are found north of Iskardo (Skardo) in Baltistan, which are probably triassic
and carboniferous in age. Further eastward beds belonging to these periods
are known to occur in the Chang-
_ cheng-mo valley and the Kara-

koram range, only isolated details
are, however, known, and it is
impossible to give a connected
account of them, but they cannot
be passed over without a notice
of that remarkable group of fossils,
allied to the Foramintfera, known
as Syringospheridz, which are
found in dark shales, below a lime-
stone taken to be of triassic age,
Fig. 10.~Syringosphera verrucosa, Duncan. on the northern side of the Kara-
koram pass. ‘They are small, rounded or oval bodies of about an inch in
diameter, and had long been known as Karakoram stones. Almost the
last work of the late Dr. Stoliczka was the collection of a number of speci-
mens which were described by the late Professor Duncan® under the
generic names of Syringosphera and Stolicekaria, the former including
seven, and the latter one species.

 

 

 

 

 

Marine carboniferous rocks are known to occur in Afghdnistdn, in
the Herdt province, in the Hindu Kush,? and north of the Safed Koh,

1 Records, 1X, 53) (1876); Memoirs, XXII, | Records, XXII, 80—88, (180).

202, (1883). 8 C.L. Griesbach, Records, XVIII, 62, (1885) ;
2 Scientific Results of the Second Yarkand XIX, 49, 240, (1886); XX, 17, (1887).
Mission, Syringospheride, Calcutta, 1879:
CAKBONIFi:.ROUS OF TENASSERIM.

Chap. VI.J 14!
where they have undergone considerable metamorphism and are pene-
trated by granite veins.?

According to Dr. Fleming boulders of Productus limestone were found
by him in the streams which flow eastwards from the Suldi:ndn range,?
but subsequent observers have not been able to detect carboniferous rocks
in this range.

The marine carboniferous rocks are overlaid, in Herat and Turkestdn,
by a series of plant-bearing sandstones with seams of coal, which appears
also to be represented south of the Safed Koh? These are lithologically
and stratigraphically the equivalents of the Gondwana system of the Indian
Peninsula, and as such their description will be deferred to a subsequent
chapter.*

In Tenasserim the only other region where marine deposits of carboni-
ferous age are known to exist, there is a great accumulation of psendo-
porphyritic sedimentary beds known as the Mergui group,’ whose
principal feature is derived from imbedded fragments of felspar. The
rock in its normal form is earthy, but highly indurated, passing into
slaty masses without the conspicuous felspar fragments on the one hand,
and on the other into grits and conglomerates. Resting on these
grits, are dark coloured earthy beds, finely laminated, with hard quartzose
grits. These rocks cannot be than 9,000 feet in thickness, and in

places they must be 11,000 or

2,000 feet.

They have only been noticed

hitherto near Mergui, and coat is known of their relations.

The beds of the last group in the Tenasserim valley are succeeded by
the Maulmain groups of hard sandstones, often in thin and massive layers,
with thin earthy partings, sometimes in fine laminz, the prevailing colour

1C.L. Griesbach, Records, XXV, 71, (1892).
The statement that “lower Silurian fossils from
the Khyber hills were found by Dr. Falconer in
the gravel of the Cabul river” was nade by

. Colonel (then Captain) H. H. Godwin-Austen
in 1866, in a paper which appears to have been
drawn up from field notes without means of
access to published info.mation, and no refer-
ence is supplied to any original authority ;
Quart. Four. Geol. Soc. XXII, 29, (1866). No
notice of the discovery of such fossils can be
found in Falconer’s published writings, and
the only original published statementis in a
footnote to a paper by Captain Vicary, Quart.
Four. Geol. Soc. VAI , 45, (1851). Vieary him-
self obtained “a small Spivifer, Orthis in abund-
ance, a Terebratula and some Polyparia”
from limestone boulders in the watercuurses
near Peshawar. Ina footnote he adds, “Dr.

 

Falconer obtained specimens of Spirifer,
Orthis, and other palzozoic forms from these
mountains several years age.” /1s>,ina note
by Sir R. Murchison »prefixed to Vicary’s paper,
the discovery of palzozoic fossils is mentioned.
Now, it is quite possible that the fossils collected
by Falconer and Vicary have been examined
and their age determined, but as this is not
stated, some doubt remains whether the fossils
may not have been carboniferous, as they were
said to be by Verchere | four. As. Soc. Beng.,
XXXVI, pt. ii, 21, (1867)], the Orthis being
perhaps O:thisina or Stieptorhynchus creni-
stria, formerly inciuded in the genus Orthis.

2 Quart. Four. Geol. Soc. 1X, 348, (1853).

3C. L. Griesbach, Records,X XV, 79, (1892).
4 Infra, p. 196.

5 T. Oldham,
32, (1856).

Sel. Ree. Govt. India, X,
142 GEOLCGY OF iNDIA—CARBONIFRROUS AND ‘TRIAS. [Chap. VIL

is a reddish tint, and some of the layers are calcareous. Some of the
more soft and earthy beds contain marine fossils. Over these sandstones
occur grey shaly beds, also sometimes calcareous and fossiliferous, with
occasional beds cf dark sandstone, then. come 150 to 200 feet of fine soft
sandstone, thinly bedded, with grey and pinkish shaly layers intercalated
and upon these again, hard thick limestone. The fossils found appeared
to be of carboniferous age, Sperzfer and Productus being the commonest
forms, but the species have not been determined, and it is rare to obtain
specimens in a state suitable for identification.

The thickness of these beds is estimated at about 5,000 feet, exclusive
of the limestone, which is itself 1,100 feet thick near Maulmain.

Near Maulmain the limestone is extremely conspicuous, and forms
large hills and ranges, extending far to the south-south-east up the valley
of the Attaran and Zami. The same rock occurs east of the Salwin, but
does not extend far into Martaban, and is wanting in the Sittaung
valley.. Fartherup the Salwin, however, in Karen-ni, and elsewhere
beyond the British frontier, large tracts of limestone occur, probably be-
longing to the carboniferous series. Limestone is said to abound in the
Mergui archipelago, and may very probably be, in parts at least, identical
with that found near Maulmain.!

Besides the rocks already described, which can be ascribed to a car-
boniferous or triassic age with more or less certainty, there are some rocks
coloured brown on the accompanying map, in south-east Afghdnista4n and
in the hills east of India proper, whose true age is doubtful.

The first of these areas to be dealt with is in southern Afghanistan,
where some unfossiliferous slates have been coloured on the map as
carbon-trias. They form the Khwd4ja Amran range, the hills north of the
Pishin valley, and on either side of the upper Zhob-valley. They are
slates and quartzites, whose similarity to the slates of the Simla area
has more than-once been noticed. In the’ first published description
they were regarded as a flysch type of the nummulitic shales and lime-
stones to the east of Quetta.? Subsequently the same observer considered
that they were more probably lower cretaceous,3 but in the absence of fossil
evidence there is no more ground for this than almost any other corre-
lation. One thing seems certain, that they are not altered nummulitics,.
for near Spira Raga, on the frontier road to Pishin, the same beds are

' Dr. Noetling’s discovery of silurian fossils ] rian not carboniferous.
in the similar limestone of the Shan hills | ? #emotrs, XVIII 32, (1881).
makes it possible that this limestone is silu- | 8 Records XVIII 65 (1885)
: ? 7 = a
Chap. VI]  UNFOSSILIFEROUS ROCKS OF DOUBTFUL AGE, 143
found within a few miles of typical nummulitics which show all the groups
characteristic of that series in Balichistan.}

In this neighbourhood and in the hills bounding the upper Zhob valley
about Hindubdgh, the beds are penetrated by intrusions of serpentinous
rock, porphyritic with crystals of diallage, precisely resembling the intru-
sive serpentine of Burma and Manipur. Among the slates a bed con-
taining subangular fragments, ranging to six feet in diameter, was observed.
The bed very much resembles the Blaini boulder bed of the Simla area,
but no great weight can be attached to a single isolated observation like
this. It has, however, taken in conjunction with the general lithological
facies, and the resemblance of the serpentine intrusions to those of the
Arakan hills, been allowed to influence the choice of the colour to be
adopted in the map, and these rocks have been coloured as carbon-trias,
with a warning note that the age is unknown and may very likely be
younger than that indicated by the colour.

There is some independent evidence of the possibility of triassic rocks
being found in Balichistdn, for Dr. Cook has recorded the finding of
Orthoceras near Kheldt. The other fossils found for the most part indi-
cate a cretaceous horizon and the section as described fits in very well
with the known cretaceous and tertiary groups of Baldchistdn ; there can
then be no doubt of its correctness. On this ground it might be natural
question the identification, but the brief description Dr. Carter gives is
incompatible with the idea that it was a Baculites or the phragmocone of a
Belemnite* It is more probable that there is an undetected unconformity,
and that both the cretaceous and the'trias are represented in the section.

Between the Irawadi valley and the Arakan coast a tract of country
has been coloured as carbon-trias on the map, with a note against it that
newer rocks are known to occur and probably form a large portion of the
area. ‘The beds so mapped compose the Arakan Yoma, a forest clad
range only traversed by a single road and by a few difficult paths at wide
intervals, and, in general, absolutely inaccessible, except along the
tortuous beds of streams. Anything like satisfactory geological surveying
becomes almost impossible in such a region, unless some well marked
and prominent beds occur to afford a clue to the stratigraphy, or fossili-
ferous belts are numerous. In the Arakan range acither is the case, the
rocks of the main range consist of rather hard sandstones and shales,
greatly squeezed, contorted, and broken, traversed by numerous small veins
of quartz, often slaty, and sometimes schistose, but there is a marked,
deficiency of .any conspicuous strata. The few bands of limestone
which occur are thin, isolated, and as a rule unfossiliferous, The rocks on

1R.D. Oldham, MS. Report, 1891. It is | they appear to be cortiruers with them.

not established that these rocks are the same | ? Your. Bemtay, Br. Roy. As. Soc. V1, 190,
asthose of the Khwaja Amrdn range, though | (1892),
44 GEOLOGY OF INDIA— CARBONIFEROUS AMD TRIAS. [Ckan, VI.

the western, or Arakan, side of the range seem, on the whole, less altered
than those on the eastern, or Pegu, slope, and unaltered nummulitic
rocks appear, on both sides throughout a great part of the area, although
not continuously on the outer spurs.

The crushed, hardened, and somewhat altered rocks of the Arakan
Yoma were criginally separated by Mr, Theobald from the newer-looking
nummulitics under the name of axials, and considered as comprising
the oldest tertiary beds and their immediate predecessors in the series.
Although there is a well marked difference between the nummulitic beds
and the axials, there is no distinct break between them. The two pre-
sent an appearance of conformity, and it is far from clear that some of the
axials are not merely nummulitic strata, greatly crushed and contorted.
But subsequent to thé preliminary examination of the area, a cretace-
ous ammonite was found in Arakan, and amongst some rather obscure
fossils discovered near the former frontier of British and Native Burma, west
of Thayetmyo, were a few specimens referred by Dr. Stoliczka to the
typically upper triassic Halodia lommelt. \t became, therefore, neces-
sary to distinguish both triassic and cretaceous beds amongst the axial
rocks of the Arakan range.

To the former has been referred a series of hard sandstones and
shales, with grits and conglomerates, and a few bands of impure lime.
stone, which form the crest of the Arakan range at the old frontier
of Lower Burma, and extend southward, nearly to the parallel of Prome.
The only characteristic beds are some white speckled grits, interbedded
with shales and sandstones, and attaining a thickness of 1,300 feet, in
the Hlwa (Lobwa) stream, 35 miles west of Thayetmyo; a band of dark
blue shale, part of which is calcareous, 33 feet thick below the grits with
conglomerate; and some thick bedded shales, passing into massive sandy
shales with hard nodules interspersed, attaining a thickness of 110 feet,
and containing “a Cardta and some undetermined Gasteropoda. The
calcareous conglomerate passes into a rubbly limestone, and appears
identical with the beds containing the supposed Halob/a lommel¢. To the
northward a band of limestone, much thicker and purer than that of the
Hlwa stream, has been traced in several places. The speckled grits and
conglomerates are, however, more conspicuous and more ch:racteristic,
and it is mainly ky means of them that the area of supposed triassic beds
was mapped,

The whole thickness of the group appears to be rather less than 6,000
feet, the characteristic beds just noticed being near, but not at the base of
the group. To the eastward these beds are in contact with nummulitic
strata; to the westward it is believed that cretaceous beds come in, but the
country is difficult of access, and has not been surveyed. ‘Ihe area occu-
pied within the limits of Lower Burma is elcngately triar gular, broadest
Chap. VLJ ROCKS OF THE ARAKAN YOMA, 145

at the frontier, where it extends for fifteen miles from east to west, and
terminating in a point to the southward, west-by-north of Prome.?

The remainder of the rocks forming the Arakan Yoma, excluding those
of cretaceous age, are either unfossiliferous or the few organisms which
have been detected, mostly the indistinct remains of ‘plants and mollusca,
are insufficient to afford any trustworthy indication of age. They have
been classed by Mr. Theobald as Negrais rocks, the name being derived
from Cape Negrais, the south-western point of Pegu, and the extreme
southern termination of the Arakan Yoma.

The Negrais rocks differ in no important particulars from the beds
already noticed. They consist principally of hardened and contorted sand-
stones and shales, intersected throughout by numerous small veins of quartz
and carbonate of lime, Limestone is not of common occurrence. Where
seen, it does not generally appear in regular strata, but in huge detached
blocks imbedded in the shales and sandstones, as if the latter had yielded
without fracture to the pressure which dislocated the limestone. Conglo-
merates also occur, sometimes passing into breccias.

The alteration of these beds is most capricious and irregular. [re-
quently for a long distance they are apparently unchanged, except in being
somewhat hardened; then they become cherty, slaty, or sub-schistose,
and cut up by quartz veins. One not uncommon form of alteration is
exhibited by the rocks affecting a greenish hue, due to the presence of
chlorite, such rocks being generally much cut up’ by quartz veins. In a
few instances, apart from the serpentine intrusions to be mentioned pre-
sently, irregular dykelike masses of either serpentine or a decomposed
steatitic rock are found, but this is far from being of frequent occurrence.
A more common form of alteration, seen along the coast north of Cape
Negrais, is apparently due to the infiltration of silica in large quantities,
and is shown by the intense, and often abrupt, alteration of beds of sand-
stone into cherty masses.

No satisfactory classification of these the main rocks of the Arakan
Yoma has been practicable. They must be of great thickness, but the
stratification is too confused for a clear idea as to the succession of dif-
ferent strata to be formed, in the absence of any well defined horizon.
Some of them appear to be acontinuation of the Ma-f, or cretaceous group,
but on the other hand it is impossible to draw any definite line of
boundary between the hill rocks and the nummulitics of Pegu. In

' It must be remembered that the specimen ( nummulites, and make it probable that most
ascribed to Halobia lommeli was a mutilated | of the rocks of the Arakan Yoma are lower
and ill preserved one, Recent investigations, | tertiary. As the results have not yet been
conducted while this work was passing through | fully worked out, the text is allowed to stand

the press, and as yet incomplete, have shown | substantially as originally written.
that the supposed trjassic rocks contain
146 GEOLOGY OF INDIA—CARBONIFEROUS AND TRIAS. [Qhap. VI.

Pegu, away from the base of the hills, comparatively soft, unaltered, fossi-
liferous beds, belonging to the older tertiary period, are found, which appear
to rest upon the hill beds, for, away from the axis of the range, both have
an eastwardly dip. The two rocks contrast strongly, the nummulitics be-
ing soft and unchanged, the hill beds hardened, crusked, and in places
almost schistose, but it is impossible to fix a precise limit to either.
The two are never seen in contact, there is no evidence that they are
faulted against each other, and there appears to be a belt, often two or three
miles wide, of rock in an intermediate condition. It appears possible
that the rocks of the Arakan Yoma comprise representatives, slightly
altered, of both cretaceous and nummulitic rocks, but there is no clear
proof that these Arakan Yoma beds are identical with the Pegu nummu-
litics, and it appears best to distinguish the hill rocks by a separate name,
though it has hitherto proved impcssible to draw a line between the
two.

From the foregoing description it will be seen that fossils have only
been found at two places near the northern limit of the coloured area, and
that there is no certainty as to the extent of the older rocks they indi-
cate. The Negrais group was originally regarded as very possibly num-
mulitic, the lithological difference and greater induration, as compared
with the undoubted nummulitics of Pegu, being attributed to the dis-
turbance it had undergone, It is very doubtful, however, whether the
explanation is sufficient, and in view of the probability of their distinct-
ness, and of the fact that they are described as exhibiting a greater degree
of induration than the beds which. were supposed to be cretaceous, it has
been considered advisable to adopt the course pursued in the preparation
of the accompanying map, and colour them the same as the rocks known
to be of carboniferous and triassic age, appending a warning note that
their true age is unknown.

The intrusive serpentine which has already been noticed generally
occurs as irregular shaped bosses of varying dimensions,! but dykes
also occur, especially north-west of Prome. The rock is a characteristic
dark coloured serpentine. It frequently becomes a gabbro, contains
porphyritic crystals of bronzite, and is intersected by veins of gold coloured
chrysotile, or, sometimes, of carbonate of magnesia. Occasionally it ap-
pears to be replaced by a form of greenstone which may possibly be dis-
tinct, although the two rocks occur in the same neighbourhood. The hills
formed of serpentine may be distinguished at a distance by their
barrenness, They appear to support little except grass and a few bushes,
while the greenstone hills are covered with luxuriant forest. In all

‘ None are sufficiently large to be marked on the map issued herewith.
Chap. VL] SERPENTINE INTRUSIONS OF THE ARAKAN YOMA, 147

probability the serpentine and greenstone outbursts were originally the
same or nearly the same, and the former rock has undergone a chemical
change.

In the neighbourhood of some of the larger masses of serpentine the
sandstones and shales are converted into greenstone and chloritic schist,
but the effect varies, and in some instances the neighbouring rocks appear
almost unaltered, It is, however, worthy of notice that, except far to the
northwards, all the outbursts of serpentine appear confined to the Pegu,
or eastern, side of the range, and that, as has already been stated, the rocks
on this exhibit, as a rule, more alteration than those on the western
slopes of Arakan. ‘To the northward, near the northern frontier of Pegu,
serpentine occurs on the highest hills of the Yoma, and, in one instance
at least, on the western side, but elsewhere all the outbursts detected
are not only east of the main range, but near the eastern limit of the hill
rocks. Nota single intrusion has been detected in the unaltered num-
mulitic rocks.

It is unnecessary to describe the distribution of the serpentine masses
in any detail. They are principally collected in three groups, the most
northern of which consists of the largest mass known, a horseshoe shaped
intrusion, some five miles in length, forming the Bidoung hill, nearly due
west of Thayetmyo. Several masses occur north-north-west of Prome,
and one of these, forming a long dykelike mass for about five miles
along the boundary between the nummulitics and the supposed trias,
appears to alter the triassic rocks, but not the nummulitic beds, although
the latter are greatly crushed. Probably the difference is owing to the
eastern boundary being a fault. The third group is west of Henzada, where
twenty-one distinct and isolated intrusions occur, scattered over a length of
twenty-six miles from north to south, close to the edge of the unaltered
nummulitic area. The largest of these masses is about three miles long
by perhaps half a mile broad, but the majority are less than a mile in
diameter. Besides the principal groups a few small and unimportant
outbursts are found isolated here and there, but none are found south
of the area west of Henzada.

Further north a series of slates and indurated sandstones, which very
much resemble the axial beds of Burma, is found in Manipur. They
occupy the hills surrounding the valley of Manipur, and are penetrated by
intrusive serpentine of the same type asin Burma. The intrusions, more-
over, are confined to he neighbourhood of the eastern limit of the hill rocks.
Our only information regarding the geology of Manipur! is derived from

1 Memoirs, XUX, 217, (1883).
2
148 GEOLOGY OF INDIA~—CARBONIFEROUS AND TRIAS. [Qhap, VI.

rapid traverses, where the movements of the geologist were determined by
political considerations, and there is consequently no detailed information
available. It was believed, however, that two unconformable, pretertiary,
rock series were observed, the upper one being composed of red slaty shales
overlaid by limestones, lithologically identical with those of the Ma-f group
in Arakan. Some beds of volcanic ash, observed on the slopes of the
Kachao mountain, were believed to be attributable to this group, but the
correlation is questionable.

The rocks seen in Manipur, show some resemblance to those of the
carbonaceous system of the Simla area, though the disturbance they have
undergone is less intense. A bed of conglomeratic slate, containing
rounded boulders of quartzite imbedded in a fine grained matrix, resem-
bling in structure the conglomeratic slates of the Blaini group, was seen
on the road between Manipur and Kohima associated with black car-
bonaceous slates.

Further to the north, in upper Assam, Mr. Mallet distinguished, under
the name of Disang, a group of shales overlaid by sandstones, which are
separated by a faulted boundary from the coal measures and overlying
tertiary rocks lying between them and the alluvial plain of the Brahma-
putra valley.’ Sofar as their lithology goes, they agree fairly well with
the older rocks of Manipur and the Nagé hills, and have been coloured the
same on the map, though their true age is very uncertain.

} Memoirs, X11, 286, (1876).
CHAPTER VII.

THE GONDWANA SYSTEM.

Gonxowhna System—Probably of fluviatile origin— Relation to present river valleys—Division
into groups— Lower GonpwAnas—Talchir group—Karharbéri group — Damuda series —
Barakar group—lIronstone shales—Rdanfganj group —Motur and Bijori_ groups—K4mthf
group--Pdnchet group—Almod group—UprerR GonpwAnas—Mahddeva beds— R4j-
mahal serits—in the Réjmahdl hills—and on the east coast—Kota-Maléri group—Chiki-
ala group—Jabalpur group—Plantbearing beds of Cutch and Kathidwér.

The upper palzozoic and older and middle mesozoic formations of
other countries are represented in the Indian Peninsula by a great
system of beds, chiefly composed of sandstones and shales, which, except
for some exposures along the east coast, appear to have been entirely
deposited in fresh water, and probably by rivers. Remains of animals
are very rare in these rocks, and the few which have hitherto been found
belong to the lower vertebrate classes of reptiles, amphibians, and
fishes. Plant remains are more common, and evidence of several succes-
sive floras has been detected. The subdivisions of this great plant
bearing series have been described under a number of local names, of
which the oldest, and best known, are Talchir, Damudal, Mahddeva, and
RAjmahdl, but the term Gondwana has now been adopted by the Geological
Survey for the whole system. This term is derived from the old name
for the countries south of the Narbadé valley, which were formerly Gond?
kingdoms, and now form the Jabalpur, Nagpur, and Chhatisgarh divisions
of the Central Provinces. In this region of Gondwdnathe most complete
sequence of the formations constituting the present rock system is to be
found.

Taken as a whole, the Gondwdna system has a wide extension in the
Indian Peninsula, but in extra-peninsular India, its representatives have
hitherto only been detected in north-western Afghdnist4n and along
the base of the eastern Himélayas in Sikkim, Bhutan, and the Ak& and

} More correctly Talcher and Damodar, but {| ? For the information of non-Indian readers
the spelling in the text has been so universally | it may be well to add that the Gond is one of
used that it is retained when the names are | the principal Dravidian, or so called aborigi-
used in their acquired geological sense. The | nal, tribes, who are believed to have inhabited
more modern and correct spelling is adopted | the country before the advent of the Aryan
when they are used geographically. Hindu race.
GEOLOGY OF INDIA—GONDWANA SYSTEM.

C Chap. VII.

150
\

Daphla hills! Representatives of the highest Gondwdna groups are found
in Cutch, resting upon marine jurassic rocks and capped by neocomian
beds; some rocks containing plant remains, which underlie jurassic
limestones in the desert between the Indus and the Ardvallis, closely
resemble portions of the Gondwana series in lithological characters.
while representatives of beds high in the GondwéAna series, in this case
frequently containing marine fossils, extend down the east coast. But,
with these exceptions, no representatives of the system are found in the
Peninsula north of the valleys of the Narbada ?and Son, nor south-west
of a line drawn from the sea at Masuli patam through Kamamet and
Warangal, north-east of Haidardbdd, till it enters the trap area near
Nirmal. The main areas of Gondwana rocks are in the R4jmahdl hills
and DAmodar valley in Bengal, the Tributary Mahdls of Orissa, Chhatis-
garh, Chutid NAgpur, the upper Son valley, the Sdtpura range south of
the Narbadé valley, and the Godavari basin.

It has already been mentioned that, with the few exceptions noted, the
whole of the Gondwana series is believed to consist of strata deposited in
fresh water, and the only question which arises is whether the beds are
lacustrine or fluviatile.: ‘The coarseness of the rocks in general, the pre-
valence of sandstones, and the frequent occurrence of bands of conglo-
merate, render it improbable that these strata are of lacustrine origin, while
the absence of mollusca almost throughout is, on the whole, rather more
consistent with river than lake deposits, although it is difficult to account
for on either hypothesis. The few fish and reptiles which occur might have
inhabited either lakes or rivers, and the Esther¢xz, which are common in
several subdivisions of the series, might either have lived in lakes or in
the great pools and marshes which often occupy so large an area in broad
river valleys. The plants might have been preserved amongst either
lacustrine or fluviatile deposits, except that it is difficult to conceive the
formation of beds of coal at the bottom of lakes. It is more probable that
the coal originated in marshy forests, such as frequently occur in the valley
plains of rivers. The physical characters of the strata, the frequent alter-
nation of coarse and fine beds, the frequency of current marking on the
finer shales and of oblique lami nation, due to deposition by a current, in
the coarser sandstones, and the circumstance of the upper portions of a
bed, such as a coal seam, being locally worn and denuded when a coarse

1 Mallet, Memoirs, Vol. XI, 14, (1874) ;
Godwin-Austen, four. As, Soc. Beng., XLIV,
pt. ii, 37, (1875); La Touche, Records,
XVIII,121,(1885). Perhaps the occurrence of a
representative of the Rajmahal stratified traps
on the flanks of the Khdsi hills might be
quoted as another instance, but though the
identification of the two sets of beds is highly

probable, it has not been confirmed by the
discovery of fossils.

®? Outcrops have been found north of the river
Narbad4 westward of Hoshangébdd, but far
south of the watershed. The Narbadd, above
the neighbourhood of Jabalpur, runs south of
the general line of division, and Gondwana
rocks occur north of the river.
Chan. VILJ RELATIONS TO OTHER ROCK SYSTEMS. 151

sandstone is deposited upon it, a phenomenon of frequent occurrence,
are quite consistent with the theory of deposition in a river valley, but op-
posed to the conception of lacustrine origin. A river constantly changes its
course, and deposits coarse sediment near its channel and finer materials
from the overflow of its flood waters, the area within which each form of
sediment is deposited varying frequently. In a lake the coarse deposits
must be limited to the margin, and finer sediment accumulates away from
the shore, where there is no current to sweep away the surface of a re-
cently deposited coal or shale bed, and to throw down coarse sand in its
place. On the whole, the evidence is decidedly in favour of a fluviatile
origin for the Gondwana rocks, and it is probable that they were deposited
ina great river valley, or series of river valleys, not unlike those which form
the Indo-Gangetic plains at the present day. ‘There is a possible excep-
tion in the lowest beds of the series, the fine silts which form the base-
ment beds of the Talchir group. These may be of ‘lacustrine origin, but
there is no clear proof that they are, and their remarkably persistent char-
acter throughout an immense tract of country is rather opposed to the
idea of their having been formed in a lake or a series of lakes.

Concerning the relations of this great series to the older and newer
formations in India but little can be said. No older fossiliferous deposits
are known in the area to which the Gondwana rocks are restricted, and
wherever these rest upon any older formation, there is complete uncon-
formity between the two. The areas in which the upper Vindhyan and
Gondwana systems are exposed being distinct from each other, the latter
have nowhere been found in contact with the former, which are the
next series in descending order, but pebbles of upper Vindhyan rocks are
occasionally found in Gondwana rocks.!- ‘Ihe Talchir and Damuda forma-
tions in the country south of Nagpur, on the Godavari below Sironchd, and
in the Mahdnadi valley near Sambalpur, occasionally rest unconformably
upon strata believed to belong to the Cuddapah or lower Vindhyan series,’
but in general the Gondwdna beds are found to have been deposited upon
metamorphic rocks.

On the other hand, the rocks of the Gondw dna series are but rarely
covered at all by a higher formation, except where the Deccan traps.
and their associated infratrappean formation, the Lametd group, rest
unconformably upon the various subdivisions of the Gondwana series, from
the lowest to the highest, in the Narbadd valley and the Nagpur country.
There are, however, localities in India in which sedimentary formations of
cretaceous age rest upon upper Gondwina beds. ‘The first of these is in

1 Manual, ist ed., 205; Memoirs, 1X, 304,| * Supra, p. 91.
1872.)
152 GEOLOGY OF INDIA—GONDWANA SYSTEM. [Chap. VIL

Cutch, where the Umia group, containing some fossil plants, found also
in the uppermost Gondwdna beds in the Narbadd valley, underlies a
stratum containing. Cephalopoda of upper neocomian (Aptian) age. The
second isin the Narbada valley near Bdrwai, where Bagh beds (upper
greensand or cenomanian) rest unconformably on representatives of the
upper Gondwana series. The remaining two localities are near the east
coast. One is in southern India, at Utatir, north of Trichinopoli, where
the plant beds containing RAjmahal fossils underlie the Utatér (ceno-—
manian) group, unconformably in places but elsewhere with apparent
conformity. J.astly near Ellore, where the upper Gondwana beds con-
tain Rajmahdl plants, and marine fossils of upper jurassic age occur
in the higher layers, the age of the strata resting unconformably upon
the Gondwéna strata is not equally well defined. The overlying beds
consist of two fossiliferous bands, one underlying a flow of basalt believed
to belong to the Deccan trap series, the other interstratified between the
lower basaltic flow anda higherone. The igneous beds, like the Deccan
traps elsewhere, are believed to be of uppermost cretaceous or lowest
tertiary age, but the fossils in the upper, or intertrappean, bed differ from
those in the lower, or infratrappean, and it has not hitherto been practicable
to refer either to a definite horizon, Neither bed, however, can be older

than upper cretaceous.

The manner in which the areas of Gondwdna rocks are distributed
throughout the country is peculiar, and there is still some difference of
opinion concerning the interpretation to be placed on their mode of
occurrence. As a general rule, these rocks are found occupying basin
shaped depressions in the older formations, and such depressions some-
times, though not always, nor even generally, correspond to the existing
river valleys. Occasionally the. basins of Gondwana beds are scattered
over the surface of the country, as in Birbhéim, and in this case there
can be no doubt of their representing the undenuded remains of strata
which were once continuous over a much larger area. Whether the
basins now remaining owe their preservation to disturbance of their ori-
ginally horizontal position, and to their having been preserved from denu-
dation through having sunk to a lower level than neighbouring portions of
the same bed, or whether they were originally deposited in hollows in the
older beds, is a point on which opinions differ. ‘There can be no question
that the former is the explanation of these basins having been preserved
in some instances, but cases may also be cited in favour of the latter
view, and it is certain that the Gondwana beds were originally deposited
on an uneven surface,

A few instances will suffice to show the phenomena presented in the
Damodar valley in western Bengal, where some of the most important
Chap. VIL] RELATIONS TO EXISTING RIVER VALLEYS. 153

and best known Gondwéna coal-fields occur. A number of detached
basins are found, all in low ground on the banks of the river, and all
presenting the very remarkable peculiarities that the lowest groups
appear on the northern side of the basin, that there is a general dip from
north to south, and that all are cut off abruptly on the southern edge,
which in most cases is a straight or nearly straight line. Similar geolo-
gical relations exist in many other areas, although the beds are not
always, as in the Ddmodar area, confined to the valley of a single river.
Thus, in the great basin of south Rewd and Sarguja, again in the Satpura
area, and especially in the Tdlcher field in Orissa, the rocks dip from
one side of the basin, and are cut off on the other, but in all these cases
the general dip is north, not south, and the beds are abruptly cut off
along the northern border. The exact directions of the abrupt east and
west boundaries vary, but they are always the samé, or nearly the
same, throughout each tract of country, that is to say, the boundaries of
different fields are parallel to each other, and they are also, as a rule,
identical in direction with the foliation of the underlying gneiss. In
some cases, and especially in the northern part of the great area which
occupies so large a portion of the Godavari valley, both boundaries, which
run nearly north-west to south-east in the last named case, are straight,
nearly parallel, and abrupt.

These abrupt boundaries are almost invariably accompanied by con-
siderable disturbance of the beds in their neighbourhood, In some cases
there is strong evidence that such boundaries are great faults, one of the
best proofs being that the fault occasionally divides, as along the north-
ern edge of the TAlcher field, and beds belonging to the lowest group
are exposed between the different subdivisions of the main dislocation,
the lowest Gondwdna group (the Talchir in the instance mentioned)
being faulted against Kamthi beds, much higher in the Gondwana system,
on one side, and against metamorphics on the other. In some cases, as
along the boundary of the Tdlcher field and also on the eastern portion
of the northern boundary in the Sohdgpur field, the line of fault is marked
by a breccia, containing fragments of the Gondwana sandstones. It is
generally considered that all the fields which are bounded by an abrupt
line cutting them off on one or both sides (and these, as: will be seen,
comprise a very large majority of the basins known) occupy areas of de-
pression, produced subsequently to the deposition of the beds by a fault
along the abrupt boundary, the connection of existing river valleys
with these Gondwdna areas being dependent on the fact that, the Gond-
wana rocks being much softer than the Vindhyan, transition, or meta-
morphic beds upon which they rest, the rivers have worn their way through
the easiest channel,—in short, that the existing drainage, so far as it
coincides with the distribution of the Gondwdna rocks, has been deter-
mined by the disposition of those rocks, produced by disturbance and
154 GECI.OGY OF INDIA~GONDWANA SYSTEM. [Chap. VIL

denudation, and has no necessary connection with their original areas of
deposition.

A different view is held by others. They consider that, with a few ex-
ceptions, there is no sufficient evidence of faulting, that the appearance
of straightness in the boundaries is partly fallacious and due to the rocks
being ill seen at the surface, that the abrupt boundaries are caused by the
deposition of the Gondwana rocks against cliffs forming the original sides
of river valleys, and that the present disposition of the beds is a close
approximation to that of the original areas in which they were deposited.
They consider further that the vertical development of the different groups
varies so much within small distances that there is no reason to believe that
any great thickness of beds abuts against the abrupt cliff like boundaries,
and that there is evidence in some cases that the different groups thin
out towards the margins of the existing basins. They conclude that the
present river valleys differ but little from those which existed in mesozoic
times.

It is possible that there may be some truth in both views. It should
be remembered that the conflict of opinion in this case is between ob-
servers who have chiefly been engaged in mapping widely separated
regions, The view that the present basins closely correspond to ancient
areas of deposition being supported chiefly by observations made in the
Son and Narbadé valleys, and the opposite opinion, that the present Gond-
wAna basins are chiefly due to faulting, being held by geologists who
have especially studied the Gondwdna rocks of Bengal, Orissa, and the
Goddvari valley. The strongest arguments against the existence of
faults along the abrupt boundaries of the various Gondwana fields is
founded on the fact that, in the Sdtpura field to the south of the Narbadd
valley, certain of the uppermost Gondwdna beds overlap the boundary,
but this may be due to the circumstance that the supposed line of fault,
which cuts off the field on the northward throughout the greater portion
of its extent, is more ancient than the topmost groups of the Gondwdna
series. A difficulty in the way of admitting that the abrupt bound-
aries of the Damodar fields are due to deposition against inland cliffs
is to be found in the improbability that all such precipices should be found
on one side of a river valley, while there are some important observations
in favour of the limits of the basins in the Damodar valley being due to
disturbance. Talchir and Damuda beds are found on the Hazdribagh table-
land, immediately north of the Damodar valley, at a height of about 1,0co
feet above the surface of the same rocks in the valley itself, and the
presence of fragments, apparently derived from lower Gondwana beds, in
a conglomerate at a similar or higher elevation on the Chutid Nagpur
highland to the southward points to the former existence of the parent
rock at a still greater elevation. In either case there is evidence of
Chap. VILJ SUBDIVISION OF THE SYSTEM, 155

disturbance for the low level exposures must have been depressed, or the
high level ones elevated ; in other words, the Gondwdnas must have under-
gone disturbance since they were deposited, and this disturbance cannot
have been without effect on the present limitation of the outcrops,

The tracts of country occupied by rocks of the Gondwdna series are,
as a rule, covered with a poor sandy soil and ill suited for cultivation. The
result is that, in many parts of India, they form wild uninhabited forests.
Such tracts are always the last to be surveyed topographically, and, as a
rule, minor details are omitted onthe maps prepared. Moreover the upper
Gondwana recks are principally sandstones and decompose readily into
loose sand, which covers the whole surface of the country and greauy
conceals the rocks. These two-circumstances—deficiency of maps and
concealment of the surface—have combined to delay the geological sur-
vey of the upper Gondwdna formations, and to render the examunation
of the beds exceptionally tedious and difficult.

The groups of which the Gondwana system is composed vary greatly,
both in number and mineral character, in the several isolated areas in
which they are found, the variation being much greater amongst the middle
and upper than amongst the lower members of the series, The two
lowest Gondwana groups, the ‘Valchir and Barakar, which consist largely
of shales, whilst the uppermost formations are chiefly composed of coarse
sandstone, grit, and conglomerate, preserve their mineral character almost
unchanged throughout the area in which the lower Gondwana Leds are
known to occur,

The system may be divided into an upper and a lower series, the.dis-
tinction having been first established in western Bengal, where it is of a
most trenchant nature, characterised by a marked stratigraphical discord-
ance, by an almost complete absence of any species common to the
two divisions, and an utter change in the type of the flora, equisetaceous
plants prevailing in the lower subdivision, and cycads and conifers in
the upper,! ferns being found commonly in both. Some Lgudsetacex
occur, however, in the upper Gondwdnas, and several species of cycads
and conifers in the lower, but the genera are in most cases distinct in
the two subdivisions. As the examination of the Gondwdna system in
the Sdtpura ranges and in south Rewé has progressed, it has been found
that the stratigraphical break there is not nearly so marked, and it is pos-
sible that a number of distinct floras will ultimately be found, bridging over
the gap in western Bengal.

'The Mahddeva series has, however, hitherto proved almost unfossiliferous.
 

(Ohap. VIL

 

 

 

“HRUSIEL |*  aysyeQ* oaryspen | * rysey | + amysjer|* — sryorey,
: 5 . ° ayoyey
“pqueysey | * uyqreqreyy | * pqreysey | * eqreysey oe
‘Teyeieg }* seyeiwg |* seyeieg|* aeyereg {+ aeyeieg|+ seqeieg }
*sareys : fs *VNYMGNOY)
s 7 INO," ene auojsuoly nas aaa > Featured. YaMO'T
ta! “us \
a -up Sur soft | +(aySurq7) )| + fueSjury me e
e cpnyour) rw
rawyy \
s *  gpowly si *  yeyourg = ae 7" $8 yaqourg
<
z
a : ft
Zz. ‘eppaepng + andfeiqnq
5 pue dey
{ “05 ‘yes yIV es Fyzeuryoeg * BASPPURT " * yewley Yor
. “FUR pue [pyeutiey
s oe coord) om |
a pue 49
Zz sradtig ‘ures see ase see
= -udearsey | 1gyey-ejoyy | ° semua ; Lo gata
a ¢ Npaaryes . : adap)
. pues ainyjna vise
o “vd ‘Anodiy|* erry |* andjeqef[ |*  sndyeqef “ ae = )
S j ‘andjeqe? pue ww
Pps} ew ace ee ove eee ane see a
5 ‘
&
O ‘
. *sfaqyea |,
¥ aa *u013 Aayea *u018 ae eer Lyqeyey pue ts
ee spe emdy¥S—"A pu peur repowyq—111 sean ees reeutlyt—y “AONANOES TVYINTD
a L

 

 

 

 

 

 

 

 

 

"4430 yova AQ sdnosd vupmpuon ay) fo uorvjuasasgax 31qUvq04¢ ay] Surmoys ajgn yz
Ohap. VIL] TALCHIR GROUP, 137

].—THE LOWER GONDWANA SERIES.

The lowest member of the Gondwdha system is known as the Talchir
group, thus named from its having been first clearly distinguished in the
small district of Télcher? one of the tributary mahdls of Orissa. When
present—and it is rarely absent over a large area—this group forms the
base of the GondwdAna series, and consists in general of fine silty shales
and fine soft sandstone. The shales are usually of a greenish grey or olive
colour, sometimes slaty. .They are of exceedingly fine texture, traversed
by innumerable joints, and break up into minute, thin, angular frag-
ments, sometimes elongate or acicular, which cover the surface of the ground
in places. Occasionally the shales have a dull Indian red colour, but this is
not common. They are frequently mentioned in the Survey reports under
the name of mudstones and needleshales. Not unfrequently they are some-
what calcareous, and in some places large concretionary massés of impure
carbonate of lime bave been found amongst them.

The most characteristic sandstones are soft, fine, and homogeneous
in texture, composed chiefly of quartz and uudecomposed pink felspar, and
in colour pale greenish grey, buff, or pale pinkish, almost of a flesh tint.
They are frequently rather massive, though distinctly stratified, but they are
also commonly interstratified in thin layers with the shales. In many
places they break up, where exposed on the surface, into polygonal frag-
ments, three or four inches across, whence they have been called tesselated
sandstones.

These beds pass into coarser sandstones of less marked character,
which vary in colour, and are sometimes, though rarely, conglomeratic.
It is an almost invariable rule, contrary to what is found to be the case
in most rocks, that in the Talchir group the beds of finest texture, the,
shales, are found at the base, and that the sandstones are higher im.
position, the coarser sandstones, moreover, overlying those of finer texture.,
A thin coal seam has been found amongst the Talchir beds in the
Jhilmilli field, in Sargdja*, but this formation is, as a rule, distinguished
by the absence of coal seams, and even of carbonaceous shale. 9

There are three peculiarities of the Talchir group which still require
notice, as all of them are of considerable importance.

The first is the frequent occurrence, amongst the shales and fine sand-
stones, generally towards the base of the group, but very frequently some
hundreds of feet above the bottom, of pebbles and boulders, always rolled and
usually well rounded, varying in size from small fragments quarter of an inch
or an inch across to huge blocks fifteen feet in diameter and thirty tons in
weight, fragments from six inches to three feet in diameter being common,

1 Memoirs, \, 46, (1856). | * Manual, Ist ed., 205.
158 GEOLOGY OF INDIA—GONDWANA SYS5TEM. (Chap. VIL

The distribution of the boulders is most iiregular, in some parts of the area
occupied by Talchir beds none are to be found over many square miles of
country, but generally some are met with at intervals, and occasionally
large numbers occur within a limited tract.

In very many instances there is every probability that the boulders
have been transported from a distance, no rocks of similar character
being found in the neighbourhood. If only one or two such cases had been
observed, it might be supposed that the rock, from which the blocks were
derived, had formerly existed in the immediate vicinity and been
removed by denudation, but the cases in which there is reason to believe
that the rounded blocks have been transported from afar are so numerous
that this theory cannot be accepted. The boulders, it should be remembered,
are frequently found imbedded in the finest silt, It is evident that deposi-
tion from water in rapid motion is here out of the question, as any stream
which could have moved and rounded the boulders would have swept
away the silty matrix in which they are deposited, and the only
suggestion, as to the cause of their occurrence, which appears to account
satisfactorily for their presence, is to suppose that they were originally
rounded by torrents and then transported to their final position. by
ice. This theory has received strong confirmation from the discovery of
smoothed and scratched surfaces on some of the large boulders found on
the banks of the Pengangd river, about ten miles west-south-west of
Chanda, Central Provinces.'. The surface of the limestone rock under-
lying the Talchirs was also in this case found to be polished, scratched,
and grooved. ;

‘The second peculiarity is the remarkable resemblance to a volcanic
rock occasionally presented by the more compact forms of shale, and by a
variety of the sandstone. So great is the similarity between the shale-and
a consolidated volcanic ash that two experienced surveyors have, at different
times, marked the beds as trappean, whilst the sandstone occasionally
simulates a decomposed basalt in colour and mode of weathering.

The third noteworthy feature of the Talchir beds is their pawer. of
resisting disintegration, and the entire barrenness, provided they are not
covered by alluvial deposits derived from other rocks, of the ground
where they appear at the surface, a natural consequence of their not
decomposing to form soil. In many places along the edges of the coal-
fields, where the Talchir beds occupy the ground, it is possible to walk for
miles through very thin jungles, free from grass, over a surface composed
entirely of the finely comminuted greenish grey shales, .

South of the Pengangé river a peculiar rock was found by Mr. Fedden
in the Talchirs near Charli, and again in the Khairgaon nala west of

1T, Oldham, Memoirs, 1X, 324, (1872); Fedden, Records, VIII, 16, (1875).
ia heterophylla, Bat.

 

Volt

       

teris valida, H’stm.

Jeurop'

N.

 

 

Gangamopteris cyclopteroides, Fstm,

Nosggerathiopsis hislop, Bunb,

Gangamepteris major, F'stm,

Caleutta Phototype Co,

TALCHIR KARHARBARI PLANTS.
Chap. VILJ TALCIIR GROUP. 159

Nandgaon. It is a calcareous sandstone, whose calcareous cement has
assumed the form of ophitic crystalline masses quarter of an inch across.
These crystals have been irregularly attacked in the weathering and the
rock split up into a number of irregular aggregates of quartz grains,
separate from each other, but held together by the interlocking of their
irregularities. The rock, as a whole, has consequently a certain amount of
flexibility analagous to that of what is known as itacolumite.!

The Talchirs preserve all their peculiarities throughout the area in which
they occur,—an enormous tract of country, extending from the flanks of the
Rajmahdl hills to the Godavari, and from the Raniganj field on the borders
of the alluvium of Lower Bengal to the neighbourhood of Hoshangdbdd,
Nagpur, and Chanda.

The thickness of the Talchirs nowhere appears to exceed about 8v0
feet, their extreme measurement where fully developed in part of the
Ra&niganj coal-field.

The fossils? hitherto discovered in the Talchir rocks are very few in
number. Of animal remains only the wing of a neuropterous insect and
some annelid tracks have been discovered, whilst the plant remains consist
of Noeggerathtopsis hislopi and three ferns, Gangamopteris cycloptcroides,
G. angustifolia, and Glossopterts communis. The only evidence of
vegetable life hitherto found has been in the higher beds of the group,
and there is a remarkable absence of plants in the lower shales, which are
admirably suited for preserving vegetable impressions. Even in the upper
beds of the group fossils are of singularly rare occurrence.

Reference has already been made to the possibility of a lacustrine ori-
gin for the Talchir beds, or at least for the lower portion, The chief rea-
son for suggesting that these beds may have been deposited in lakes is
the great thickness of very fine sediment accumulated at the base of the
group, and the very frequent occurrence of much finer beds below than
above. ‘The latter, on the hypothesis of a lacustrine origin, may be ex-
plained by the gradual silting up of a lake basin, in which fine sediment
would be deposited at a distance from the margin, whilst coarser beds
would be thrown down by rivers as their deltas advanced into the lake
and filled it up. This evidence, however, is quite insufficient by itself to -
prove that the Talchirs are a lacustrine deposit, and it is at least equally
probable that they were formed ina river valley, like the overlying members _
of the Gondwdna system.

At the same time the large size of the boulders and their generally

1 Memoirs, XIII, 16, (1877). Compare the | longing to this and other groups of the
K4lidna flexible sandstone (supra, p. 72); see | Gondwana system ,are taken from Dr. Feis-
also, for a more detailed description, Records, | mante]’s descriptions in the Pal, Ind., series
XXII, 54, (1889). ii, xi, xli, (1877-86).

2 The determinations of fossil plants be-
ene“ SKOLOCY OF INDIA—GONDWANA SYSTEM. [Ohap. VIL

rounded aspect suggest that they were produced by rapid flowing streams,
whose beds hada steep gradient. The great unconformity between the
Talchirs and the underlying rocks points to a long continuance of dry
land conditions, unfavourable to the accumulation of sediment, and one of
the first effects of those land movements which caused its accumulation
might well be to split up the river valleys into large lake basins, and
steep stream valleys leading into them, Were this the case, all the con-
ditions essential to the explanation of the features ordinarily exhibited
by the Talchir boulder bed would be existent, if winter ice were super-

added.
This explanation is not, however, sufficient to account for the smoothed,

polished and striated surfaces of the fragments included in the Talchir
conglomerate, and of the underlying Vindhyan limestones in the Pen-
gangé valley, which appear to be due to the action of a true glacier. The
boulder beds largely developed near Bap, in western R4jputdna, which can
hardly be other than of Talchir age, being unconformably superimposed
on the Vindhyan limestones and older than the upper Gondwana beds
of this district, contain numerous well glaciated fragments, and a similar
boulder bed near Pokaran is seen to rest on a surface of older rock, which
is not only smoothed and striated, but exhibits typical roches moutonnées.
In the last named instance there is some possibility that the boulder bed is
older than Talchir,? but excluding this, there is evidence enough that
glaciers must have descended to low levels in Talchir times.

The coal bearing rocks of the Karharbdri coalfield were originally
assigned to the Barakar group in the publications of the Geological
Survey, on account of their mineral character and their position imme-
diately above the Talchir beds, The examination of the Karharbdri
fossil flora has, however, shown that, whilst all the species known to be
found in the Talchir beds are represented, one of them (Gangamopteris
eyclopteroides) being the commonest fossil of the Karharbéri beds, many
of the common Damuda fossils are rare or wanting, and several very re- _
markable species are found which have not hitherto been detected in the
Damuda series. The peculiar excellence of the coal, and its superiority to
that obtained from the majority of the Damuda seams, have led to exten- '
sive mining operations in the Karharbéri field, and it has consequently been
possible to obtain good collections of the fossil plants.® It has also been
noticed that the coal of Karharbari differs in structure from that of the
Damuda series generally, and a partial re-examination of the field appears
to justify the inference that there is also a slight distinction between the

? W. T. Blanford, Records, XX, 49, (1887). I. J. Whitty, Superintendent of the East
? Supra, p. 106, Indian Railway Company's collieries at
* These have been chiefly collected by Mr. Karharbari.
Chap. VIL] KARHARBARI GROUP. 161

Karharbdri and Barakar sandstones, although it isas yet uncertain whether
a passage may not eventually be found between the Karharbéri group
and the Barakars. The paleontological evidence hitherto obtained tends,
however, to connect the former with the Talchir group, and it appears
best, for the present, to keep the Karharbdri rocks distinct from the
overlying Damida series, under the name of the coalfield in which they
were first distinguished. ~

‘The rocks of the Karharb4ri group consist almost solely of sandstones,
grits, and conglomerates, with seams of coal. Very little shale occurs, the
little which exists being associated with the coal seams. The sandstones
are mostly white, grey, or brown, and felspathic, often gritty and conglo-
meratic, from containing large fragments of felspar and pebbles of quartz.
The chief distinction between the constituents of the grits and conglo-
merates forming the Karharbdri group, and those which make up so large
a portion of the Barakars, is that in the former, and especially in the
coarser grits and conglomerates, a large proportion of the fragments of
felspar and quartz are angular or subangular, whereas in the Barakars the
pebbles are, as a rule, particularly well rounded. The coal of Karharbéri
is rather dull coloured and tolerably homogeneous in structure, the layers
of very bright jetty coal, which are so conspicuous in the Damuda seams,
being in general few and ill marked. The seams appear to be somewhat
variable in thickness, but to undergo very little change in composition
throughout the small field in which they are found. Some of the seams,
both in the Barakar and Rdniganj subdivisions of the Damuda series,
furnish fuel equal in quality to that extracted at Karharbdri, but they are
much more distinctly laminated.

The Karharbdri beds rest with apparent conformity on the ‘falchirs, but
the former completely overlap the latter in places, within the limits of the
little Karharbdri field, and the mineral characters of the two groups are
strongly contrasted. In the west of the Karharbdri basin the Talchirs
attain a thickness of about 500 or 600 feet, whilst, within a distance of less
than four miles to the eastward, the Karharbdari beds rest upon the gneiss.
It is probable that the highest rocks seen within the coalfield may be of
Barakar age, and there is some slight appearance of the Karharbdri beds
being overlapped by these higher strata, but the overlap is notclear. The
whole thickness of the Karharbdri group is probably about 500 feet.

Outside the limits of the Karharbdri field the Karharbdri group has
been recognised, on palzontological grounds, at Mohpdni in the Narbadé
valley, in the Ddltonganj coalfield and, with considerable degree of proba-
bility, in Hutar and Rew4. The coal in a seam, lying very little above
the Talchir group, in the Raniganj field resembles the Karharbdri coal in
mineral character, and it is probable that this group will be found to have
a wider distribution than is now known, and to be represented in all those

M
162 GEOI.OGY OF INDIA—GONDWANA SYSTEM. (Chap, VIL

sections where the Talchirs are described as conformable to the beds above
them.

The only fossils which have-so far been formed in the Karharb4ri beds
are plants of which the following is a list?; those species distinguished
by an asterisk having also been found in the Damuda beds, while a dagger
marks the forms which have been found in the underlying Talchirs :—

EquiseTacER— FILICES,—contd,
Schigoneura, cf. meriani. Sagenopterts (2) stoli skana.
* Vertebraria indica.
eae CycaDEACcEz-—
ene evi Jobhavonih Glossozanites stolicgkanus.
eames, ae tee tVeggerathiopsis hislupi.
23 (?, buriadrea.
: ConIreRE—
i major. yg:
obliqua. : t Eurypyllum whittianum.
+ 9 angustifolin. Voltata heterophylla.
a cf. spathubata. | Albertia, sp.
44Glossopteris communis. |} SEEDS—
* a damudica. Samaropsis, sp.
a decipiens. - Carvdiocarpum, sp.
* es andica, Carpchithes millert.

 

The Talchir-Karharbari groups are succeeded by a great series of beds,
the Damuda series, which was first examined and described in the
coalfields of the Ddmodar valley. Nearly all the coalftelds of the Indian
Peninsula owe their mineral wealth to the presence of these beds, the
Karharbdri being the only other important coal bearing group,. and the
quantity of valuable minerals contained in the roeks of the Damuda series
is probably greater than that of all the other rock groups of India com-
bined.

The Damuda series in Bengal has been found to consist of three
subdivisions, known in ascending order as the Barakar group, Ironstone
shales, and Rdniganj beds. The first and lowest is also found in the
Son, Mahdnadi, Narbad4, and Godavari valleys, the upper subdivisions
being represented by groups differing in mineral character from the
Bengal beds.. In the Sdtpura area the Damuda subdivisions are known
as the Barakar, Motér, and Bijeri groups, and in the Godavari valley,
above the Barakar group, there also the only coal bearing formation, a
single member of the upper Damuda beds occurs, and is known as the
Kdmthi group. A similar arrangement prevails in the Mahdnadf and
Brahmani area, only two Damuda subdivisions being found, which appear
to correspond to those of the Godavari region.

The mineral characters and geological relatiens of all these different
groups must be described separately, It is sufficient for the purpose at

,

! Pol, Ind., series xii, IIT, pt. i, (1879-81); 1V, pt. 2, (1882).
 

Spbenopteris polymorpha, F'stm.

 

 

Stem of Schizoneura écndwanensis,

 

Schizoneura gondwanensis, F’stm,

Phyllotheca indica, Bunb, Vertebruria indica, Royle.

Calcutta Phototype Co

DAMUDA PLANTS.
   

Alethopteris jindleyana, Royle.

 

Glossopteris angustifolia, Bet

Gicasepteris communis, F'stm. Glossopteris conspicua, F'stm,

Calcutta Phototype Co.
DAMUDA PLANTS,
DAMUDA SERIES,

Chap. VILJ

present to note that all consist of sandstones and shales, with more or
less ferruginous bands, and that some contain coal. Slight unconformit
between the different groups has been noticed in places, and the Bees
beds are frequently unconformable to the Talchirs. The whole thickness
of the Damuda series is 8,400 feet in the RAniganj field, and about 10,000
feet in the Sdtpura basin. It thus constitutes the most important portion
of the Gondwana system.

‘The examination of the floras of the different groups of the Damuda
series has shown that there are slight differences, but as nearly all the
species of the Barakar group, and without exception all the more abund-
ant forms, are also found in the RAniganj group, they have been united in
one list given below. The letters /., m., u. prefixed signify that the
species have been found in the Lower, Middle, and Upper, or Barakar
Ironstone or Raniganj beds, respectively ; a dagger that it is also known
from beds of Karharbdéri age, and an asterisk that it ranges into rocks
newer than the Damuda series :—

163

EaulSETACE& — m. u Glossopterts vetifera.
* +1. u. Schigoneura gondwanensts. m. u. 55 conspicua.
u. Phyllotheca indica. u. " ingens.
uP ty rotusta. u. ra divergens.
|. u. Trisygia(Sphenophylium) spe-|*tl.moue 3 damudiva.
ctiosa. * Lm.u. » angustifolia.
* +1. u. Vevtebraria indica. u. 5 leptoneuru.
u i Formosa,
Firices— u. ” orbicularts.
1. Cyathea, cf. tehihatcheffi. u. Gungamopteris anthrophyoides.
lu. Sphenopteris polymorpha. u. r whittiana.
u. Dicksonia hughes. u. a5 hughest.
u. Alethopteris whithyensis. flim , eyclopteroides.
u. 33 lindleyana. u. Belemnofteris wood-masontana.
u. 35 phegopteroides. u. Anthrophyopsts, sp.
u. Pecopteris affints. lu. Dictyopteridium, sp.
u. Merianopteris major. u. Sugenopteris (?) longifelia.
*}m. u. Macroteniopteris daneotdes. u, 3 polyphylla.
u. 5 feddeni. u. Actinopteris bengalensis.
u. Paleovittaria kurzi. Cyslencre—
ed ne ae ee anne u. Pterophyllum burdwanense.
Sa l. Platypterigium balli.
i r6., SONI 1 N rathiopsis hislopi
* +1. m.u. Glossopteris communis. Filth wy Bh OeE ze mee Hinge
L 3 intermittens. ConIFERE—
U4 stricta. + wu. Voltsia heterophylla.
u, - ? musefolia. u. Rhipidiopsis densinervis.
*TlLomvu. es indica, 1, 5 gingkoides.
La. 7 browntant. 1. Cyclopitys dichotoma.
lu. A inter media. #t 1, u. Samaropsis, ch. parvula.

 

iii, (1880-81), IV, pt. ii, (1889).
164 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL.

Although there is little difference between the floras found in the
various subdivisions of the Damuda series, the characters and relations
of the minor groups require separate notice, and of these groups the
lowest and. the most important is the Barakar. This group derives
its name from a river which traverses the western portion of the Rani-
ganj coalfield, and then falls into the Daémodar within the limits of
the field.!. In the higher portion of its course the Barakhar river receives
the streams which drain the Karharbéri coalfield.

The Barakars have an equally extensive range with the Talchirs, and
consist of conglomerates, sandstones of various kinds, shales and coal.
The sandstones are often coarse and felspathic, a variety of frequent
occurrence being rather massive, white or pale brown in colour, soft at
the surface and not much harder below, consisting of grains of. quartz
and decomposed felspar. ‘The weathered surface of this sandstone fre-
quently exhibits small projecting knobs, apparently due to calcareous
concretions. One of the most striking distinctions between the sandstones
of the Talchirs and those of the overlying formations consists in the
felspathic constituents of the former being, as a rule, undecomposed, while
in the Damuda series the grains of felspar are almost invariably converted
into kaolin.

Besides the whitish felspathic sandstone, another typical Barakar rock
is a conglomerate of small, well rounded, white quartz pebbles. These
are sometimes found scattered over the surface and serve to indicate the
presence of the conglomerate, where it is not exposed in section. The
matrix ot the conglomerate is usually white sandstone,

It must not be supposed that white is the only colour of the Barakar
sandstones. Brown, red, yellow, and other tints are to be found, and pre-
dominate in many places. The whitish felspathic sandstone is however a
typical rock, preserving its character in localities as far apart as Ranfganj
in Bengal and Chanda in the Central Provinces, being well developed
in the Godavari valley, but it is subordinate and forms but a small portion
of the group to the eastward. Here the greater portion of the Barakar
rocks consists of shales, grey, blue or black, frequently micaceous, and
more or less sandy, occasionally associated with argillaceous iron ore, and
often containing seams of coal. Not unfrequently the shaly beds are
interstratified with hard flags.

The coals of the Barakar group vary greatly in quality and character in
the different coalfields. They all, however, agree in having a peculiar
laminated appearance, due to their being composed of alternating layers
of bright and dull coal, the former purer and more bituminous than the
latter, which, in many cases, is shale rather than coal. The best coals are

1 Memoirs, Ul, 212, (1863).
hap. VILJ BARAKAR GROUP, 165

those in which the bright layers predominate, but nearly all seams hitherto
discovered are somewhat inferior to average European coal of the carboni-
ferous formation, and there is a yeneral tendency to variation in the thick-
ness and quality of each seam within short distances. At the same time
excellent fuel has been obtained from some Barakar seams. Some coal
beds are of immense thickness, single seams (including partings of shale)
amounting to as much as 35 feet in the Raniganj coalfield, 50 feet near
Chand4, and no less than go feet at Korba in Bildspur. Some of the
Barakar coal exhibits a peculiar spheroidal structure, and round balls of
various sizes, up to more than a foot in diameter, break away from the
mass when the coal is mined. So thoroughly are these rounded that
they were taken at first for rolled fragments, derived from some older
formation.?

In places the Barakars rest quite conformably upon the Talchirs, and
the two groups appear to pass into each other. In general, there is an
abrupt change in mineral character, but the only case which has hitherto
been found in which there is clear evidence of denudation having re noved
portions of the lower beds, during the depusition of the higher group,? is in
the Rdmgarh coalfield, where rolled fragments derived from the Talchirs
have been found in the beds of the Barakar group. The Barakars, however,
overlap the underlying Talchirs in many places and rest upon the metamor-
phic rocks, and in some coalfields, as in that of Rdnfganj, there appears to
be overstep as well, the highest beds of the Talchirs disappearing first, as if
they had suffered from denudation. It yet remai1s to be seen whether
representatives of the Karharbdri beds do not intervene in those cases
in which there is an apparent passage between the Barakar and Talchir
groups.

The Barakars appear nowhere to exceed the thickness of 3,300 feet, a
development which they attain only, so far as is known, in the Jharid field.
In no other field, except RAmgarh, do they exceed 2,0c0 feet.

Above the Barakar group in the Rdniganj and a few other fields
of the Damodar valley, there is found a great thickness of black or grey
shales,? with bands and nodules of clay ironstone (carbonate of iron,
mixed with clay), some of which is of the carbonaceous variety known
as black band. Towards the base these beds become more sandy, and
interstratifications of sandstone occur amongst them. The shales disinte+
grate slcwly, and consequently the tract covered by this group is barren,
and frequently elevated, but the rocks are not as a rule well exposed on
the surface, although their presence is indicated by fragments of ironstone
being scattered about.

‘See Four. As. Soc. Beng. XVII, 59, (1848); 2 Yemoirs, V1, 113, (1867).
XVII, 412, (1844); XIX, 75, (1850). 3 Memoirs, 1, 40, (1869).
166 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL.

The greatest thickness attained by the ironstone shales is about 1,500
feet in the Bokdro coalfield, and they are nearly as thiek in the R anigan)
field. As arule, they are quite conformable to the underlying Barakars,
the slight unconformity, which has been observed in places, is very possibly
local, but one case! has been noticed where a break in time may be indicated.

Fossils are not common, and most of the species recorded were ob-
tained from the South Karanpurd coalfield,

The highest group of the Damuda series, in the Damodar valley,
derives its name ‘of Raniganj from the principal town of the mining
district of Bardwd4n, and comprises a great thickness of coarse and fine
sandstones, with shales and coal seams.? The sandstones are moderately
coarse, as a rule in thick massive beds, white or brown in colour, and
obliquely laminated. They are usually more or less felspathic, the felspar
being converted into kaolin. Bands of rather calcareous, fine, hard, yellow
sandstone, often weathering out at the surface in nodular fragments, are
common and characteristic of the group. Conglomerates are of rare
occurrence. Shales form a much smaller portion of this group than they
do in the Damodar area of the subjacent Barakars. They are sometimes
black and carbonaceous, sometimes bluish grey, and occasionally red or
brown, more or less mixed with sand or stained by iron, and small bands
of argillaceous ironstone occasionally eccur, though they are not common.
The coal is composed of alternately bright and dull layers, as in the
Barakars,

This group is of considerable thickness in the Rdniganj field, being
as much as 5,000 feet from top to bottom where fully developed, and it
is possible that this is less than the original thickness, for the next group
in ascending order rests upon the denuded surface of the present. The
Raniganj group diminishes in. thickness in the other fields to the west-
ward, and appears to be represented by groups of different mineral charac-
ter beyond the limits of the Damodar drainage.

As a general rule, the Raniganj beds are conformable to the ironstone
shales, but the higher group oversteps the lower, and rests on the Bara-
kars, in the Bokdro coalfield, near Hazaribagh.

No animal remains have been found in the rocks of this group, but
plants are abundant and comprise nearly all those in the Barakar groups,
besides a number of species that are not known from any lower horizon.

The lithological distinction of the threefold division of the Damuda
series, and the overlying PAnchet group, which will be noticed further on,
has only been recognised with certainty in the coalfields of the DAmodar

© Memoirs, II, 42, (1853). I * Memoirs, Ill, 46, (1863).
Chap. VILJ MOICR AND BIJORT GROUPS. 167

valley. The Radniganj and Pdnchet groups have been recognised, palzon-
tologically, in south Rewd, though they have not been mapped, but in the
Mahanadi and Godavari drainage areas the Barakars are overlaid by a
great series of beds which have been described in different areas by
various names.

In the Sdtpura ranges, south of the Narbadd, the lower Gondwdna
beds, above the Barakars, have been divided into two groups,—the Mottr
and Bijori. This area has not had the same attention paid to it as that of
Bengal, the Godavari valley, and Orissa, and the classification of the beds
above the Barakars must be regarded as purely provisional.

The Motiér group? derives its name from a village of that name
situated about 12 miles south-south-east of Pachmarhi, on the dividing
ridge between the valleys of the Denwa which runs into the T4wa,
a tributary of the Narbadd4, and the Kanhdn, which is a tributary of
the Godavari. The village is on the road from Badntr and Chhind-
-wdrd to Pachmarhi, and was at one time used as a sanitarium,

The beds of this group somewhat resemble the Pdnchets of Bengal in
mineral character. They consist of thick, coarse, soft, earthy sandstones,
grey and brown, sometimes with red and mottled clays and calcareous
nodules. Shales occur, but they are usually sandy and very rarely carbona-
ceous. It is probable that the Motir group is unconformable to the Bara-
kars. No collections of fossils have hitherto been made from the beds of
the Mottr horizon.

The highest members of the Damuda series in the Sdtpura region
are exposed in the upper Denwa valley, at the southern base of the
Mahadeva or Pachmarhi hills. For the rocks of this horizon the name
of Bijori has been proposed,* from a small village rendered famous
by being the locality whence the only distinctly vertebrate fossil, except
‘Brachiops, yet obtained from the Damuda series, was procured.

The rocks of the Bijori horizon are characteristically Damudas, and
comprise shales, occasionally carbonaceous, micaceous flags and sand-
stones.

Nothing definite is known of the relations between the Bijori and Mottr
groups, nor has the thickness of either been determined, but the greater
portion of the 3,000 to 4,000 feet of beds, intervening between the
Motir beds and the base of the Pachmarhi sandstone, may be assigned
to the Bijori group.

The most important fossil hitherto found in the Bijori beds is the
specimen already referred to, which is the skeleton of a Labyrinthodont
allied to Avchegosaurus, described by Mr. Lydekker under the name oi
Gondwanosaurus bijoriensés,®

1 Memoirs, X, 161, (1873). 3 Pal, Indica, series iv, I, pt. 4, (1885).
° Memoirs, X, 15¢, (1873). |
168 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL

Besides the labyrinthodont, the following plants have been identi-
fied:—?

EquiIsEeTACcEE— Fiiices,--contd.
Schizoneura gondwanensis. Glossopieris damudiea.
Vertebraria indica. * vetifera.
Trizygia speciosa. as angustifolia.

Fitices— Gangamopteris, sp.
Dicksonia, sp. -CONIFFRE—

Glossopteris communis. Samar opsts, cf. parvula.

The general facies of this flora corresponds best with that of the
R4niganj group in Bengal, with which it may be correlated in a general
way, as long.as exact contemporaneity of origin is not asserted.

In the God4vari valley, and in Chhatisgarh and western Orissa, the
beds which overlie the Barakar group have been described under the
names of Kamthf and Hingir, respectively, but in spite of some minera-
logical differences, the two seem to represent each other so closely that
they may be united under the first mentioned and older name.

The name Kdmthi is derived from the military station so called,
twelve miles north-east of Nagpur, and the station again derives its name
from a village on the opposite side of the Kanh4n river, where there is a
famous quarry which has yielded a large number of fossils. The term
Hingir is derived from a zaminddri of that name situated north of
Sambalpur.®

The typical Kdmthf rocks consist of conglomerates, grits, sand-
stones, shales, and clays. The conglomerates contain pebbles of quartz.
The grits are sometimes hard and silicious, so much so as to be quarried
for quernstones, but usually they are soft and argillaceous. They
are frequently stained by iron, and are often intersected by hard fer.
ruginous bands of a dark brown colour. The sandstones are of every
shade of colour, and vary greatly in character. They comprise fine grained
micaceous beds, white in colour, with blotches and irregular streaks of red,
and one of the most characteristic beds of the formation is a very fine
argillaceous sandstone, hard, massive, and homogeneous, resembling a
shale in structure, except that it exhibits no trace of lamination, yellow in
colour below the surface, but becoming red when exposed. It passes into
red shale. Another characteristic bed is a hard grey grit or sandstone,
ringing under the hammer and breaking with a conchoidal fracture. ‘The
clays are red or green in colour, and chiefly prevail in the upper portions
of the group.

These typical beds, with the exception of the clays, are chiefly de-
veloped near Nagpur. Elsewhere the Kamthis consist mainly of soft, porous

1 Pal, Indica, series xii, HI, pt. ti, 17, (1880). 3 Records, VIE, 112, (1875).
2 Records, 1V, 50, (1871); Memoirs, 1X, 305, (1872).
Chap. VILJ KAMTHi GROUP, 169

sandstone, brown or white in colour, and conglomeratic in places, often
with hard, ferruginous bands, and a few red shales. Here and there,
however, a band of one of the characteristic rocks is met with towards the
base of the formation.

The chief peculiarity, which distinguishes the K4mthf group from the
RdAnfganj and Bijori groups. is the absence of carbonaceous markings. In
other Damuda groups, with the exception of the ironstone shales, the
remains of plants generally retain a portion of their original carbon, but
this appears very rarely to be the case amongst the Kamthis.

The thickness of the Kdmthi group has not been determined, but it is

-undoubtedly considerable, probably 5.000 to 6,000 feet at least. The beds

belonging to this group generally appear conformable to the Barakars,
but it is extremely doubtful if the conformity is more than apparent, for
the KAmth{ beds overlap the Barakars in a most irregular manner, and
the break in conformity between the two is well marked in places, The
Hingir beds, both near Sambalpur and in the Tdlcher coalfield, certainly
rest unconformably in places on the Barakar group.!

The fossil plants of the Kamthi group comprise the following species :?

D. Phyllotheca indica. Glossopteris musefolia.
D. Vertebraria indica. »  leptoneura.
Pecopterts, sp. Gangumopteris hughest.
D. P. Glossopteris communis. Angiopteridium, cf. macclellandi.
D.P. » indica. D. Macroteniopteris daneoides.
D. ss browniana. D. ss feddeni.
D. » damudica. Neggerathiopsis hislopi.
as stricta.

In the foregoing list the letter D prefixed to the name of a species
signifies that it is also found ,in the Damuda series of Bengal, and the
letter P that the same species is known from the Pdnchet group. Of the
former, all are found in the RAnfganj group, but not alJ in the lower groups,
one species Angiopteridium macclellandi has been found in the RAjmahal
group of the upper Gondwdnas. The character of the flora would lead us
to regard it as homotaxial with the Rdnfganj group or possibly newer.

In the neighbourhood of Méngli, a small deserted village lying at
the northern extremity of the Wardha Gondwéna basin, about fifty miles
south of Nagpur and thirty-five north-west of Chandd, some quarries have
long existed, from which a very fine red and yellow sandstone is obtained
and employed in building, chiefly for ornamental purposes and for carv-
ings. The stone is precisely similar to that of Silew4da and other
typical exposures of the Kémthi group, near Nagpur, and the coarser

associated sandstones of MAngli differ in no way from the ordinary Kdmthi
grits.

? Records, VIII, 113, (1875). ® Pal. Indica, series xii, ItI_ pt. ii, 19, (1880).
170 GLOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL.

The quarries of Mdngli have become well known by name to Indian
geologists, and even to those of other countries, having furnished to
Mr. Hislop the first Labyrinthodont amphibian fossil (47achiops laticeps)
detected in India! They have -also yielded a species of Estherta anda
few plant remains. ‘he latter are so poor that very little dependence can
be placed upon their determination, One is believed to be coniferous, and
has been referred to Padissya ;* another is a stem of
afern. Thespecies of Esther7a has been named
EL. mangaliensts by Rupert Jones.? A smaller
variety closely resembles the Es‘her¢a found in
the Panchet group of Bengal and may be identical,
but the identification is not quite certain, as
the Pdnchet fossil is so poorly preserved that

Fig. 11.—Estheria mangali- some of the specific characters, depending upon
ensis, Rupert Jones (enlarged : :
3 diameters). the microscopical texture of the shell, cannot be
ascertained.

 

The uppermost beds of the lower Gondwanas in the Sdtpura range have
been distinguished under the name of Almod from a village at the south
base of the Pachmarhi escarpment. ‘The rocks consist of sandstones witha
few carbonaceous shales, from which no fossils have been obtained, Their
relations to the groups above and below require further investigation. No
unconformity has been traced and their sole importance comes from their
position between the Mahddevas and Damudas, and the consequent pos-
sibility of their representing the Panchets of Bengal.

The term Pdnchet was originally applied to two groups of beds in
the Rdniganj coalfield It is now restricted to the lower of these
groups, the upper Pdnchets of the Ddmodar valley being referred to
an upper Gondwdna age, and ascribed to the Mahddeva series. The
name was derived from an important zamindéri, which still com-
prises a large tract in the southern portion cf the Rdniganj coalfield and
formerly included much more, and the same name is that of a large hill,
the basal portion of which consists entirely of Panchet beds,

The great mass of this group consists of thick beds of coarse felspathic
and micaceous sandstones, often of a white or greenish white colour, with
bands of red clay from a few inches to twenty feet inthickness. The felspar,
in the sandstones, is occasionally undecomposed, which is never the case in

1 See Quart. Four. Geol, Soc. X, 472, | stem belonging to the Lycopodiaceous genus
(1854); XI, 37, (1855). Knorria. ‘

2 Feistmantel, Records, X, 26, (i877). The * Pal. Soc, Mem. Foss, Estheria, p. 78,
identification seems doubtful, for Sir Charles | (1862).
Bunbury suggested the possibility of the same * Memoirs, Il}, go, 126, 132, ete., (1869).
 

yh

 

MH

Samazopal

      

Oleandridtum cf.

stenonvuron, Schenk

 

Thinnfeldis, cf

odonlopleroides, Morr

Prcopteris concinria, Presl

clupteris pachyrhacis, Gopp

 

 

Di-vnuden onentelis, Huxley

Lithographed & Vrinted at Geological Survey Office

PANCHET FOSSILS.
Chap. VILJ PANCHET GROUP. 175

the Damudas, Conglomeratic beds sometimes occur in the upper portion
of the group, but they are not common. At the base of the group grey
and greenish grey sandstones and shales are usually found in very thin
beds, and often highly micaceous, In places the greenish micaceous
clays are met with higher in the group.

The Pdnchet rocks are distinguished from the typical Damudas by the
presence of red clay and the absence of carbonaceous shales, and by the
sandstone being, as a rule, much more micaceous. But rocks of the
Pdnchet character are found in parts of India interstratified with the
Damudas, as in the Motiér group.

The thickness of the present group in the Ddmodar valley nowhere
exceeds about 1,800 feet. It rests with slight, but distinct, unconformity
upon the denuded surface of the Rdnfganj group, and in some places the
Pdnchets completely overlap that group and rest upon lower beds, the
unconformity being most marked in the Bokdro coalfield. Fragments of
coal and shale, apparently derived from the Damudas, have occasionally
been found in the conglomerates of the Panchet group.

The most important remains of animals hitherto found in the lower
Gondwana rocks have been derived from the Pdnchets. In the upper
portion of the group there is, in the Ranfganj coalfield, a well marked
conglomeratic band containing reptilian and amphibian bones. These are
isolated from each other and sometimes slightly rolled The specimens
obtained have been examined and described by Prof. Huxley and Mr.
Lydekker! and comprise the following forms, besides a few others whose
affinities are doubtful. The plants were described by the late Dr.
Feistmantel.2, A dagger in the following list indicates species also found
in the Damudas :—

 

ANIMALIA. _ PLANTA.
AMPHIBIA— EauiseTaAcER—
Lab yrinthodontia— t Schizoneura gondwancnsis.
Gonioglyptus longirostris. + Vertebraria indica.

” huzleyt. FILICES ~— is
Glyptognathus fragilts. Pecopteris concinna,
Pachygonia incurvata. Cyclopteris (?) pachyrhaca.

REpPTiua— Thinnfeldia, cf. or ontopteroides.
Dicynodontia— Oleandridium, cf, stenoneuvon.
Dicynodon orientalis. t Glossoptesis eS
. . oa.
Ptychosiagum (Ptychognathus) orten- f phon i.
tale. t angustifolia.
Dinosauria— CycapEacER—
Epicamopdon (Ankistrodon) indicus. Scales.
CRUSTACEA— Conirere—
Estheria mangaliensis (?). t Samaropsis, ci. parvula.

2 Pal. Indica, series iv. Indian pretertiary | °® Pal Indica, serizs xii, III, pt. ii, pp. 5I— 56,
Vertebrata, (1880).
172 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL

IIl._—Tue UppierR GONDWANA SERIES,

The unfossiliferous beds of the lower part of the upper Gondwdnas
have, in the more recent publications of the Geological Survey, been
generally referred to under the name of Mahadeva. This name, first
applied to the sandstone of the Pachmarhf hills,! was subsequently extend-
ed so as to comprise all the beds of the Sdtptra basin above the
Damudas of the lower Denwa valley, except the Jabalpur group, and
has utimately come to be a convenient, because indefinite, term
to apply to the soft sandstones and conglomerates, of obscure strati-
graphical position, seldom containing any fossils except fragments of
carbonised wood, which there is good reason tv believe belong to the
lower part of the upper Gondwdnas, though they cannot be assigned with
certainty to any particular horizon, ‘

The Mahadeva rocks consist chiefly of very thick massive beds of
coarse sandstone, grit, and conglomerate. These are frequently ferrugin-
ous, or marked with ferruginous bands, as in the Kdmthis, . They are asso-
ciated with clays, and occasionally with bands of impure earthy lime-
stone, The sandstones form high ranges of hills, and often weather into
vertical scarps of great height, forming conspicuous cliffs that contrast
strongly with the black precipices of the Deccan traps and the rounded
irregular masses of the more granitoid metamorphic rocks.

In the typical area of the Sdtpura region the Mahddeva rocks attain
a thickness of at least 10,000 feet, nine-tenths of which consist of coarse
sandstone, grit, and conglomerate. They appear to be unconformable to
the underlying Damudas, as the series overlaps the upper members of
the lower Gondwana series.

In the SAtpura region the Mahddeva formation has been subdivided -
into three groups,—the Bdgra, Denwa, and Pachmarh{,— each of which re-
quires a few remarks.

‘The name of the Pachmarhi? group is derived from a village on the top
of the hills of the same name, and the site ofasanitarium. The group con-
sists of massive sandstone, whitish or brownish in colour, usually soft, often
containing small subangular pebbles, and occasionally intersected by hard
ferruginous bands. As a rule, the stratification is obscure, oblique lami-
nation being common, and the different beds of which the group is
composed exhibit great irregularity in superposition and often overlap each
other. The.hard ferruginous partings are most irregularly interspersed
throughout the mass, usually as thin beds, though not always perfectly
parallel to the planes of stratification. Sometimes the impregnation with
iron is confined to pipes or nodules. Fragments of these ferruginous bands:

1 Four. As, Soc, Beng., XXV, 252, (1856) ; 2 Memoirs, X, 155, (1873).
Memoirs, 11, 183, 315, (1869).
Chap. VIII MAHADEVA SERIES. 173

are often scattered in quantilies over the surface, and serve Lo distinguish
the outcrop of the Pachmarhf group from those of the underlying beds.!

The Pachmarhf group comprises, where thickest, 8,o00 feet out of the
10,000 found in the Mahddevas of the Satpira hills.

The middle group of the Sdtpura Mahddevas is named ® after a
stream which rises on the south side of the Pachmarhi range and,
turning round the eastern end of the ridge, forms its northern boundary
throughout, falling finally into the Tawa. The course of this stream,
north of the Pachmarhf hills, is the area of the Denwa rocks, which,
presenting a marked contrast to the massive Pachmarhi sandstone, are
principally composed of soft clays, pale greenish yellow and bright red,
mottled with white in colour, forming thick beds interstratified with
discontinuous and subordinate bands of white sandstone, and very rare
courses of earthy limestone. The sandstones are locally conglomeratic.
In short, in mineral character the Denwa rocks are a repetition of the
Motir group in the middle of the Damuda series, and resemble the
Panchets of Bengal.

The thickness of these beds in the Denwa valley is about 1,200 feet.
They appear in places to pass into the underlying group, although they
are quite distinct in the typical area.

The Denwa group is the only one which can be correlated to those
of other parts of India on palzontological grounds. Vertebrate remains
have been found in it, and more abundantly in south Rewd, which show
that it is the equivalent of the better known Kota-Maleri group of the
Godavari valley. The name is consequently one which will probably drop
out of use as the relations of the rock groups are more completely
worked out.

-

The uppermost group was named Bagra$ from a hill fert built upon it,
where the river Tawa cuts its way through a spur of the Sdétpura hills,
south-east of Hoshangdbdd. It is largely composed of conglomerates,
often coarse, frequently with a deep red sandy matrix. It is more
calcareous than the other Mahadeva groups, and bands of calcareous sands
and clays and limestones, sometimes dolomites, are of frequent occurrence,
The group is very irregular in composition. The greatest thickness does
not exceed 600 to 800 feet, and in places it overlaps the Denwa shales
and rests directly on the Pachmarhi sandstones.

The generally unfossiliferous nature of the Mahddeva beds, their soft-
ness and ease of weathering, render their recognition with certainty a
matter of difficulty. In the coalfields of the Damodar valley some soft
pebbly sandstones, which were formerly regarded as upper Pdnchet, are now

1 It should not be fergotten that similar 2 Memoirs, X, 153, (1873).
ferruginous layers are found in the Kd4mthis. 4 Memoirs, X, 150, (1873).
174 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VII,

regarded as very probably Mahddevas. Their relations to the underlying
rocks are difficult to make out, the junctions being greatly obscured by
pebbles and detritus derived from the newer grits, but there appears to be
some unconformity.

-In the TAlcher field and in Chhatisgarh the uppermost soft pebbly sand-
stones are believed to be Mahddevas, but have not been coloured as upper
Gondwanas on the accompanying map, as their age has not been satisfac-
torily established, and the outcrops are small,

Mahadeva sandstones are found in the Narbadd valley, running out
from under the scarp of the Deccan trap at Bdrwai,' and further west
certain sandstones, underlying the cretaceous of the Narbada valley, are
probably of Mahddeva age, but as they have been held to be cretaceous
the question of their age will be discussed when dealing with that

system.”

In the RAjmahdl hills the lower Gondwdnas are overlaid by a thick
band of coarse sandstone, which was at first associated with the overlying
beds, but has since been separated, as it is unconformable to them. It is
lithologically very similar to some beds in the Damodar valley, which are
believed to be of upper Gondwana age, and possibly is a representative
of them.

The Dubrdjpur group, as this band of sandstones and conglomerates
is called, takes its name from a village® in the Rajmahdl hills, situated
about forty miles north by east of Suri. The component beds are sand-
stones of several varieties, grits and conglomerates, for the most part ferru-
ginous. Fine grained beds are not common, although shaly sandstones
are occasionally met with. Most of the coarser beds are ferruginous, and
one form of conglomerate, of frequent occurrence, consists of quartz
pebbles in a ferruginous matrix. <A precisely similar bed is found in the
supposed Mahadeva beds of the D4modar valley.

Along the western scarp of the Rajmahal hills the rocks of the Dubrajpur
group rest partly upon the Damudas and partly upon the metamorphic
rocks, the Damudas (Bardkars) being repeatedly overlapped by the
Dubr4jpur beds in a manner which shows the two to be quite unconform-
able. The greatest thickness of the Dubrdjpur group in the R4jmah4l area
does not exceed about 450 feet. Some specimens of a cycadeaceous plant
(Pizlophyllum) were once found near the southern extremity of the hills
in the uppermost beds underlying the Rajmahal trap, but there is some
little doubt as to, whether the fossiliferous band may not belong to the
Rajmahdl series itself.

1 Records, VIII, 73, (1875). 3 Pal. Indica, series ii, f, 1, 41863);
3 Infra, p. 253. Memoirs, XIUI, 198, (1877).
Chap. VILJ RAJMAHAL GROUP. 175

The Rdjmahdl series derives its name from a range of hills in Bengal,
extending north and south from the Ganges to the neighbourhood of
Suri in Birbhim, and, unlike the other members of the Gondwdna
system is confined to the neighbourhood of the eastern margin of the
Indian Peninsula. Some species of fossil plants, identical with Rdjmahdl
forms, have been found in other localities, but they are either isolated, or
associated with plants belonging to a different flora.

In its typical locality the kdjmahdl group of the Rdjmahal series con-
sists of a succession of basaltic lava flows or traps with interstratifications
of shale and sandstone. The sedimentary bands are held to have been
deposited in the intervals of time which elapsed between the volcanic out--
bursts, by the circumstance that the different bands of shale and sandstone
differ from each other in mineral character, and also that the upper surface
of the shaly beds has sometimes been hardened and altered by the contact
of the overlying basalt, whilst the lower surface is never affected. ‘The
sedimentary bands are chiefly composed of hard white and grey shale,
cartonaceous shale, white and grey sandstone, and hard quartzose grit.

       

     

  

         
  
  

on LA
Tm ae
- Ae | ”

THD IDEN ee
— TTY
a
UN

Fig. 12.— Radiating columnar trap, RAjmahdl hills.

The trap rocks are all dark coloured dolerites, They vary in character
from a fine grained, very tough and hard rock (anamesite), ringing under
the hammer, and with the edges of its fracture almost as sharp as those
of a quartzite, to a comparatively soft, coarsely crystalline basa't. The
latter usually contains olivine in large quantities. Many of the trap rocks

1 Four. As. Soc. Beng. XXII, 263, | 209, (1877); Pal. Ind., series ii, J, 1, (1863).
(1854); Memoirs, Ul, 313, (1860); XIII,
176 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VIL

are amygdaloidal, the enclosed nodules usually containing some form of
quartz, either agate, chalcedony, or rock crystal. Occasionally, but less
frequently, zeolites are found, stilbite being the commonest, natrolite less
abundant, and analcime has also been detected. It is not usual to find the
cavities lined with green earth, as is so frequently the case amongst the
amygdaloids of the Deccan trap. The basaltic flows above the sediment-
ary bands are, as a rule, compact,

Very little light is thrown on the source of the basaltic rocks by any
observations within the RA4jmahdl area. Dykes are rare, and there is only
one instance known of an intrusive mass which may mack the site of an
old volcanic outburst. This is about 22 miles south south-east of Colgong
cn the Ganges, close to a place called Simra, where a group of small coni-
cal hills occurs, composed of pinkish trachyte, porphyritic in places, and
surrounded by Damuda rocks. The surface of the ground is much ob.
scured by superficial deposits, but there appears good reason for suppos-
ing that the core of a volcanic vent is here exposed. It appears not an
unfrequent occurreace that the later outbursts from a volcano are more
silicious than earlier eruptions, and that a volcanic core, even when the
lava flows have been doleritic, should itself prove trachytic, when exposed by
denudation. This may be due to the solution of the highly silicious meta-
morphic rocks through which the outburst took place by the molten lava
remaining in the fissure after the eruption, and the consequent conversion
of that lava from a basic into an acid rock.

Trap dykes and intrusions, believed to be of Rajmahdl age, are
abundant in the coalfields of the Damodar valley, and both dykes and cores
of basalt are common in the portion of Birbhtim lying south-west of the
Rajmahdl hills. It is possible that the principal vents lay in this direction,
or they may have been in the region now covered by the Ganges alluvium.
The difficulty of determining the original source of eruptive rocks will be
again illustrated in the case of the Deccan traps.}

The bedded basaltic traps of the RA4jmahdl hills, with their associated
sedimentary beds, attain a thickness of at least 2,000 feet, of which the
non-volcanic portion never exceeds 100 feet in the aggregate. They rest
with general parallelism on the grits and coarse sandstones of the
Dubrdjpur group, but nevertheless several instances of overlap take place,

1 Some doubt still attaches to the deter-| tions of these traps in the feld cannot be said
mination of the true age of the Rdjmahdl | to havebeen so close as to preclude the pos-
traps. Detailed examination has shown that | sibility that the supposed interbedded traps
a close lithological resemblance exists be- | are really intruded along the planes of bed=
tween them and the Deccan traps; Records, | ding. The correlation of the RAjmahal traps
XX, 104, (1887); XXII, 226, (1889). This | with the Sylhet traps is an important point,

is not, however, in itself sufficient to prove|as (if it is correct) the former must be

their contemporaneity. On the other hand | o'der than the Deccan trap period.
the examination of the stratigraphical rela-

yy.
   
    

   

Palissyu conferta, Fstm

 

Gleichenites bindrabunensis Schimp

 

Cveadites confertus. 0. & M

Thotyozamites Falcatus 0. & M

 

Otorarites bengealensis. 0. & M.

Lithographed & Printed at Geological Survey Office.

RAJMAHAL PLANTS.
       

Oldhara

Pecopteris lobate

 

Morris.

Pierophyllum rajmahalense

 

Pterophyllum princeps 0. & M.

Macrotcemopteris lata. Oldham.

Geological Survey Office.

lithographed & Printed at

PLANTS.

RAJMAHAL
Chap. VIL.) RAJMAHAL GROUP, 177

and, in one locality at least, there is evidence of the Dubrdjpur beds having
been denuded before the deposition of the Rajmahal group.

From the extent of the area throughout which these dykes are de-
veloped, conclusions may be drawn as to the original limits of the volcanic
action coincident with the period of deposition of the Rdajmahdl group,
The number of trap dykes gradually diminishes in tne coalfields of the
[)Amodar valley from east to west, until finally, in the Karanpurd field
south-west of Hazdribdgh, volcanic intrusions disappear almost entirely,
and none appear to be known further west, until basalticdykes of different
age, which apparently are contemporaneous with the much newer
Deccan trap, make their appearance. Ouiside of the coalfields it is
difficult to distinguish the dykes belonging to the Rdjmahal period from
older eruptions, but there is not the same abundance of extensive basaltic
intrusions in southern Monghyr, Hdzariba4gh, and Chutid Ndgpur asin
Birbhim, So far as can be judged, the region immediately north of the
Raniganj coalfield was one of the foci of eruption and it is far from im-
probable that the bedded traps of the Rajmahal hills had originally a
considerable extension to the south-west and south, though, as no single
outlier has been preserved, it is impossible to feel sure of the inference.
There is, however, considerable probability that a large tract in the
Damodar valley, including the whole Rdniganj field may have been once
covered with bedded traps.

The great difference of age between the R4jmahdl group on the
one hand and all the lower Gondwana rocks, including tke Damudas
and P4nchets, on the other, is well illustrated by the change in the
flora and by the very much greater amount of disturbance to which the
Damuda rocks- have been subjected. The Rdjmahdl traps are almost
horizontal, and no faults have been observed in them, while the dykes
which abound in the RAnfganj field,and are almost certainly of Rdjmahdal
age, are newer than the faults of the coalfield.

By far the greater part of the Rajmahal fossils have been obtained
from two bands of fine grained whitish or greyish shales—the upper 25
to 30 feet thick, the lower 10 to 15,—separated from each other by
lava flow, and having other beds of trap, with intercalations of sandstone
and shale, above and below,

The following is a list! of those hitherto described from this group :—

EauISETACER-— Fiviczs,—contd.

Equisetum vajmahalense. _ Sphenopterts hislopi.
FILicEs— ” meimbranvsd.

Sphenopteris aiguta, Dicksonia bindrabunensts.

1 The list is taken from O. Feistmantel, original specific names have be2n restored jn
Pal. Indica, series ii, 1, 143, (1877). In the place of P. indica, D. indicus and HK. rai-
case of Thinnfeldia salicifolia, Dietyozamites mahalensis, as there does not seen to be suffi.
faleatus and Echinostrobus indicus, the cient reason for the change.

N
178 GEOLOGY OF INDFA—GONDWANA SYSTEM.

Fitices—contd.

({Chap. VII.

CycaDeAce & —conta.

Hymenophy lites bunburyanus. Pterophyllum rajmahalense.
Cyclopteris oldhami. 5 fissum.
Thinnfeldia salictfolia. 5 cf. propinquum.
Alethopteris indica. Zamites proxtmus.
Asplenites macrocarpus. Ptilophyllum acutifolinm.
Pecopteris lobata. 53 cutchensz.
Gleichenia bindrabunensts. Otozamites bengal ensis.
Angiopteridium macclellandt. 5 abbreviatus.
os spathulatum. 3 oldhami.
a ensts. Dictyosamites falcatus.
Macroteniopteris lata. Cycadites confertus.
= crassinervis. » - vagmahalensts.
a ovata. Cycadinocarpus vajmahalensis.
4 morrisi. Williamsonia microps.
Daneopsis rajmahalensis. Ptilophyllum ef. W. gigas.
CycaDEACEE—
Pterophyllum distans. ConiFeRE—
ss cartertanum. | Palissya indica.
ee morrisianum. ss conferta.
i medlicottianum. Chirolepis gractlis.
oi princeps. Cunninghamites dubgosus.
* crassum. Echinostrobus indicus.

 

The first thing which must strike any one in looking over the above
list is the great change in forms of life between the upper and lower
Gondwana series, so far as we are yet acquainted with them. It is
highly probable that intermediate beds may hereafter be found, but for the
present there seems to be, in Bengal at least, just as great a break in the
flora as inthe stratigraphy. The most striking distinction is that the
prevalent forms in the lower Gondwdnas are Egursetacez and ferns of the
Glossopteris type, Cycadeacez being rare, whilst in the upper Gondwanas,
and especially in the RAjmah4l group, Cycadeacee prevail, their individual
abundance being so great that they frequently form the mass of the
vegetation. In fact, the cycads, and especially Péilophyllum acutifolium
are just as abundant and characteristiz in the Rajmahdl group, as
Glossopteris and Vertebraria are in the Damidas,

The Rajmahdél beds are represented along the east coast by a series of
small outliers, most of them too small to deserve detailed notice here, which
are interesting, as they appear to comprise some rock groups of later age
than the R4jmahdl, and contain marine fossils associated with the plants.

The most northerly of these isthe Athgarh basin, a tract of sandstone,
some twenty miles long from north to south, and eighteen miles from east
to west close to the town of Cuttack, on the western margin of the alluvial
plain. Some carbonaceous shale, occurring in the lower portion of the
sandstones, has been supposed to indicate the presence of Barakar heds,
but no Damuda fossils have been found. The relations of the remaining
Chap. VIL] RAJMAMAL OUTLIER OF THE ELLORE. 179

part of the rocks are very obscure, and they were believed to be of K4mthi
age till the discovery of characteristically R4jmahdal fossils in them.

On the right bank of the Godavari, near Thalapddi, about
ten miles above Rajdmahendri, a well marked belt of upper Gondwana
beds commences, which extends for sixty miles, from the Godavari to be-
yond Golapilli west of Ellore. The width of this belt varies from ten to
fifteen miles. There is a general dip to south-east or east-south-east at 5°
to 10°, and the beds rest unconformably, throughout a considerable portion
of their area, upon various members of the Kdmth{ group, but they
overstep this group, both to the east and west, and rest upon a sloping
floor of gneiss, which has the appearance of a plane of marine denu-
dation formed after the deposition of the K&mthf rocks, as the latter
rest upon a much more uneven surface of the metamorphic formations.
This appearance of resting upon a surface which had been fashioned
by denudation after the deposition of the lower Gondwdna beds, quite
agrees with the peculiar distribution of the Rdjmahdl group and its
associates, which evidently were accumulated in a distinct area from
that in which the Gondwana beds of the Godavari valley were deposited.
To the south-east the upper GondwAna beds of the Ellore area disappear
beneath the Cuddalore sandstones and the alluvial deposits of the
Godavari delta, except west of Rajdmahendri, where the Gondwdnas are
covered by outliers of the Deccan traps.

The rocks of the Ellore area are peculiarly interesting, because they
appear to contain representatives of groups higher than the R4jmahAls,
associated with beds in which the typical RAjmahdl flora is well preserved.
Dr. King, who surveyed the rocks of the Godavari district, classed the
upper Gondwdna beds in three subdivisions, thus distinguished in de-
scending order :? 1. Tripetty sandstones.
2. Ragavapuram shales,
3. Golapilli sandstones.

The Golapilli sandstones consist of brown and red sandstones and
conglomerates which form a broad plateau near Golapilli, capped by
conglomerates and _ gravels, probably belonging to tbe Cuddalore
sandstones,

The following plant fossils have been obtained from the Golapilli
beds :—®

Finices— | FILices,— contd.
, Alethopteris indica, Angiopteridium ensis.
Pecopteris macrocarpa. a spathulatum,

1 Feistmantel, Records, X, 68, (1' 77); Pal.

2 Records, X, 56, (1880); Memoirs, XVI,
Indica, series xii, I, 187, (1879). ,

211, (1889).
8 Pat, Indica, se ies ii, I, 163, (1877).

N 2

 
180 GEOLOGY OF INDIA—GONDWANA SYSTEM. (Chap. VII.

CYCADEACER— CYCANEACE ® —contd.

Pitlophyllum acutifoliune. Williamsonia gigus.
59 cutchense.

Dictyozamites indicus. ConwERa—

Pterophyllum morrisianum. Pulissya conferta.

carlertianum. ‘5 inlica.

oe kivgianum, Chetrolepis, cf. muensteri.
~ distans, Arancarites macvopterus.

All the above, with the exeption of Plerophyllum kingianum, the
Cheirolepis and Araucarttes macropterus are characteristic Rajmahal

forms.

Resting upon the Golapilli beds, in the neighbourhood of IHore, there
is found a thin band of white and buff shales, having a few interstratifica-
tions of sandstones towards the base, not more than roo feet thick. No
unconformity has been detected between these shales and the Golapi'li
sandstones, but there appears to be some difference in the flora, for, while
the plants of the Golapilli standstones are all Rajmahdl forms, except a
few species peculiar to the beds, the flora of the overlying shales com-
prises, in addition to several forms common to the beds below, a few
species allied to Jabalpur plants. The shales have been called Ragava-
puram, from a village situated about twenty-six miles north-north-east of
Ellore. A list of the plants will be found in the tabular statement on
page 183.

With the plants are some marine shells, chiefly casts, amongst which
are some Ammonites, apparently allied to middle jurassic species, the
principal form being near A. ozs, but distinguished by having the ribs
simple throughout. A, of7s belongs to the subgenus Stephanoceras, and
to the group of A. macrocephalus, and is found in the Chdri and Katrol
beds of Cutch (callovian and oxfordian). Besides the ammonites, Leda)
Pecten, Gervillia, etc., occur, the Leda being especially common and
characteristic.

Above the shales just noticed there is another thin band of dark brown
and red sandstones and conglomerates, chiefly ferruginous, with silicious
and argillaceous bands, and beds of concretionary clay ironstone named
from a pagoda called Chinna (little) Tirupati (Tripetty), which stands
upon a scarp composed of them, about twenty miles north-north-east of
Ellore. Towards the bottom these sandstones become softer and less
ferruginous. In the main area, near Ellore, these Tripetty beds are only
40 feet in thickness.

The Tripetty beds in the main area have only yielded fossil wood,
but from some outlying patches, near Innaparazpdlaydm about twenty-four
miles north by east of Cocondda, supposed to belong to the same band, Dr.
Chap. VIII RAJMAHAL BEDS OF GNGOLE, Sy

King obtained two Pv/gonte,—T. smect and 7. ventricosa,—both of which
are characteristic of the Uma beds of Cutch.

The sequence of upper Gondwdua beds in the neighbourhood of Ellore
is very instructive. The whole series rests unconformably on the Kamthis
(lower Gondwana), and although the whole thickness of the upper Gond-
wana series is trifling, apparently not exceeding 209 or 300 feet, it com-
prises representatives of the Rdjmahdl and Umia groups, and of an inter-
mediate furmation, Yet these thin bands exhibit no marked unconformity.
The middle group is overlapped at both ends, it is true, but there is nosign
of any important break. It is clear that the country must have undergone
very little disturbance in the interval between the deposition of the differ-
ent groups, and, judging from this instance, it is impossibl: to argue from
the small amount of discordance between successive subdivisions of the
Gondw4na series, that the period of time which elapsed between the differ-
ent groups was of small amount. No notice would, in all probability, have
been taken of the distinctions between the different beds at Ellore, but
for the fossils, and many similar subdivisions might be practicable in such
groups as the Kdmthi or Pachmarhi if the strata were fossiliferous,

South of the Kistna river Mr. Foote has detected a threefold division
of the Gondwana beds, similar to that of the Ellore cert and has distin-
guished the groups seen near Ongole as follows :—!

3. Pavulur sandstones,
2. Vemdvaram shales
1. Budavada sandstones,

Besides numerous remains of marine organisms, not yet determined,
which have been obtained from the two lower groups, the Wantdyacani
shales have yielded a tolerably rich flora of a type similar to those of the
Ragavapuram and Sripermatir groups. To avoid needless repetition of
names and to exhibit more clearly the relationships of the flora of these
groups, they have been combined in the tabular statement on page 183, from
which it will be seen that nearly half the Ragavapuram species are also
found in the Vemdvaram beds, and that the relationship of the floras to
those of the Rajmahal and Jabalpur groups is very similar in each case.
It is, therefore, tolerably certain that the similarity of grouping of the beds in
the two areas is not merely accidental, but that the three groups in each
region are respectively equivalents of each other.
The only Vemdvaram fossil, apart from the plant remains, which has been
determined, is a macrurous crustacean regarded by Dr, Feistmantel? as

1 Memoirs, XVI, 69, (1880). correct locality is given by Mr. Foote in Me-
? Records, X, 193, (1877). The locality is motrs, XVI, 63, (1879).
there given by mistake as Sripermatir. ‘The
182 GEOLOGY OF INDIA—GONDWANA SYSTEM. [ Chap. VIL,

probably identical with the liassic Bryon barroviensis. There are, however,
some important differences which make the specific identity of the two
forms doubtful.

The upper Gondwana beds near Madras are divided into two groups;!
the lower of which has been named from Sripermatir, a town 25 miles
west-south-west of Madras and a well known locality for fossil plants.
The group is composed of white shales, containing plants, associated with
sandstones, grits and micaceous sandy shales. Corglomerates occur, espe-
cially towards the base, where they are coarse and occasionally contain
boulders of great size, but all the conglomerates are loose in texture and
not compact. A boring recently put down at Place’s garden near Madras
has penetrated beds of carbonaceous shale, overlapped at the surface,
which have raised hopes of finding workable coal. It is not at present
known whether these belong to an outlier of the lower Gondwd4nas or not.

It is in the Sripermatur shales that the fossils of the group are found.
They consist of both animals and plants. The shells are ill preserved and
have not been determined. They comprise two or three species of Am-
monites and several lamellibranch bivalves, The Cephalopoda were regard-
ed by Dr. Waagen? as resembling neocomian rather than jurassic forms,
but the species cannot be determined, owing to the poor state of preserv-
ation, too much weight must not be attached to the opinion, though
it is of interest in connection with the resemblance between the R4jmahal
series and the Uitenhage series of South Africa, now regarded as neo-
comian in age. The occurrence of Zrigonia smeed and T. ventricosa, both
South African Uitenhage species in the outher north of Coconada, has
already been mentioned, and it is not necessary to make any further refer-
ence to this subject here, as it will be more fully treated of in the next
chapter.

The Sripermatir group is overlaid by a set of beds of coarse compact
conglomerate, with intercalated sandstones and gtits, which have been dis-
tinguished as the Sattavédu group, from a series of moderately elevated
ridges of the same name, lying about thirty-five miles north-west of Madras.
Only imperfectly preserved plant remains have been obtained from the
Sattavédu beds, The junction with the Sripermatdér beds is ill seen, the
groups appear to be conformable, and it is doubtful whether there is suffi-
cient justification for the separation of the upper beds as a separate group.

The upper Gondwdnas of the Trichinopoli district occur as narrow
outcrops along the western edge of the cretaceous beds, which they se-
parate from the gneissose rocks, being quite unconformable to bath. The

* Memoirs X, 64, (1873). | * Pal. Indica, series ix, p. 236, (1875).
Chap. VILJ UPPER GONDWANAS OF EAST COAST. 183

Tabular Statement showing the distribution of the fossil plants of the
Ravavapuram, Vemdvaram, Sripermatir groups, and Utatir beds
of the Rajmahdl series)

(+ Signifies that the species is also found in the Rajnahal group* in the Jabalpur group,

«

 

 

@ : :
a1 ee
= | 8] & 3
SB) ee] 2 | #
2 > oa 5
Fitices—
Thinnfeldia subtrigona . . ‘ < . % . & Bs
Dichopteris ellorensis  , . . - . < Pa ad don ss
* Alethopteris whitbyensts . . . . . * * &
5 indica ‘ ‘ . . z - silll) | ade *
Pecopteris reversa . ‘ Z ‘ s -| * *
+ Angiopteridium spathelatuns . * ‘ . 5 | * x *
+ Pr macclellandi ‘ : . oo cal) x * 5.4
+ Macroteniopteris ovata(?) . . . . . zi] vse ee . ?
CycaDEAcEa—
Anomozamitesjungens . ‘ s is . - ‘ ¥* ee 326
$s lindleyanus . ‘ i ‘ . low st *%
t i fissus . . . ‘ . . * ‘
Péterophyllum footeanum 3 . . . , ‘ *
+ Zamites proximus ; ‘ iz 5 . e f * =
* Podozamites lanceolatus ‘ go oy eS ot .| * ie
+ Otosamites abbreviatus . . . . . 7 »| * %- *
a varinervis . 5 ; . . . ‘ ‘ % *
FA bunburyanus ‘ : . . . %* * s
* 5 hislopi - 7 ‘ ‘ * *
i parallelus . 3 ¥ . ‘ . . aoe x *
5 acutifolius . . . . . . of % i
angustatus . . . . ri é ses es x%
t Ptilophyllum acutifolium 5 . ‘ 2 . -| * % x
t ye eutcherse .. * . . . . . ” %& * %
+ Dictyozsamites indicus . . : . . . . % %
Cycadites cowstrictus . . . . . . 2 * we ah
Conirer&
+ Palissya conferta : “ ‘ . 7 . al ee we * %*
+ 5 indica . ‘ . ‘ ‘ : * 2
* »  jabalpurensis , . . ‘ . 5 ‘ ” * dive sa
* Araucarites cutchensis . ‘ 3 . ‘ : F * * p
* ii macropterus 3 x > . a sie *
Pachyphyllum peregrinum . ‘ . ‘ : ‘ * *
i heterophylilum . 5 . A . ¥* a "
+ Echinostrobus rajmahalensis . ‘ 5 a * ey
* - rhomticus é “ # . F ‘ % .
~ is €%, ANSUS s . ‘ i x 4 .
™ Taxites tenerrimus 7 7 < 7 , . wi | ne se
sy planus i s ‘ . 2 : : - x | * *
Gingko crassipes . r . . é . x * :

 

 

 

 

 

 

1 Pal. Indica, series ii, 1, 199, (1879).
184 GEOLOGY OF INDIA—GONDWANA SYSTEM. [Chap. VIL

most important of the outcrops is that near Utatdr, and the rocks consist
chiefly of soft sandy clays and micaceous shales, with sandstones and a
coarse conglomerate of rounded gneiss pebbles at the base.!

The Utatdr outcrops are the most southerly known to be of Gondwana
age, but Mr. Foote has recorded some exposures of shales and conglomer-
ates in the Madura district which closely resemble the beds of the coastal
Gondwanas. No fossils were, however, found, so the identification is not
fully established.

Reference has already been made to the resemblance between the floras of
the Ragavapuram, Vemdvaram, Sripermatiér groups, and of the Utatir out-
crops and, to save repetition, a separate list of fossils has not been given in
each case,but the whole united in the tabular statement on the previous page.
From this it will be seen that not only are there a certain number of forms
common to two or more of the groups, but that in each case the flora is
characterised by a large proportion of R&jmahdl species, and a much smaller
proportion of forms that are only known from the Jabalpur and Umia groups.
Apart from the presence of Jabalpur species, and species peculiar to these
outliers, their flora is distinguished from the trae RAjmahdl flora by the
absence of broad leaved Tentopteridex,and the greater abundance of certain
forms, such as Angiopteridium spathulatum and Dictyozamites indica,
which though represented in the RAjmahdl beds, are found only in a
smaller proportion.? The presence of the genus Macroteniopteris in the
Utatur beds, and the smaller proportion of purely Jabalpur species,may show
that they are nearer in age to the Rajmahdl group than the others, but,
with this possible exception, we may take the groups as being at any
rate approximately of the same age, intermediate between that of the
Rajmahal and Jabalpur groups, as is shown in the tabular statement
on page 156.

In the Godavari valley the principal representatives of the upper Gond-
wanas are the Kota and Maléri groups, more commonly referred to as the
Kota-Maléri group, the name being derived from those of two villages
long known to Indian geologists, the former by the discoveries of
lish teeth and fossil fish by Dr. Walker and Dr. Bell® in 1851, the latter
by the late Revd. S. Hislop’s discovery of reptilian bones.* The village of
Kota is on the left bank of the Prdnhita or Waingangd4, about 8 miles

1H. F. Blanford, Memoirs, IV, 39, “1 V 2 c

* Pal, Indica, series ii, 1, 199, (1879) * Quart. Four. Geol. Soc. XX
i s ; . 7 1 S0C., > 280, (1864).
® Quart Four. Geol. See. VII, 272, (1851); z : aie
Chap. VILJ

above its junction with the Goddvari. Maléri is about 32 miles north-west
of Sironché

The combined group, usyally spoken of as a whole, is slightly, but dis«
tinctly, unconformable to the underlying Kamthfs, and is divided into two
subgroups, which were separately mapped by Dr. King.!’ The lower, or
Maléri, consists essentially of bright red coloured clays, interbedded with
soft, light coloured and open textured sandstones subordinate in thickness
to the clays. The fossils are found in the red clays, coprolites being
much the most abundant; besides them three species of Ceratodus and
two genera of reptiles, Ayperodapedon and Parasuchus, have been
distinguished,

The Kota subgroup, which overlies the Maléri, consists principally of
coarse, loosely compacted sandstones, with some subsidiary bands of shale,
and three very strong bands of limestone, from which all the animal remains
have been obtained, the few plants being all derived from sandstone
bands,

Though the distinction between these two subgroups is traceable in the
field, they are so closely associated that the fauna may be treated as a
whole and, as the few fossils of the Maléri subgroup have already been
mentioned, there will be no difficulty in separating them in the subjoined
ist of the fauna and flora of the combined group :—

UPPER GONDWANA, M‘DRAS, 188

ANIMALIA.

CrusTacEa—
Estheria ketchensis.
Candona kotahensis.
InsEcTa—
Undetermined—
PISCES -—-
Lepidotus deccanensis.
is longiceps.

” breviceps.
»  pachylepis.
9 calcaratus.
Tetragonolepis oldhami.
3 analis,
FiLices—

Angtopteridium spathulatum,

CYCADEACER ~

*t Ptilophyllum acutifolium.
Cycadites, sp.

 

‘PisCEs.—contd.
Letragonolepisr ugosus.
Dapedius egertoni.
Ceratodus hunterianus.

3 hislopianus.
+8 virapa,

REPTILIA——
Hyperodapedon huxieyi.

sp.
Pachygonta incurvata.
Belodon, sp.
Parasuchus hislopi.

Massospondylus, sp.

PLANTA.

ConirERé&.
t Palissya conferia.
* 4 jabalpurensis.
*T 4, tndica.
F Cheirvolepis, cf. muenstert.
* Avaucarites cutchensis.

1 Memoirs, XVII, 267, (1881).
186 GEOLOGY OF INDIA—GONDWiNA System. (Chap. VIL

In the list an asterisk prefixed to a species shows that it is also known
from the Jabalpur group, a dagger that it is found in the Rajmahdl group of
the Rdjmahdl hills, or, inthe case of Checvolepis muenstert, in the Golapilli
beds. .With the exception of the last mentioned species, all the plants are
found in the Sripermattr or some of the other groups of the same age on
the east coast. We may consequently regard the Kota-Maléri beds as
somewhat later in age than the Rajmahdl group, and nearly equivalent to
the Sripermatir, though possibly somewhat newer.

The paleontological relations of the animal remains will be treated in
the next chapter, and all that need be noted here is the occurrence of the
Panchet form Pachygonia incurvata, The animal is believed to be speci-
fically,! certainly generically, identical with that of the PAnchet group, and
its presence here along with a flora which indicates a much newer age is
remarkable,

The discovery of animal remains in the Denwa group of the Sdtpuras
and south Rewa has been already noticed. In the former area they are
represented by scutes and vertebra of Parvasuchus and Mastodonsaurus,
in the latter by remains of two species of Hyperodapedon and Parasuchus.
The material is not sufficient for establishing those specific identifications
which would alone allow us to assign the groups to the same horizon, but
the resemblances are suggestive and the position of the Denwa group,
below the Jabalpur and separated from it by the Badgra group, places it
stratigraphically on very mucli the same horizon as is indicated by the plant
fossils of the Kota-Maléri group. ~

Resting on the Kota-Maléri beds is a group, which was separated by
Dr. King? under the name of Chikidla, from a village of that name situated
close to their boundary, though actually upon the Kota sandstones. They
extend along the eastern side of the Gondwana outcrop, with a width of
eight to ten miles, for nearly seventy miles, from Rebni in the Wardha basin
to the reach of the Godavari below Enchapalli.

The group is composed of soft sandstones and heavy bands of conglo-
merates of white quartz pebbles. Clay bands and seams of shale are fre-
quent and the group is very ferruginous, the iron ore is collected and
worked into iron to a considerable extent.

The relations of this group to the Kota-Maléri are obscure, but there
appears to be a slight unconformity. It was believed by Dr. King, on
the ground. of its lithological similarity, to represent the Tripetty sand-
stones of the coastal region—a correlation which is not inconsistent with
its relations to the Kota-Maléri group.

Lydekker, Records, X, 34, (1877); XV, | 2? Memoirs, XVIII, 140, (1881).
25, (1882),
   

Otozamites gracilis, Schimp, Otozamites hislopi, Old.

 

Bracbyphyllum mammiliare, L.& H.

  

Otozamites gracilis, Schimp. Otozamites hislop, Olah. Palissya indica, (ldh.

  

Palissya jabalpurensis Fstm, Alethopteris medliottiana, Oldb.
Podozamites lanceolatus, Schimp,

 

Podozamites lanceolatus

var, spatbulatus F'stm,

Calcutta Phototype Co.,

JABALPUR FOSSIL PLANTS.
JABALPUR GROUP.

Chap. VIL.) 187

In the first account of the central portion of the Narbada valley, a
group of rocks was distinguished as ‘upper Damuda.’ It was, however,
pointed out at the time that this group was not only unconformable to
the ‘lower Damuda,’ but that it contained a very different flora. When
a true upper Damuda group was subsequently found in the Raniganj coal-
field, it became desirable to distinguish the Narbadé beds by a different
name, and as they are well developed in the immediate vicinity of Jabal-
pur, they have been named from that town.

The Jabalpur group consists of clays, shales, and earthy sandstones,
with some thin beds of coal. The clays and soft shales, which are the
most characteristic beds of the formation, are pale coloured, usually white,
pale lavender grey, or pale red. The sandstones are generally coarse and
conglomeratic. Carbonaceous shales are met with in several places, and
occasionally one or more thin bands cf jet-coal, very different in character
from the coal of the Damuda formaticn. Limestone is rare. At the base
of the formation, when resting upon gneissic rocks, there is frequently
found a coarse, compact sandstone, so hard and compact as almost to
resemble a quartzite. It is often conglomeratic, and the matrix contain-
ing the pebbles consists of white earthy rock in a porcellanic condition.
Occasionally, but rarely, this bed is calcareous.

The thickness of the Jabalpur group does not appear to have been
determined with any accuracy. It is, however, of no great vertical extent,
and so far as is known nowhere exceeds 1,000 feet. The relations of
the Jabalpur group to the underlying Mahddevas have not been examined
in detail, but they appear to be generally conformable.

The following is a list of the fossils found in this group, those found
also in the Umia beds of Cutch being marked with an asterisk, whilst those
met with in the Rajmahdal group are distinguished by a dagger *—

FiricEs— CYCaDEACEE—

“+ Sphenopteris, cf. avguta. Podozamites lanceolatus.
Dicksonia, sp. A spathulatus. a
Alethopterts lobifolia. 3 hackett.

Py medlicottiara, Otogzamites hislopi.

eh whitbyensis. i gracilis.
Macroteniopteris satpurensis. 55 distans.
Glossopteris, cf. communis, os angustatus. -

Sagenopteris, sp. Pterophyllum nerbuddaicum.

1]. G. Medlicott, Memoirs, 11, 176, (1860).
The Jabalpur formation was at this time not
clearly distinguished in places from the Maha-
devas, the former being supposed to be the
lower; in reality the Jabalpur formation is not
only newer than the Mahadeva, but it appears
to be the latest member of the whole Gond-

 

wAna series, with the possible exception of
the Umia group of Cutch, or of some of the
uppermost groups of the east coast. Further
accounts of the Jabalpur group will be found
in Records, 1V, 75, (1871); and Memoirs, X,
142, (1873).

2 Pal, Indica, series xi, II, 83, (1377).
188 GEOLOGY OF INDIA—GONDWANA SYSTEM. C Chap. VIL.

' Cycapraceas— contd. Contrer£-- contd.
* + Ptilophyllum cutchense. * Araucavites cutchensis.
T+ 3 acutifolium. Brachyphyllum mammillare.
+ Williamsonia, cf. gigas. 4 Echinostrobus expansus.
Cycadites, ch. gramineus.. Echinostrobus rhombicus.
Taxites tenerrimus.
ConIFERE— Gingto lobata.

Phénicopsis, sp.

* + Palissya indica. Edis bsp
» Sp.

»  jabalpurensis.

It will be seen that nearly as many RAjmahdi as Umia species, five of the
former and six of the latter, are found in the Jabalpur group, so far as the
flora has hitherto been determined. It should, however, be remembered
that the known species of the Rajmahdl flora are nearly fifty in number,
while those of the Umia flora are much less numerous, about twenty-two.!
Moreover, the Jabalpur beds are distinguished by a conspicuous want of
raany of the commonest and most characteristic RAjmahdl plants, such as
the broad leaved species of Pterophyllum.

On the whole, the Jabalpur beds are probably on nearly the same
horizon as the Umia beds of Cutch, but possibly represent a period inter-
mediate between the Umia and Rajmahdl groups, though nearer to the
former. At the same time the circumstance that no representative of the
Jabalpur flora has yet been found on the east margin of the Indian Penin-
sula, to which the RAjmahdl flora is confined, suggests that the distinction
mav be due to the beds having been formed in different botanical regions.
Bearing in mind, however, the large amount of evidence which exists to
show that the greater part, if not the whole, of India proper was a land
area in Gondwana times, this idea of the country having been divided into
distinct botanical regions is less probable than the theory of a difference
in age between the Rdjmahdal and Jabalpur groups.

The plant bearing beds of the Umia group in Cutch are only mentioned
here because of their relations to the uppermost beds of the Gondwana
series. The name Umia is derived from a village about 50 miles north-
west of Bhuj, the chief town of Cutch. The group will receive a fuller
description -under the head of the jurassic formations, and an account will
there be given of its mineral character and animal fossils.*

The special interest of this group in connection with those just enu-
merated is due to the fact that beds containing plants, several of which
are identical with those of the Jabalpur beds, are interstratified with rocks
yielding marine fossils.

1 Dr. Feistmantel enumerates twenty-eight | and others stems not identified generically.
in his Memoir, but some are only varieties 2 Infra, pr 223.
Chap. VIL]

UMIA GROUP. 189

The following is a list of the plants from the Umia beds,! a dagger in-
dicating species fcund in the Jabalpur group :—

ALGE— CycapesCe&z—contd,
(?) Chondrites dichotomus. ‘ Otozamites contiguus.
FiLices— sg tmbricatus.
Oleandrid:um vittatum. % cf. goldiei.
Teniopteris densinervis. Cycadites culchensis.
fAlethopteris whitbyensis. Cycadolepsts pilosa,
Lecopteris tenera. Williamsonia blanford:.
Pachypleris specifica. ConirERE—
. brevipinnata. Palissya bhoopoorensis,
Actin op teris, sp. t ue cf. indica.
CYCaDEACER — s cf. laxa.
tPtilophyllum cutchense. Pachyphyllum divaricatum.
t is acutifolium. tEchinostrobus expansus.
” brachyphyllum. tAraucarites cutchensis.

 

At a somewhat lower horizon in the rocks of Cutch, a few plants have
been found near a village named Narha, in the northern part of the prov-
ince, in beds interstratified with the Katrol group, the Cephalopoda of
which are considered by Dr. Waagen as corresponding to those of the
Kimmeridge and Upper Oxford beds of Europe. These plants consist of
the following species ?:—

Sphenopteris, cf. arguta. Otozamites, cf. contiguus..

Alethopteris whitbyensis. Avaucarites cutchensis.

The three last are apparently identical with species found in the Umia
beds, whils: Sphenopteris arguta isan English lower oolite species, found
also in the Jabalpur and Rajmahd4l groups. The Alethopterzs and Arau-
carttes are also Jabalpur forms. This evidence, so far as it goes, tends
to show a great persistency in the flora, and it may indicate that the
Jabalpur beds are a little older than the Umia group, since the connection
of the flora found in the Katrol beds of Narha with that of the Jabalpur
group is quite as strong as with the Umia plant fossils,

In northern Kathidwdr a series of soft white and ferruginous sand-
stones, with pebbly bands, is exposed, of which a few fossil plants have
been obtained, which are comprised in the following list,3 an asterisk and
dagger indicating species that are also found in the Umia and Jabalpur
groups respectively :— ;

FILices—-

*® + Alethopteris whitbyensis.
Pecopteris, sp.
Tectopteris, sp.

CycaDEACE x —
+ Podozamites lanceolatus.

* + Ptilophyllum cu chense.

! Pal. Indica, series xi, 11, 63, (1876).
2 Memoirs. VX, 213, (1872); Pal. Indica,

ConiFERE—
T Palissya jabalpurensis.
T Taxttes tenerrimus.
* + Echinostrobus expansus,
* + Avaucarites cutchensis.

series xi, [I, 80, (1876).

3 Memoirs, XXI1, 81-82, (1884).
190 GEOLOGY OF INDIA ~GONDWANA SYSTEM. (Chap. VIL

The Kdthidwdr beds have been regarded as the equivalents of the Umia
group in Cutch, and the lithological resemblance and geographical prox-
imity are certainly in favour of correlating them with the only group, of the
more extended series, which appears to have been deposited under similar
conditions, but four of the seven species of plants recognised are found both
in the Umia and Jabalpur groups and three in the latter alone. The
palzontological relationship is consequently closer with the more geo-
graphically remote beds, and if the Kathidwdr sandstones are the equi-
valents of the Umia group, they indicate a greater approximation in age
between the latter and:the Jabalpur group than a direct comparison of the
two floras would necessarily imply.
  

Araucarites kachensis, F'stm, Ptilophytlum catehense, Morr,

   

Oleandridium vittatum, Bégt, Alethopteris whitbyensis, Bgt.

 

Prilophyilum cutchense, Morr. Pachyphyllum divaricatum Bunb, Oleandridium vittatum, Béget,

Calcutta Pnototype Co.

CUTCH (UMIA) FOSSIL PLANTS.
CHAPTER VIII.

HOMOTAXIS OF THE GONDWANA SYSTEM,

Comparison of Gondwdna and European floras—Palzontological relations of the animal re-
mains —Asiatic representatives of the Gondwdna system—Representatives of the Gond-
wana system in Australia atid in South Africa—Correlation of the rock groups in the four
continents—A former land connection between India and Africa—Bearing on the doctrine
of the permanence of oceans—and on the hypothesis of secular variations in latitude.

In the preceding chapters all reference to the very interesting question
of the age of the Gondwana system, as compared with the geological se-
quence in Europe, has been purposely omitted. The subject is one whose
interest and complexity deserves a special treatment, and it would have
been impossible to deal with its bearing on several of the most important
problems of theoretical geology that are still unsettled, without interrupting
the general description of the stratigraphy of the Gondwdna system.

The writer of these pages is fortunate in the fact that the bitter con-
troversy which long raged over this question, is now practically extinct.
The history of this controversy would be an interesting one, showing,
as it does, how truth is ultimately arrived at by the contact of conflicting
opinions, each involving an element of falsehood and each containing a
large amount of truth. But the purpose for which this book has been
written, and that for which it will principally be consulted, is the state-
ment of the present state of our knowledge, and a recapitulation of the past
would be a task as uncongenial in its execution as unprofitable in its re-
sult. The main points for which the members of the Geological Survey have
contended were accepted at last, even by the talented palzontologist who,
alone among his colleagues, disputed them, and the last smouldering
embers are mere minor and unimportant differences of opinion as to the
exact position of certain individual groups of the Gondwana system.

The most obvious method of determining the age of the rock groups of
the Gondwdna system would be a direct comparison of the fossils they
contain with those of Europe. Unfortunately this method leads to very
unsatisfactory and inconclusive results,

If we take the flora of the Damuda series we find, according to Dr.

4!
192 GEOLOGY OF INDIA—GONDWANA Homoraxis. [Chap. VII,

Feistmantel, the following species identical with, or allied to, European
forms ?:—

PERMIAN—
Macroteniopteris feddent, Fstm., allied to M. abnormis, Gutg.

Trias —

Schisoneura gondwanensis, Fstm., allied to S. paradox 1, S-himp.
Merianopteris major, Fstm., allied to M. angusta, Heer.

Jura oF Russia anp S1BERIA—
Phyllotheca indica, Buub., allied to P. sibivica, Heer.
1. robusta, Fstm., allied to P. stschurowskit, Schmalh.
Cyathea, cf, tscihatchefi, Schmalh., probably identical.
Dicksonia hughesi, Fstm., allied to D. saportana, Heer.
Samaropsis, cf. parvula, Heer., probably identical.
Rhipidopsis, cl. gingkoides, Schmalh., probably identical.

while Belemnopteris woodmasontana is said to closely resemble the living
form Hemionot?s cordata found in Southern India.

If we take the Rajmah4l flora the evidence is no less conflicting. The
following list gives the relationships with European forms as determined
by Dr. Feistmaatel? :—

PERMIAN—
Macroteniopteris latu, O. & M., allied to M. abnormis, Gutbg.
Pterophyllum carterianum, O. & M., allied to Pt. blechnoides, Sandbg.

Trias—
Gleichenta bindrabunensis, Schimp., allied to G. gracilis, Heer.
Dancopsis rajmahalensis, Fstm., allied to D. marantacea.

:

Raetic—
Equisetum vajmahalense, Schimp., allied to Z. muensteri, Stbg.
Thinnfeldia salictfolia, O. & M., allied to T. decurrens, Schenk.
Alethopterts indica, O. & M.,, allied to Asplenites rossertt, Schenk.
Angiopteridium maclellandi, O. & M., allied to A. muensteri, Gépp.
Macroteniopteris lata, O. & M., allied to M. gigantea, Schenk.
» cf. propinguum, Gopp , probably identical.
Pterophyllum fissum, Fstm., allied to P. comptum, L. & H., and minus, L. & H.
aistans, Morr., allied to P. braunianum, Gaon, ,

Pierapiyllam princeps, O. & M,, allied to P. brauni, Schenk.
Otozamites, cf. brevifolia, Brgt., probably identical.
Palissya indica, Fstm., allied to P. braunt., Endl.
Cheirolepis, cf. muensteri, Schimp., probably identical.

3 gracilis, Fstm., allied to C. muenster7, Schimp.

Lias—
Equisetum rajmahalense, Schimp., allied to £. liasinum, Heer.
Angtopteridium macclellandi, O. & M., allied to A. muenstert, Gopp. -
Cycadites vajmahalensis, Oldh., allied to C. linearis, Stbg.

> Pal. Indica, series xii, III, pt. ii, (1880),1 ? Pad. Indica, series ii, I, pt. ii, (1$77) .
Chap. VIII. AFFINITIES OF THE GONDWANA FLORAS. 193

OoultTE—

Sphenopteris arguta, L. & H., identical.

AHymenophyllites bunburyanus, Fstm., allied to Tympanophora racemosa, L. & H.

Alethopteris indica, O. & M., allied to A. whitbyensis, L. & H.

Pterophyllum fissum, Fstm., allied to P. minus, L. &. H., a Rhezetic species also,

Williamsonia, cf. gigas, Carr., probably identical.

Araucarites macropterus, Fstm., allied to A. brodiei, Sap.

These two floras were regarded by an eminent paleontologist as
indicating a triassic and a liassic age respectively, but the most striking
points about them are, firs/ly, the want of definite evidence of a difference
of age corresponding to the great stratigraphical break and palzontologi-
cal contrast between the two groups and, second/y, the extremely hetero-
geneous nature of the alliancesexhibited by the flora. For the last of
these an explanation will be found in the sequel, but. it is also largely due
to the absence of any true test of relationship in fossil plants. The shell
of a marine mollusc, the test of a crustacean or the cup of a coral give
real clues to the zoological position of the animal they once formed parts
of, but leaves or, in the case of ferns, fronds, either barren or with the fructi-
fication too obscure to be determinable, are all that we have when dealing
with fossil plants, and these, which have the least weight in determining the
relationships of living plants, are often only imperfectly preserved, When a
number of leaves are found, all showing the same shape and venation,
there is a considerable probability that they belonged to the same
species, but when small differences are observed, which lead to their being
classed as belonging to-distinct species, there is no certainty that they did
not belong to plants widely separated from eéach other in all important
characteristics, while plant remains that are classed under distinct genera
or even families may have belonged, not merely to closely allied species
but may have formed different parts of one and the same plant. It may
consequently result that a plant fossil may most resemble one of a very
different age, which possibly would prove to be widely distinct from it did
we but know the whole of both.

The alliances of the PAnchet flora are shown in the following table! :—

Trias—

Schizoneura gondwanensis, Fstm., allied to S paradoxa, Schimp,
Ru a£T1Ic—

Pecopteris concinna, Presl., identical.

Cyclopteris pachyrhaca, Gépp., identical.

Oleandridium, cf. stenoneuron, Schenk., probably identical.

Thinnfeldia odontopteroides, Morr , allied to T. rotundata, Nath.
Jura—

Samaropsis, ck. parvula, Heer.

In the last case, however, the resemblance is of little value.

! Pal. Indica, series xii, III, pt. ii, 51, (1880).
194 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [Ohap. VIII.

If we pass upwards to the Jabalpur and Umia groups we find a much
more homogeneous flora, as is indicated by the following statement of the
alliances with European fossil plants :—

ALLIANCES OF THE JABALPUR Frora.!
Lras =

Otozamites gracilis, Schimp., identical.

Lower OoLiTE—

Sphenopteris, cf. arguta, L, & H., probably identical.
Alethopterts lobifolia, Schimp., identical.

3 whithyensis, Gopp., identical.
Podoszamites lanceolatus, L.& H., identical.
Williamsonia, cf. gigas, probably identical.
Cycadites? cf. gramineus, Heer., probably identical.
Avaucarites cutchensis, Fstm., allied to A. phillipsi, Carr.
Brachyphyllum mammillare, L. & H., identical,
Echinostrobus expansus, Schimp., identical.

ALLIANCES OF THE Umia Fiora?
RyzTic—

Oleandridium vittatum, Schimp., identical.
Aetincpteris, sp. allied to A. peltata, Schenk.
Lower Oo1LirE—
Oleandridium vittatum, Schimp., identical,
Alethopteris whitbyensis, Gépp., identical.
Pachypteris specifica, Fstm., allied to P. lanceolata, Brgt.
Otozamites, cf. goldiet, Brgt., probably identical.

4 imbricatus, Fstm., allied to O. bronguiarti, Sap.
Cycadites cutchensis, Fstm., allied to O. samioides, Leck.
Williamsonia blanfordi, Fstm., allied to W. sp, Carr.
Taxites cf, laxus, Phill. probably identical.
Pachyphyllum divaricatum, Bunb., identical.
Echinostrobus expansus, Stbg., identical.

Araucarites cutchensis, Fstm., allied to A. brodiei, Carr.

Urrer Jura (Kimmeridge ?)—
Gycadolepsis pilosa, Fstm., allied to C, hirta, Sap.

Reviewing the evidence of the Gondw4na plants, we find that, the two most
important of the floras, those of the Damuda and Rdjmah4l series, do not
show a definite relation to any single horizon of the European sequence,
nor do they show any distinct evidence of a difference of age. The flora
of the Panchet group has a much more defined relationship to the rheetic,
and the beds might have been referred to this age on the evidence of the
plants alone, were there not other comsiderations, to be detailed below
pointing to an older date. The Jabalpur and Umia floras show a still

‘Pal. Indica, series xii, II, €3,(1877). 1 ? Pal. Indica, series xii, Il, 63, (1876).
Chap. VIILJ AFFINITIES OF THE GONDWANA REPTILES. 195

greater definiteness of relationship, in this case to the lower oolite, and
though the latter group was described as newer than the former, the
difference in age of the two is possibly not so great as to introduce any
difficulty. The Cutch plant beds are, however, found resting on marine
deposits whose Cephalopoda show that tuey are of uppermost oolitic age, and
are overlaid conformably by beds containing upper neocomian ammonites ;!
there is consequently a direct conflict here between the evidence of the
marine mollusca and the fossil plants, and the question of which is to be
preferred arises.

This will not be discussed here in detail, but the explanation is to be
found inthe diversity of the forms of terrestrial life inhabiting distant,
regions of the earth at the present day. There is a much greater difter-
ence between the terrestrial faunas and floras of Africa, Australia and
America than between the animals inhabiting the Atlantic, Indian and
Pacific oceans, and it is a common circumstance to discover fossil remains
of animals and plants, without any living representatives in neighbour-
ing lands, but allied to forms still living in a distant region. Such was
also the case during the Gondwdna epoch, and, as will appear, the distinc-
tions at its commencement were even more trenchant than at the present
day.

If we turn from the plants to the animal remains found in the Gond-
wana system, the evidence is little less ambiguous. The Gondwanosaurus
from the Bijori group belongs to the family Archegosauridz, which in
Europe is principally carboniferous and permian, though a specialised form
ranges into the trias. The affinities of the Indian specimen are said to be
permian, On the other hand, Brachyops laticeps, from the Mdngli group,
which is believed to be of about the same age as the Bijari group, is
in Europe only allied to Rhkinosaurus, a jurassic form.? Its uearest allies
are to be found in the Karoo beds of South Africa,

The reptiles of the Pdnchet group exhibit but little connection with
European forms, and their connection with the South African fauna will be
noticed further on.

In the Kyta-Maléri and Denwa groups the genera Belodon, Hyperoda-
pedon, and Mastodonsaurus are all represented in the upper trias of
Europe, the first and last being also known from rhetic beds, Parasuchus
belongs to the same typically mesozoic group of crocodiles with biconcave
vertebrz as Belodon, and is placed withit, by Prof. Huxley, in a section of
the family which is almost confined to triassic rocks in Europe. Of the
fishes represented in the Kota-Maléri group the genus Lepfdotus ranges
from the lias to the lower chalk, and the Kota species were regarded

1 Infra p. 286. 3 Quart. Four, Geol. Soc., XXXI, 427, (18750

2 Pal. Indica, series iv, 1, 13, (1885).
O02
196 GEOLOGY OF INDIA—GONDWANA HoMOTAXIS. [Chap, VITL

by Sir P. Egerton, as showing liassic or oolitic affinities, Tetragonolepis
is only known from liassic beds, and Dapedius is also a liassic genus.
Ceratodus is principally triassic, but species have been found in beds of
later date, and the genus is still living in Australia.

Here the fauna of the Maléri group indicates an earlier age than that
of the Kota, and in this agrees with the relative stratigraphical position of
the two groups, but the fossils indicate a much greater difference in the
age of the two rock groups than their intimate stratigraphical association
suggests, and we have very much the same paleontological contradiction
as there is between the land plants and marine animals of the Cutch
jurassics.

From the coalfields of Tongking a fossil flora has been described * which
contains certain Gondw4na forms, and exhibits a much closer relation toa
definite European horizon than any of the groups of that system. Out of a
total of nineteen species, ten are found in the rhetic beds of Europe, and
eight of the remainder in India; of the latter Phyllotheca tndtca,
Palzovittaria kurzt, Macroteniopteris feddent, Glossopteris browniana,
and Noegerathiopsis hislopi, are Damuda forms, and the R4jmahdl and
Sripermatér groups are represented by Angioptertdium spathulatum,
Tzxniopteris ensts and Otozamites rarinervis, Here the evidence, so far
as it goes, is distinctly in favour of regarding the Tongking beds as rhetic
in age and intermediate oe Damuda and R4jmahdl series in India,

or more or less contemporaneousjwith the Pdnchets, whose flora has also a
. . " iN
thetic facies. :

Outside the limits of India proper, in north-western Afghdnistdn, a
series of coal bearing sandstones, intercalated with marine beds, and having
at its base a boulder bed precisely similar in character to that of the

Talchirs has been described by Mr. Griesbach.1 The general classification
of the beds as adopted by him is as follows :—

 

 

 

 

Age, Formation, Localities,
Jurassic . . . | Densely red grits and sandstone,
shales with plant remains. Upper Almar stream near Pain-
Trap. guzar; Astar-ab below Paisnih.
Dark bluish grey grits and sand- |( Khorak-i-Bala, north of the Kara
stone; plant remains. Koh,
Ash-beds,

Sandstone and black alum shales | Dodb north of the Kara Kotal.
tons plant impressions; marine
ossils.

3 Pal. Indica, series iv, I, 2, (1875), ‘‘Ganoid | details in the text are taken from Pal. Jndica,
Fishes from the Deccan.”

? series xii, IV, Introduction, pp. xv-xvii, (1886).
2R. Zeiller: ‘‘ Examen de la flore fossile des | % Rec a 1X f 5
couches de charbon du Tongking, Aznales ot ae

s i. red grits were afterwards said to be nevco-
des Mines, 8th series, II, 299, (1882). The | mian. Records, XX, 94, (1887). x
Chap. VIL. J GONDWANAS OF AFGHANISTAN, 197

 

 

 

Age. Formation, Localities,

 

Upper Trias or Rhz- | Light coloured sandstones and | Kotal-i-Sabz (north slope of Kara
tic. | shales with coal seams. Koh), Shisha Alang.
Upper .| Great thickness of marine sand- | Chahil; Shisha Alang.
stone, limestone, and shales with
coal seams. Schisoneura sp., etc.
UpperTrias ¢ Middle. | Brown sandstones and shales with | Chahil, north slope of Kotal-i Sabz.
coal seams. FAguisetites colum-
naris.
Lower .| Marine sandstone and limestone | Chahil.
beds. Halobia lommeli, Monotis
salinaria,

 

 

Permo-Carbon . . | Altered shales (mica-schist, etc.) | Saighdn; Ak Robdét Kotal, north.
with graphitic and anthracitic : :
seams. Clay shales with impure
coal, The whole traversed by
hornblendic granite.

Coarse conglomerate in greenish | Pali Kotal and gorge; Ak Robiat.
matrix, altered by granite.

Massive dark limestone with bra-| Ditto ditto ditto.
chiopod casts.

 

 

 

There can be little doubt that these beds are the equivalents of the
Gondwéna system of the Indian Peninsula and will be of the greatest im-

portance in determining the age of the various members of that system,
when they have been more fully studied, At present the subdivisions of the
series of beds seen in Afghdnistén cannot be correlated with those of the
Gondwdnas in detail, while as regards the correlation of the beds with the
European sequence, it must be remembered that no fossils from this area
have been critically determined, and the correlation, depending merely on a
field determination of two or three species, may be upset by a fuller
study of the fauna and flora as a whole. It is necessary to bear this in,
mind as the horizon indicated by Mr. Griesbach, for the coal bearing beds
of Afghdnistdn, is higher than that which we shall have cause to regard as
the horizon of the Barakar or Rdniganj groups, their probable equivalents
in India. The country in which these Afghdn Gondwdnas are deyeloped i is
not open to detailed examination by Europeans, such information as is avail-
able having been obtained during rapid journeys through the country, and
until a closer and more detailed examination of them can be made they have
not the importance, from the present point of view, that their proximity to
India, and the occurrence of beds containing marine fossils, intercalated with
those which contain the fossil plants, would otherwise give them.

If, instead of looking to the west or the east, we turn to the south, we
198 GEOLOGY OF INDIA—GONDWANA HoMOTAXIS. [Chap. VIII,

will find in Australia a series of beds which clears up the vexed question of
the homotaxis of the Gondwdnas in a wonderful manner.

From Bacchus marsh in Victoria three species of Gangamopteris have
been obtained, but of this very limited fauna one species is identical with,
and the other two allied to, Karharbdri forms. This would in itself suggest
a correlation of the Bacchus marsh with the Karharbdri beds, and there is
further evidence in the presence of large blocks of granite and pebbles of
rocks that must have travelled long distances, imbedded ina fine grained
matrix of mud. Like the ‘Talchir boulder bed, the beds are of glacial origin,
and this, combined with the paleontological evidence, justifies us in
regarding them as the equivalents of the Indian Talchirs.

By itself this would be of little importance, as the Bacchus marsh ex-
posure is small and the beds cannot be palzontologically connected with any
others of known age. But there is in New South Wales a much better and
more complete section, which is in fact the standard one for Australian
geology, so far as the period we are dealing with is concerned. The
sequence has there been divided into the following groups :—

. Wianamatta shales.

Hawkesbury sandstones.

Newcastle beds, or upper coal measures.

. Upper marine beds with carboniferous fauna.

. Stony creek beds, or lower coal measures. { Muree beds.
Lower marine beds, with carboniferous fauna.

4Ku74
ry eaaga

The marine beds are important on account of the fossils they contain,
and because they give us a fairly detinite geological horizon to start from.
Their equivalence to the glacial beds at the base of the speckled sandstone
in the Salt range, and the close alliance of the fauna of these last named
with that of the Australian marine carboniferous beds, has already been
referred.to,” and need not be recapitulated here,

The only paleontological evidence bearing on the correlation of the
Bacchus marsh beds, with the sequence in New South Wales -is the
occurrence of Gangamopteris angustifolia in the Newcastle beds. But
there is weightier evidence of a different character, which renders it certain
that the marine beds, and not the overlying Newcastle beds, are the true
equivalents of the glacial boulder clays of Bacchus marsh. ,

The marine beds, and especially those immediately associated with the
lower coal measures of Stony Creek are composed of a fine grained matrix
of sand or shale, enclosing numerous delicate Fenestelle and bivalve shells
with their valves still united, which had lived, died and been tranquilly
preserved where they are now found, thus proving, as conclusively as

’ Report on the geology of the district of | p. ro.
Ballan by Richard Daintree, Melbourne, 1866, 2 Supra, p. tan.
Chap. VIII. ] COAL MEASURES OF AUSTRALIA. 199

the texture of the matrix in which they are preserved, that they could
never have been exposed to acurrent of an great rapidity. Scattered
through this matrix there are are numerous more or less subangular
blocks of stone, of all sizes, ranging up to several feet in diameter, some
of which exhibit most characteristically developed glacial striz. It was
not mere velocity of current that brought these fragments, and deposited
them where they are, for to move even the smallest of them would require
a current that would have swept away the matrix in which they are
imbedded, and destroyed the delicate fossils with which they are asso-
ciated. They must have been floated to their present position and dropped
on to the bottom of a tranquil sea, andtaking into consideration their
abundance, as well as the distinct traces of glacial action that some
of them exhibit, the only agency than can be appealed to is that of floating
icebergs.!

We find then that the marine carboniferous deposits of New South
Wales were formed during a period of exceptionally cold climate, and it is
to the latter that we must look for an equivalent of the glacial beds of
Bacchus marsh rather than to the overlying Newcastle beds, which indicate
a more temperate climate at the time of their formation.

The lower coal measures consist of a comparatively thin band of sand-
stones and coal, intercalated between the lower and upper marine beds.

~The fact of this intercalation has been questioned, on account of the
supposed mesozoic age of the flora obtained from them. But no one who
had actually examined them in the field doubted the intercalation, and
itis now too well established to be questioned. Under these circumstances
it is important to see what are the plants which co-existed with a marine
fauna of carboniferous type. The following is the list given by Dr.
Feistmantel ? :-—

Phyllotheca australis. Glossopterts elegans.

Annutaria australis. ss primeva.

Glossopteris browniana. Neggerathiopsis prisca.
i clarket,

At a glance the flora can be seen to be of the type of the lower Gond-
wAnas, in India, and to differ totally from the European flora of corre- }
sponding age. The same alliance with the lower Gondwéna fiora is to be [
seen more conspicuously in the following list-of the plants of the Newcastle
beds:—

Phyllotheca australis. Sphenopteris alata,
Vertebraria australis. 3 flexuosa,

! Records, XIX, 39, (1886); Quart. Four, ( and notices of previous literature, is given by
Geol. Soc., XLIII, 190, (1887) the same author in Sitzungsber. K. b6hm, Ges-

2 O. Feistmantel, Paleontographica, supple- | Wiss. 1887, pp. 55-77, and in Four. Roy, Soe,
ment, 1878-79. A resumé—, with lists of fossils | New South Wales, X\V, 103, (1881).
200 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [Chap. VIII

Spienopteris germana. Glossopterts reticulum.
” hastata, ” teniopteroides.
” . Labifolia, 35 wilkinsont.
’ plumosa, Gangamopteris angustifolia,
Glosiopteris browniana. ¥5 clarkeana,
= ampla. Caulopteris (2) adamsi.
e cordata. Zeugophyllites elongatus.
» elo gata. Néggerathiopsis media.
i linearis. spathulata.
” parallela, Brachyphyllum australe.

Of these, Gangamopteris angustifolia and Glossopterts browntana are
also found in India, the former in the Talchir and Karharbdri groups, the
latter in the Damuda series. ‘Ihree species of Glossopteris, G. linearts, G.
ampla, G. parallela, are represented in the Damuda flora by the allied forms
G. angustifolia, G. communis and G. damudica. The Australian Phyllotheca
is very closely allied to, and has been considered identical with, P. indica
of the Damudas, Vertebraria, which is common in both the floras, is only
known elsewhere by a distantly related species from the jurassic of Siberia.
These specific relationships between the Damuda and Newcastle flora are
strengthened by the general resemblance in the type of the floras as a
whole. All the principal and more characteristic genera of the Newcastle
beds are represented in both floras, and the genus Glossopterzs in both
cases includes about one-third of the total number of species,

The paleontological evidence would of itself be almost sufficient to
justify the correlation uf the Newcastle and Damuda series, and when we
bear in mind that their position, relative to the underlying glacial beds, is
similar in both cases, it becomes certain that the two coal bearing series
must be more or less completely the equivalents of each other.

Above the Newcastle comes the Hawkesbury group of sandstones and
shales. The few fossils that have been obtained from this and the over-
lying Wianamatta group are scarcely sufficient to establish their correlation
with any particular division of the Gondwana system, none of the species
being found in India except Thinnfeldia odontopteroides, a Panchet form,
but there is evidence of a recurrence of glacial conditions in the Hawkesbury
beds which is worth noting. Large angular fragments of shale, similar
to that interbedded with the sandstones, are found imbedded in atconfused
manner, with their original bedding planes lying at all angles, in a matrix
of sand. They occur nearly always immediately above the shale beds,
and are accompanied by well rounded quartz pebbles! It is difficult to
account for the facts that have been described without the agency of ice, in
one form or another, but they are in no way comparable with the proofs

1C. S. Wilkinson, Four, Roy. Soc. New South Wales, XIII, 105, (1880).
Chap. VIII] EVIDENCE OF COLD IN THE PANCHET GROLP, 201

of glacial action exhibited by the marine beds. The evidence indicates the
action of winter ice rather than of actual glaciers.

No evidence of the recurrence of glacial conditions has been recorded
in India, but there is an indication of a return of cold in the undecomposed
felspar, found in the sandstones of the Pdnchet group. Undecomposed
felspar is characteristic of the sandstones associated with the boulder beds
of the Talchirs, and Prof. Green! nas remarked on the extreme freshness of
the felspar in the glacial beds of the same age in South Africa. The beds of
glacial origin in the Indus valley, which have been supposed to be eocene?
also contain an abundance of fragments of undecomposed felspar. Apart
from these observations there is an inherent probability that sandstones
containing undecomposed felspar would be found in cold climates. They
mean that the disintegration of the parent rock from which the material was
derived, together with the transport and final accumulation of the debris,
went on at a greater rate than chemical decomposition of the constituent
minerals, and this might be due either to extreme dryness, which would
retard the rate of decomposition, or to an extreme severity of climate,
which would accelerate the rate of disintegration.

Taking these considerations into account, it may well be that the un-
decomposed felspar of the P4nchet sandstones indicates a recurrence of a
cold period, less severe than that of the Talchir, and comparable to that of
the Hawkesbury group in New South Wales. Quite independent of this
the stratigraphical position of the two groups would suggest their correlation,
and the flora, which contains a mixture of purely lower Gondwana with
upper Gondwana genera, is consistent with this conclusion.

Newer than the Hawkesbury beds a number of groups of plant bearing
sandstones are found in different parts of eastern Australia, which probably
represent the upper -Gondwdnas of India, Their relations to each other
have not been fully worked out, and the recorded fossils indicate an
admixture of specimens obtained from older beds. In any case they are of
little importance in the present connection, and will not be further noticed.

In South Africa we again find a representative of the Gondwana flora
which, though perhaps less valuable for the purposes of establishing the age
of the Indian formation than the Australian beds, is otherwise of great interest
and importance. The Karoo series, as it is now generally called, consists
of a thickness of many thousands of feet of sandstones and shales, with
interbedded coal seams. Like the Gondwdnas, they were at first regarded
as lacustrine in their origin, but are now looked upon by many geologists
as having been deposited by rivers, Whether lacustrine or river deposits,

" Quart. Four, Geol. Soc., XLIV, 244, (1868). | 2 Infra, p. 346.
202 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [Ohap, VIII

they were certainly formed by fresh water, and the coal resembles that of
the Damudas in its laminated structure and in the absence of an underclay

 

- Fig. 13.—Geological sketch of South Africa.

G, granite and gneiss. S, silurian. B, porphyritic intrusions. C, carboniferous. Black,
Ecca-conglomerate. K, Karoo formations and overlying trap. White (in the neighbourhood
of Algoa Bay and Cape Agulhas) Uitenhage and younger marine deposits.

or even roots penetrating the underlying beds. More important than these

“mineralogical resemblances are the similarities shown by the fossil flora and
fauna of the two countries, as will be described immediately. But, first, it
will be well to notice briefly the distribution and relation to older rocks of
the beds we are concerned with,

The Karoo formation occupies a very large tract of dry lands, in the
interior of South Africa, forming the head waters of the Orange river
and its southern affluents, and of the principal rivers which issue on the
south coast. Northwards its extension has not been worked out, but rocks
of a similar character have been observed in central Africa. "Details of the
palzontology of these regions, which alone would render them of import-
ance in the present connection, are however wanting. Over the whole of
the area in South Africa occupied by the Karoo system, the beds lie hori-
zontally in general, except for slight disturbance of the lower. beds along
the southern margin, and form large, flat, desert, plains, known as Karoo,
from which the system derives its name,

Between the Great Fish and St. John’s rivers the Karoo beds extend to
the east coast, but elsewhere they are separated from it by a series of
ranges formed of more or less disturbed palaeozoic beds. Among these the
Bokkeveld beds contain a fauna of distinctly Devonian facies, overlaid by
Chap. VIIL.J KAROO SERIES OF SOUTH AFRICA. 203

plant bearing beds, the Witteberge and Zuurberge quartzites, containing a
flora allied to the carboniferous of Europe.

The subdivisions of the Karoo system have received a variety of names
at different times from different authors, but generally it may be divided
into three divisions as follows :—

3. Upper or Stormberg beds.
2. Middle or Beaufort beds.
1. Lower or Ecca and Kvonap beds.

In dealing with the palzontology of this system we have the great
advantage that the latest, and most complete, description of the plants is by
Dr. Feistmantel,! the same distinguished palzontologist who examined
and described the floras of India and Australia. According to him the
specimens from the lower, or Ecca, group comprised the following species :

D. Glossopteris browniana. K. Noeggerathiopsis hislopi.
K. Gangamopteris cyclopteroides.
(a variety).
The middle or Beaufort beds yielded—

Schigoneura (?) africana. D. Glossopterts stricta,

Phyllotheca (?) bD. + retifera.
D. Glossopteris browniana, K. D. 95 damudiea (a variety).
D. . angustifolia. Rubidgea mackayt, Tate.
D. se communis.

the last being, according to Dr. Feistmantel, very probably the same as
Palzorittaria kurgt, Fstm, from the Damudas.

In these lists the letter K signifies that the species occurs in ite Karhar-
bari group, D that it is found in the Damuda series, and a glance at the
list will show. that, with the exception of one doubtful species, all those
recognisable are identio al with Indian species, and there can be no
darger in correlating the Ecca and Beaufort beds with the Talchir and
Damuda.

Apart from the palzontology of these beds, the lithological character of
the lower or Ecca group would suggest its correlation with the Talchirs, for
it contains a boulder bed, composed of blocks of stone of various sizes, im-
bedded in a fine grained matrix, precisely similar to that of the Talchirs,
and now generally regarded as owing its origin to glacial action. The
Ecca group also resembles the Talchirs in containing beds which closely
resemble volcanic traps or ashes.” There can be no doubt that these
glacial beds were formed during the same cold period which has left such
conspicuous traces throughout India and Australia, and, taken in con-
junction with the fossils, they leave no room for doubt that the Ecca beds and

' Uebersichtliche Darstellung der geologisch- | Ges. Wiss., VII, Band 3, (1889).
palzontologischen Verhdltnisse sud Afrikas| 2 A.}. Green, Quart. Four. Geol. Soc., XLIV,
I Theil ; Die Karooformation und die dieselbe 244, (1888).
untedlasernden schichten. Abhand. K. bohm,
204 CEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [ Chap, VIII.

Talchirs were deposited contemporaneously. The Devonian fauna, and
carboniferous flora, underlying them have been already referred to, and as
far as can be judged from the published description, there is no great
discordance, implying a long lapse of time, between the beds in which they
are found and the base of the Karoo system.

Besides the plants, and much better known than them, are the numer-
ous reptilian remains, which have been found in the Beaufort beds. These,
too, show a very distinct connection with the much more limited lower
Gondwana fauna. The Ménglilabyrinthondont, Brachyops laticeps, Owen
is closely related to Mzcrapholis stowt?, Hux., from the Beaufort beds.
The aberrant genus Dicynodon is represented by no less than thirteen
species in South Africa, and is not known elsewhere except from the
Pdnchet group, and from reputed triassic beds in North America, while
Ptychostagum ortentale resembles P. declive, Owen, so closely that the
specific distinction is difficult.

The age of the Beaufort beds, as compared with the European se-
quence, cannot be said to be definitely established, the opinions expressed
by different palzontologists, and by the same one at different times, having
varied, but the general consensus appears to be that the reptilian fauna
represents a triassic age. They have generally been regarded as the equi-
valents of the Panchet group, on the strength of the resemblances between
the reptiles of the two, but the conjecture is open to question. ‘The
genus Dicynodon ranges into the next succeeding rock group, which also
contains one plant identical with a PAnchet species, and agrees better in
stratigraphical position than the underlying Beaufort group.

The uppermost group, of the Karoo system, again contains reptilian
remains, among which is one species of Dicynodon, and the following
plants, besides some undetermined equisetaceous stems :—

A. Sphenopteris elongata. , Anthrophyosis, sp.

A. Thinnfeldia odontopteroides. Alethopterts, sp.

A. 3 trilobata (2). A. Podogzamites elongatus.
A. Teniopteris carruthersi. 33 sp.

A. 59 daintreet. Baiera schencki.

In this list the letter A denotes that the species has been found in Aus-
tralia, in the so called mesozoic beds of Victoria, Tasmania or Queensland,
whose exact position in the New South Wales sequence has not been
fully established. The two species of Sphenopteris have been found in
South America, and only one species, Thinnfeldia odontopteroides, a
P4nchet form, has been found in India. The probability that these beds
represent the Pdnchet group has just been referred to, and the presence
of this plant makes the suggestion more probabie.

1 R, Lydekker, Records, XXIII, 10, (1890).
Chap. VIIL. J UITENHAGE SERIES. 205

It will be seen that the flora of the Stormberg beds is as distinctly
allied to the Australian one as those of the lower groups are Indian, but
Indian affinities reappear in the next succeeding rock group,

In the neighbourhood of Algoa bay a set of fossiliferous shales and
sandstones, known as the Uitenhage series, of no great thickness and not
found far from the coast, bears much the same relation to the Karoo sys-
tem of the interior, as do the coastal outliers of the Rajmahal series to
the lower Gondwdanas: of the interior of the Peninsula. Like these, they
are of marine origin and have yielded marine fossils, which were at one
time believed to indicate a lower oolitic age,’ but have since been shown to
be neocomian,? at least so’far as the upper beds are concerned.

The plant remains that have been described appear to have been pro-
cured from the lower beds of the series.3 Unfortunately, Dr. Feistmantel
did not live to complete the description of the plants of this series, and we
have only the earlier description of a more limited collection by Prof.
Tate, of which the following is a list* :—

FILices— CyYcaDEACER—
Lecopteris atherstonet. Paleosamia (Otosamites) recta.
3 rubidget. 3 (Podozamites) morristi.
a5 africana, 35 rubidgei.
Ss lobata. 5 (vel Péerophyllum) afri-
Sphenopteris antipodum. cana,

Cyclopteris jenkinsiana,

ContrERB—Arthrotazttes.

Of these, Pecopteris lobata is a Rajmahdl form, P. atherstonez and P.
rubidget are allied to Alethopteris indica, Cyclopterts jenkinsiana is allied
to C. oldhami, and the eRe is very like the Rajmahdl Echzno-
strohus indicus.

The alliances of the Cycads are vague and appear to be rather with
European lower oolite than with Rdjmahdl forms, but Pecopterds lobata
appears to be the commonest plant, and C. jenkins¢ana is also abundant.
On the whole, however, the flora is distinctly related to that of the R4jmahal
group, though the resemblances are not sufficient to establish a contem-
poraneity of origin.

Having detailed the facts, so far as they bear on the subject in hand,
we can now proceed to the discussion of the age of the different rock

groups.

IR. Tate, Quart, Four. Gal, Soe, XXII, | 8 Quart. Four. Geol. Soc, XXII, 147,
169, (1867). (1867). See alsa G, W. Stow, Quart. Four.
2 Holub. u. Neumayr, Denks., k. b. Akad. Geol. Soc., XXVII, 497, (1871).
Wien., XLIV, 267, (1882). 4 Quart. Four. Geol. Sve., XXII, 144, (1867):
206 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [ Chap. VII.

To begin with, we find, in Africa, India, and Australia alike, certain
beds containing abundant and conspicuous traces of glacial action. The
plant remains show that in South Africa, in the Indian Peninsula, and in
Victoria these are of approximately the same age, and marine fossils show
the same with regard to the beds in New South Wales and the Salt range.
The deposits in every case were formed during a period of great cold, which
was succeeded by a much more temperate climate, and it is almost impos-
sible to doubt that this wide spread change of climate must have been due
to some far reaching, if not cosmic, cause. It is consequently justifiable
to use these glacial deposits for the purpose of correlation, and to conclude
that the boulder beds of the three continents were formed contempora-
neously.

In this way we at once find in the marine fossils of New South Wales
and the Salt-range a means fordetermining the homotaxis of the Talchir
group. The former were once regarded as lower carboniferous in age, but
the bulk of the lower carboniferous species they contain range through the
-whole of the epoch. The absence of the group of Productus giganteus,
and the presence of the genus Strophalosia, point to a newer horizon, and
they are now looked upon as upper carboniferous or somewhat newer, The
reasons for ascribing a similar age to ihe Salt-range boulder beds have
already been given.

The Barakar group in India and the Beaufort beds in South Africa agree
so closely in their stratigraphical relations to the glacial boulder beds, and
in their fossil plants, that they are clearly equivalent to each other. In
the case of the Newcastle beds. of Victoria, though the palzontological
agreement with the Barakar group is less close, the stratigraphical
relations are equally intimate, and there can be little doubt that there
is no great divergence in the homotaxis of the two groups. The strati-
graphical connection, between the Newcastle beds ard the underlying,
marine carboniferous deposits, is too close to allow of any great interval
of time, and the Newcastle beds and Barakars cannot well be newer than
permian. In the Salt range a similar stratigraphical position is occupied
by the lower division of the Productus beds which are separated from
the boulder beds of Talchir age by the speckled sandstones. If not the
equivalent of the Barakar group, they cannot be much older. Their age
has already been established as corresponding to the permo-carboniferous
of the Ural mountains, and it is noteworthy that the flora of this age in
eastern Russia is as essentially and typically paleozoic as the contem-
poraneous Barakar or Karharbdri flora is mesozoic.®

1W. Waagen, Fahrb. K. K. Geol, Reichs. IV, 153, (1890).
Wien, XXXVU, 163, (1887) ; Records, XXI, * See Pal, Indica, series xiii, 1V, 175,
106, (1888) , See also Pal, Indica, series xiii, (1890).
Chap. VII] CORRELATION OF THE GONDWANA GROUPS. 207

As we ascend the sequence the evidence gets less satisfactory. The
probable equivalence of the Panchet group with the Hawkesbury beds of
Victoria and the Stormberg beds of South Africa has already been referred
to. The fauna of the Beaufort beds, once regarded as_permian, is
now more usually looked on as triassic, but the evidence is not con-
clusive. In the Salt range no certain equivalent of the Pdnchets can be
detected, though it is possible that the base change of faunaat the top of
the Ceratite beds was due to a change of ocean currents caused by, or
the cause of, that change of climate which is indicated in the Pénchet and
Hawkesbury beds. This suggestion would harmonise with the conclu-
sions that have been drawn from the study of the animal remains of the
Pdnchet group.

The age of the RAjmahdl group must remain. uncertain till the marine
fossils of the outliers on the east coast have been determined. It is older
than the Tripetty sandstones from which 7rigonia smeei and T. ventricosa
were obtained, both of which occur in the Umia group of Cutch, and the
last named also in the neocomian beds of the Uitenhage series. The
cephalopoda of the Sripermatér group were believed by Dr. Waagen to
have a neocomian facies, but the specimens are too ill preserved for very
great value to be attached to this determination.

Indirectly we can form some sort of a guess at the age of the Rajmahdl
group, for the plant beds of Cutch overlie beds containing a marine fauna
which represents an upper oolitic (portlandian) horizon, and underlie a
bed of ferruginous oolite of neocomian or aptian age. The lower oolitic
facies of the flora has been mentioned, but in view of the uncertainty that
attaches to palzobotanical evidence, when large distances intervene, and
the distinctness of that afforded in the present case by the marine fossils, it
is impossible to regard the Cutch plant beds as older than upper oolite
while they may verge into the neocomian period, The Jabalpur group,
which is closely related to the Cutch plant beds, becomes, consequently,
middle oolite at- the oldest, and the RAjmahdl series ranges backwards
from that, throwing the Rdjmah4l group of the RAjmahdl hills into the
lower oolite or even the lias. :

Taking everything into consideration, we may then accept the corre-
lation indicated below as approximately representing the true interpreta-
tion of known facts. Two important reservations must, however, be made
in this connection, In the. first place the suggestion made by Mr. H.
F, Blanford! in 1875, that the Talchir boulder bed was contemporaneous
with the permian glacial deposits of England, has never been absolutely
disproved, and as recent investigations have shown that the supposed lower
carboniferous deposits of Australia are newer than they were formerly

1 Quart. Four, Geol. Soc., XXXI, 528, (1875).
208 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. [Chap, VIII.

i
Tabular Statement of the probable equivalence of the Upper Palzozotc
and Lower Mesogotc rocks of India, Africa and Australia.

 

 

7
Europe. India. New South Wales. South Africa,

 

Upper Oolite . - | Umia.
Middle Oolite . . | Jabalpur.
Lower Oolite . « | Kota-Malérf.

Lias . .  ,|Rajméhal.

2

Rhetic . .|Mahddeva’. . | Wianamatta, :
i
Trias. . » | Pénchet : . | Hawkesbury . «|, Stormberg.
Permian . . | Réniganj .
} Newcastle 7 - |Beausfort.
Permo-Carboniferous | Barakar a. ;
i
Upper Carboniferous | Talchir > . | Lower coal measures | Ecca.
and associated mar-
ine beds.

a

 

 

 

 

 

considered to be, it is still possible that this may be the true equivalence,
in which case the Panchets might be of rhetic age, as their flora indicates.
In the second place the correlation of the R4jmahdl group with the lias is
open to question. Dr, Waagen’s statement regarding the neocomian facies
of the Sripermatér cephalopoda cannot in itself carry any weight, on
account of the very imperfect and ill preserved material he had to deal
with, but, taken in conjunction with the palzontological connection of the
Rajmahal flora with that of the lower beds of the Uitenhage series, whose
upper beds are now regarded as neocomian, it is strongly suggestive of a
later age for the RAjmahdl beds than is indicated in the table. The pale-
ontological grounds on which the Rdjmahél is considered older than the
Jabalpur group have been referred to in the last chapter, where it was
shown that, looked at from a purely local point of view, it is more probable
that the difference between the two floras, and their admixture in the Kota-
Maléri group and the upper members of the outliers on the east coast, is
due to a difference in age than to a mere difference of situation These
reasons are not, however, conclusive, and it seems possible that un ex-
amination of the marine fauna of the east coast outliers would show that the
Rajmahdl beds are newer than has been supposed, and that the differences
between the Rdjmahdl, Jabalpur, and Cutch floras do not indicate successive
periods of time so much as divergent conditions of soil or climate, existing
in different parts of the continent on which the beds of the Gondwdna
system were deposited. Further, there are grounds for taking the Umia
Chap. VIII] © RANGE IN TIME OF GONDWANA SYSTEM. 209

group out of the jurassic system and placing it at the base of the cretace-
ous, since its stratigraphical connection with the overlying beds, containing
upper neocomian ammonites, is as close as with the underlying Katrol
group, whose age is regarded as upper oolite (Oxford). Though it is im-
probable in either case that the Cutch groupsare exactly contemporaneous
with the groups they have been referred to, on the strength of their marine
fauna, there is ample room between the possible limits of error to allow of
the Umia group being either lowermost neocomian or uppermost oolite,
and the occurrence of Uitenhage neocomian 7yzgonzz in the Tripetty beds,
which are probably, and the Katrol beds which are certainly, slightly older
lends some support to a correlation with newer beds than that adopted.

It will be seen from this that there are grounds for shifting the whole
of the Gondwana groups a step higher in the sequence than the positions
they occupy in the tabular statement, but the commencement will still
remain in the palzozoic era, whence the system ranges throughout the
lower half of the mesozoic era well into its upper portion.

Apart from these elements of uncertainty regarding the exact correlations
of the different rock groups, the tabular statement does not truly represent
their relations of the different rock groups in point of time. It has been
necessary to space the rock groups equally, as it is quite impossible to
determine what relation the periods they respectively represent may bear
to each other, though it is quite certain that they are far from being equal
or nearly so, Apart from this, and with the reservations already made,
the table may be taken as representing the nearest approach to the truth
which is at present possible of attainment, and equivalence adopted as that
which appears to agree best with the known facts, taken as a whole,

The comparison of the South African and Indian Gondwana floras is of
less importance and interest from the point of view of establishing their-
homotaxis than as indicating a former distribution of land and sea very
different from what now exists. Naturalists have before now appealed to
a former land area stretching across what is now the Indian Ocean, to
explain certain relationships between the living fauna of the Indian
Peninsula on the one hand, and South Africa and Madagascar on the other,
and the name Lemuria, given to this suppositious continent, is familiar to
many. The hypothesis has of late years been discredited, at least in the
form in which it was first propounded, and for the purpose it was originally
intended to serve, and the most distinguished of the authors who have
treated this subject, Dr. A. R, Wallace, has not only denied the neces-
sity of appealing to any land connection inorder to explain the pecu-
liarities in the distribution of living animals, but has declared' that the

1 “Island Life,” ed. 1880, p. 418.
210 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS.

(Chap. VIIL

fauna and flora of the Mascarene islands, lying between the two con-
tinents, is such as to preclude the possibility of their being the remnants
of an ancient continent, as was supposed by the believers in Lemuria. Dr.
W. T. Blanford has, however, not only shown that the facts are not fully
stated in Dr. Wallace’s book, but has shown that the actual distribution of
certain genera of birds, fishes, reptiles, and land mollusca, are strongly
suggestive of a stretch of dry land having formerly extended from
Southern India to Madagascar! The question is a complicated one, but —
even if Dr. Wallace’s conclusion is granted, it no ways justifies the much
wider inference he has drawn in support of the somewhat popular hypothesis
of the permanence of continental and oceanic areas, and is quite consist-
ent with the existence of an Indo-African continent in’ pretertiary times.

The facts that have been detailed regarding the fauna and flora of the
Karoo system show that there is a closeness of relation, amounting to iden-
tity and, extending throughout the whole of the Talchir ard Damuda periods,
which is inexplicable unless there had been a continuous land communication
along which the plants could freely migrate between two areas,?
And the conclusion is vastly strengthened when we remember that through-
out the greater part, if not the whole, of this period, a very different type
of flora was flourishing in Europe and North America.

Whether the comparative absence of Indian forms in the flora of the
Stormberg beds indicates a break up of this land connection and the
establishment of free communication with what is now Australia, it is difficult
todecide, but if broken up, the presence of the Indian forms in the Uiten-
hage series suggests that the connection must once more have been
established. Itis true that only one species is identical in the two areas, but
after allowing for the uncertainty of the alliances of fossil species of plants,
the connection between the two floras seems to be real, and the differences
are such as would naturally follow from a difference in their age.

The indications of a former Indo-African land area do not cease with
the Gondwana epoch. From a study of the jurassic fauna of the world,
Neumayr came to the conclusion that a land barrier must have stretched
from Africa to India during that period, separating two distinct
faunas.2 This conclusion was especially founded on the study of the
neocomian fauna of the Uitenhage series, and has lately received a
strong confirmation in the identification of four species of Belemnites from

1 Proc. Geol. Soc. Lond., 1890, Presidential,

address p. 88,

2 Dr. A. R. Wallace (‘Island Life,” p. 398)
speaks of the “fragmentary evidence derived
from such remote periods” and the futility of
the notion that “a similarity in the production
of widely-separated continents at any past
epoch is only to be explained by the existence

of a direct land communication.” As may
be seen from what has gone before and what is
to follow, he hardly appears to have appre-
ciated the full weight of the evidence. The
subject has been treated of by Dr. W. T. Blan-
ford iu his presidential address to the Geolo-
gical Society, 1890,

8 Denkschr, k. k, Ak. Wiss.
(1885).

Wien, L, 132,
Chap. VIITIJ AN INDO-AFRICAN. CONTINENT. at

Madagascar. Three of these belong to the group of Mofoceli,a gtoup
which is typical of his equatorial fauna, while the fourth belongs to the
#astat?, a group which is distinctly southern in Europe, The only Bedem-
wete in the Uitenhage beds is not only different from any of the Mada-
gascar forms, but belongs to the group Adso/utz, which is typical of the
boreal regions in the northern hemisphere. The infetence from this is
that the neocomian beds of northern Madagascar were deposited in an
extension of the tropical sea, while those of the extreme south of Africa
were formed in a different, prebably colder ocean.?

This barrier does not seem to have been absolutely continuous
throughout the jurassic period, or there may have been a mode of communi-
cation round the north of the Peninsula of India by which some migration
took place, and so the presence of a few Cutch species, which are also
found on the east coast of India and in South Africa, ts accounted for,

In cretaceous times the evidence is even stronger. The fauna of the
cretaceous (cenomanian) beds of Bagh is closely allied to that of Arabia and
Europe, but is as distinct from that of the cretaceous beds in Trichinopoli
as is possible in the case of two homotaxially equivalent faunas, But the
Trichinopoli cretaceous fauna is very closely allied to that of the Khasi
hills and of South Africa, showing that these areas, which are separated
from each other by distances much greater than that which divides the
Trichinopoli and Bagh exposures, were parts of one marine province, and
the difference of the fauna from that of the lower Narbadd valley can
only be explained by the existence of a land barrier, separating the sea in
‘which the Trichinopoli, Khasi, and South African fauna lived, from that ia
which the Narbad4, Arabian, and European cretaceous beds were deposited.

We see then that throughout the later part of the palzozoic and the
whole of the mesozoic era, there was a continuous stretch of dry land over
what is now the Indian Qcean, which finally broke up and sank beneath
thesea in the tertiary period.

This conclusion has an important bearing on the generally, though not
universally, accepted doctrine of the permanence of continental and
oceanic areas. It is claimed, by many geologists of eminence, that the
deep oceanic areas of the present day have been oceans throughout the
whole of the period represented by the sedimentary formations of the
geological sequence, and that, if we except small volcanic islands rising
from the depths of the ocean, the dry land of every geological period was
confined to the present dry land and the shallower parts of the sea sur-
rounding it.

' Quart. Four. Geol, Soc., XLV, 333, (1389). | but the climatic inferences are vitiated by the
2 Neues Fahrb. Min. Geol., 1890, Band I, p. 1; | possibility of there having been extensive al+
W. T. Blanford, Quart. four. Geol. Soc., XLVI, | terations of latitude since these beds were dee
Proceedings, p. 98, (1890). The evidence is | posited (wide infra).
good for the distinctness of the marine areas, 3 See pp. 247, 252.

P2
212 GEOLOGY OF INDIA—GONDWANA HOMOTAXIS. (Chap. VIIL

The soundings that have been made in the Indian Ocean are not so
numerous as to preclude the possibility of there being a bank connecting
India and Africa, which would allow a bridge of dry land having existed
without imperilling this popular theory, but there are no indications of the
existence of such a bank, and the Indian Ocean is generally regarded by
the supporters of the hypothesis as one of the original oceanic areas of
the world. So the conclusion we have to draw from known and accepted
‘facts is in conflict with an hypothesis which has much to be said in its
favour and has the support of many of the most eminent geologists of the
day.

But this hypothesis is, consciously or unconsciously, to a large extent
bound up, with, and based on, ideas relative to the constitution of the earth
as a whole, which represent it as either solid throughout, or at any rate as
having a solid crust whose thickness is very considerable in proportion to
the whole diameter of the earth. And one of the consequences that follow
on this theory of the constitution of the earth is, that there cannot have
been any great changes in the direction of the earth’s axis of revolution,
or changes of latitude of places on its surface caused by the shifting of tlie
superficial crust over the internal core.

Here again we find the facts in conflict with a generally accepted hypo-
thesis, which is, however, being gradually discredited in later years. WKat-
ever may be the cause of these cold periods, of which two are now well
established in the geological history of the world and several more are
less completely indicated,! there can be no doubt that their effects will be
more widespread, more extensive, and their traces more conspicuous in
high latitudes than in low. Yet the remains of this carboniferous glacial
period so conspicuous in India, Africa, and Australia, all lie within, or only
just beyond, thirty degrees from the equator. The furthest from™the
equator lie in latitudes where the last glacial period, of pleistocene times,
has left but few traces at low altitudes, and those of a somewhat doubtful
character, while most of the remains are in latitudes to which the ice of
the pleistocene glacial period never penetrated, and many are well within
the tropics. ,

At the same time the corresponding deposits not merely of the temperate
clime in Europe and America, but even within the Arctic circle, in which
one would expect the traces of this cold period to be more abundant
more extensive, and more conspicuous, are almost free from traces Gt
glacial action. Boulders of rock imbedded in fine silt have been found
and some have shown a striation believed to be due to the action of ite,
but they are sporadic and indicate that the carboniferous beds of England
and Europe, in which they were found, lay in a latitude which was Gene the
limit which the floating icebergs could reach before melting.

J. Croll; Climate and Time,” London, 1875, Chap. XVIII. See also PP. 106, 346, A. £,
Chap. V1II.] CHANGES OF LATITUDE, 213

If we compare this with the very distinct evidence that glaciers
descended to low altitudes in the Penganga valley and the great desert of
Rajputdna, the contrast is not only striking but inexplicable, unless there
has been a very considerable change of the position of the earth’s surface
relative to the preseat position of the poles.

There is independent evidence that similar changes of latitude are
actually taking place at the present day, in the records of all the principal
observatories of Europe and America! The justice of the conclusion has
been questioned and the variations have been ascribed to instrumental
errors and ercors of observation, but their consensus is so strong that this
appears to be out of the question, and their accuracy has to a great extent
been confirmed by a series of observations recently made at Honolulu?
In fact, the only grounds for questioning the possibility of changes of lati-
tude on the earth’s surface are based on those hypotheses regarding the
constitution of the earth which have been referred to. ‘these hypo-
theses, however, are not based on direct observation, but on mathematical
reasoning of a very brilliant nature which, to quote Prof. Huxley’s well
known simile, is a mill that grinds very fine, but can only grind what
is put into it. In the present case the fundamental data on which the
investigations were based are uncertain, and the conclusions must con-
sequently be questionable. This view has frequently been verbally accepted,
and more recent investigations have gone far to modify the earlier ones,
yet such is the glamour of genius, and such the natural tendency of
human nature to mistake precision for accuracy, and to prefer a definite
statement to a guarded inference, that the idea of the general solidity of
the earth, with its consequences, that very extensive changes in its form, or
the position of its surface relative to its axis of revolution, are impossible,
and the limitation of the period which has elapsed since the earth’s surface
had cooled sufficiently to support life and to admit of the deposition of
ordinary sediments, to a comparative few millions of years, have exerted
an influence on the speculations of physical geology none the less im-
portant because usually unconscious.

It has already been stated that some of the more recent mathematical
investigations have tended to show the uncertainty of the earlier conclu-
sions, and in one of the most recent and philosophical of these.; the Revd.
O. Fisher has almost returned to the old idea of a thin crust lying on a
molten interior. According to him the actual solid crust of the earth is

TAsaph Hall, Am. Four. Sci., 3rd series, , important, as showing a distinct change cf
XXIX, 223, (1885); R.D. Oldham, Geol. Mag., | azimuth, and probably of latitude, since the
3rd decade, III, 300, (1886); G. C. Comstock, | pyramids were built.

Am. Four, Sei., 3rd series, XLII, 470, (1892). 2 Proc, Roy, Soc., L., 227, (1891).
Flinders Petrie’s observations (‘ Pyramids and 80. Fisher, Physics of the Earth’s Crust
Temples of Ghizeh,” London, 1883, p. 125) are | London, rst edn., 1881; 2nd edn., 1889,
2R4 GEQLOGY OF INDIA—GONDWAN\ HMOMOTAXIS. (Chap. VIO.

comparatively very thin, not exceeding 25 to 30 miles in thickness, and
vests on a magma which, in its essential characters, must be regarded as
a fluid. What may be the condition of the great mass of matter forming the
eentral core of the earth is unknowm and immaterial, but if Mr. Fisher's
theory of the earth’s crust is correct, it not only allows of changes of lati-
tude having taken place, but renders it exceedingly improbable that they
have not taken place. In this respect it agrees better with the observed
facts than the more usually held hypothesis, and we will see, when the
Himalayan range is dealt with, that this is not the only respect in which a
confirmation of his theory can be derived from the facts. of Indian geology.
CHAPTER IX.

MARINE JURASSIC ROCKS.

Cutch—Western Rajputdna--Salt ranee— Himdlayas— Afgh4nist4n— Doubtful jurassics of
western Garhwal.

With the exception of the fossiliferous upper Gondwana beds of the east
coast, referred to in a previous chapter, no marine beds of jurassic age
are known to occur in the peninsular area proper, and even as regards these
it is still uncertain whether they should not be regarded as more recent
in their origin. | But in the debateable tract lying east of the Indus and
west of the Ardvallis, marine jurassic rocks attain a large development
in Cutch and in western Rdjputdna.

The jurassic area of Cutch! may be considered as occupying a number
of post-tertiary islands, now connected by alluvial flats. The largest of
these islands, that forming the western and central portion of Cutch, is
about a hundred and twenty miles long, from east to west, by about forty
broad. To the north-east of it is the district of W4gad, another ancient
island nearly fifty miles from east to west, and, excluding alluvium and
“ Rann,” twenty-five miles broad. Farther north four isolated masses of
hills, chiefly composed of lower jurassic rocks, extend in a line nearly a
hundred miles in length from east to west. These are the so called islands
in the Rann,? Pateham, Kharir (Khurreer or Kurreer), Bela and Chorar.

1 The account of this province is taken partly
from a report by Mr. Wynne, Memoirs, XI, pp.
1—203,(1872), and partly from manuscript notes
by the late Dr. Stoliczka. The Cephalopoda have
been determined by Dr. Waagen, and described
in the Pal. Indicas series ix, (1873-75). It
shou.d, perhaps, be noticed that Dr. Waagen’s
views of specific distinction differ from those of
many paleontologists, and that,as he points
out, several of the forms described by him as

species might by other naturalists not be con-
sidered to rank higher than varieties.

2 The “Rann” of Cutch is an immense tract
surrounding the province on all sides, except
the south, and consisting of barren salt marsh
periodically overflowed by sea water. This
tract, which is evidently an ancient sea basin
now filled up by alluvial deposits, will be furs
ther described in a subsequent chapter on
postetertiary and recent deposits,
216 GEOLOGY OF INDIA—MARINE JURASSIC ROCKS. [OQhap. IX,

A few smaller islands also occur, but none of them are of sufficient size
to be worthy of notice.

None of the rocks found in Cutch and the adjoining islands are of older
date than jurassic. In one spot some limestones, containing upper neo-
comian Cephalopoda, are found resting upon the jurassic series, the
uppermost group of which may perhaps itself be of intermediate age, and
belong in part to a lower neocomian horizon." In general, the upper
jurassic beds disappear to the south beneath the Deccan traps, but marine
tertiary beds (nummulitic) overlap the traps and rest upon the older series
in many parts of the country, both traps and nummulitic beds being quite
unconformable to the jurassic formations.

The lowest beds are seen dipping to the south in the Rann islands,
and are locally exposed in an anticlinal which runs along the noithern
edge of the province, the intervening synclinal being, for the most part,
concealed beneath the Rann, From the anticlinal near the Rann there
is a general dip, varying in amount, to the southward. The greater
portion of the series is, however, repeated twice in consequence of a great
fault, which runs from east to west along the northern scarp of the
Charwér range of hills south of Bhdj.

By the earlier observers, including Mr. Wynne, the jurassic. series in
Cutch was simply divided into a lower and an upper group, the former
chiefly marine, the latter apparently fresh water for the most part, though,
as was shown clearly by Mr. Wynne, no marked line of division can be
drawn, for not only is there an absence either of unconformity or of any
marked break in lithological character between the two subdivisions, but
marine beds are occasionally found interstratified with the upper, and
plant beds with the lower group. The examination of the Cephalopoda
by Dr. Waagen indicated the probability that representatives of several
European jurassic groups existed in Cutch, and Dr. Stoliczka, re-examining
the beds with the aid of Mr. Wynne’s geological map ard his own
knowledge of palzontology, found no difficulty in distinguishing four sub-
divisions, the three lower of which had been included in the inferior or
marine group of previous observers, whilst the upper comprised the higher
fresh water beds, with the uppermost marine strata. The names of the
groups proposed by Dr. Stoliczka, with their homotaxial equivalents
amongst European formations, are exhibited on the opposite page.

The whole thickness of the Cutch jurassic series has been estimated by
Mr. Wynne at 6,300 feet, of which 3,000, or very nearly half, belong to

‘ Dr. Stoliczka unfortunately did not live to { (infra, p. 286) is conformable to the underlying
publish the results of his examination of Cutch, | Umia beds. He doesnot precisely state, how-
but from his rough field notes it appears prob- | ever, what are the relations of the upper bed to
able that the upper neocomiaa bed of Ukra | the lower at this spot,
Chap. IX.]

CUTCH.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

217
Classification of the Furassic sertes in Cutch.'
CUTCH. | EUROPE.
Groups. Beps ZONES. Group.

1. Beds with Crioceras and Ammo- seiner CRETACEOUS.
nites of the rhotomagensis group. |Opper Neocomian,

2, Sandstones and shales with Cy- ? ? WEALDEN,
cade@ and other plants.

Umias,

3. Sandstones and conglomerates | Upper Tithonian . | Upper jurassic,
with marine fossils, Ammonites | Lower Tithonian .| Tithonian and
(Perisphinctes) eudichotomus, fre- Portland,
quens, Irigonia smeei, T. ven-
tricosa etc,

4. Sandstones and shales with Am.| Zone of A. (Perisph) UPPER
(Phytloceras) plychoicus, A. (Op-| mutabilis. JURASSIC.
pelia) trachynotus, A, (Perisphinc-| Zone of A. (Oppelia) Kimmeridge.
tes) torquatus, pottingeri, etc. tenuilobatus.

Katrot. (| 5. Red ferruginous and yellow sand- |? Zone of A. (Pelt.)
stones (Kantkot sandstones) bimammatus.
with Am. (Stephanoceras) maya,|? Zone of A. (Pelt.)
A, (Aspidoceras) perarmatus, A. transversarius, MIDDLE
(Perisphinctes) virguloides, leio- JURASSIC.
cy mon. ; ‘(Middle oolite).
Oxford.

6. Oolites (Dhosa oolite) with Am. | Zone of A. (Amaltheus)

(Séephanoceras) polyphemus, A.| cordatus.
(Perisphinctes) indo-germanus, | Zone of A. (Amaltheus)
A. (Aspid) perarmatus, babeanus,| lamberti.

A. (Pelt.) arduennensis, etc

7. White limestones with Am. (Pelt.)| Zone of A. (Pelt.)
athleta, A. (Oppelia) bicostatus,| athleta.
etc.

CuArt . | |8 Shales with ferruginous nodules, | Zone of A. (Perisph.)
with Am, (Perisph.) obtusicosta,| anceps. MIDDLE
anceps A. (Harpoceras) lunula, JURASSIC.
punctatus, etc. (Middle oolite).
Kelloway.

9. Shales with calcareous bands and| Zone of A. (Steph.)
locally with golden oolite: Am,| macrocephalus.
(Steph.) maczrocephalus, tun idus,
bullatus, A. (Oppelia) subcos-
tarius, A. (Perisph.) funatus,
etc.

to. Light grey limestones and marls
with Am. (Oppelia)  seiriger, Mippte

Z Corals and Brachiopoda, etc. JURASSIC
PATCHAM | dis (Lower ovlite).

11, Yellow sandstones and limetones, Bath.
with 7sigonia, Corbule, Cucullee,
etc.

 

 

 

 

 

 

Waagen, Pal. Indica, series 1X, Introduction, (1875).
218 GEOLOGY OF INDIA—MARINE JURASSIC ROCKS. [Chap. IX,

the uppermost group alone, the thickness of the other groups not having
been estimated separately. It must be remembered that the base of the
whole series is not exposed, and that the upper beds had suffered from
denudation before they were covered by the traps.

The Patcham group is thus named from its occurrence in the island
of Patcham in the Rann. The lowest beds are exposed on the northern
scarp of a range of hills which runs east and west through all the Rann
islands from Patcham to Chorar. The rocks composing the range dip
south at a low angle, the crest of the hills and the surface of their southern
slopes being formed of a thick massive bed of yellowish sandstone and
limestone, which contains Corbula pectinata, Astarte compressa,a Trigo-
nia closely resembling 7. interlevigata, Cucullea virgata and other
fossils.1 Below the massive bed come shales and sandstones, all more or
less calcareous, containing a Rhynconella, near R. concinna, Lima, Ger-
villia, a small Exogyra, etc. The lowest bed seen in Patcham island is
calcareous sandstone abounding in the small Exogyra. The same lower
beds are seen in Koari Bet, asmall islet north-west of Patcham, and on
the northern flank of the range, in Kharir, Bela, and Chorar, the top of the
range in all cases consisting of the yellow calcareous rock. The thickness
of this portion of the beds is at least 500 feet.

Besides forming the range of hills in the islands of the Rann, the
Patcham limestone is exposed at four places in Cutch itself,—at Jarra, Kira
hill near Chari, Jura hill, and in Halamdn hill near Lodai—all situated along
the northern edge of the main province of Cutch, near the borders of the
Rann, In all these places they appear as inliers, exposed at the crest
of an anticlinal, and surrounded on all sides by higher beds. At Jarra,
about fifty miles north-west of Bhdj, there is a bed of white limestone
containing Scyphia, a Terebratula, and small Rhyanconel/z and, immediately
above it, a bed of corals. These rocks do not appear to be equally well
exposed elsewhere. They are at the base of the Chari group and were con-
sidered by Dr. Stoliczka as the uppermost beds of the Patcham group of
Cutch.

The lower portion of the Patcham group has yielded no Cephalopoda,
and the higher beds only eight species, all of which are rare. One is
Nautilus jumarensts, the others are Ammouttes, of which one belongs
to the sub-genus Opfpelia,three to Stephanoceras, and three to Perisphinctes.
One Stephanoceras is a variety of Ammonites macrocephalus, the typical
form of which is abundant in the next higher subdivision, and both
the other species of Stephanoceras pass likewise into the lower beds

1 As only the Cephalopoda of the Cutch beds { may require modification. Only those are
have been properly compared, it is possible | mentioned which are in all probability correctly
that some of the identifications of other fossils ' determined.
Chap. IX.J CHARI GROUP. 219

of the Chéri group. With the exception of A. macrocephalus, the only
species found also in European rocks is Ad. (Opfelia) serriger, which
was originally described from upper bathonian beds. So far as the
Cephalopoda are concerned, it would be difficult to correlate the Patch-
am group with any subdivision of the European oolites, but the Patch-
am Brachicpoda, which, however, have not been thoroughly compared,
and the position of the beds immediately beneath the strata containing
A. macrocephalus in abundance, have induced Drs. Stoliczka and Waagen
to reter the group to the horizon of the Bath oolite (bathonian),

The next group in ascending order derives its name from the village of
Chéri, situated close to the borders of the Rann, about thirty-two miles
north-west of Bhij. This village has been known since the time of Captain
Grant, the earliest geological explorer of Cutch, as an admirable locality
for fossils, and especially for Cephalopoda, of which large numbers are
found in the calcareous sandstones exposed around Kira hill.

The Chéri group is composed of four subdivisions, each marked by its
mineral characters and by the fossils it contains. The group, as a whole,
is much more shaly than any of the other subdivisions, but it contains
hard bands of limestone or calcareous sandstone forming ridges, which
are usually distinguished by characteristic forms of Ammonites.

The lowest of the four zones or subgroups consists of shales, usually
of a grey colour, with occasional bands of golden oolite, and sometimes
nodular shaly limestone. The rock called golden oolite (which is not
peculiar to India, but which is also found in the jurassics of Europe, and at
about the same horizon) is very characteristic and easily recognised. It is
a rather coarse grained limestone, composed of calcareous grains, which
are coated with a very thin ferruginous layer and are surrounded by a
matrix of carbonate of lime, so that the stone has much the appearance at
first sight of a rock with golden coloured mica. In places, as at Chéri
itself, the golden colite is thick and conspicuous, but it is locally distri-
buted and often wanting. The most characteristic fossils of these lowest
Chati beds are Ammonites (Stephanoceras) macrocephalus, and allied
forms of the sub-genus Stephanoceras.

Above the macrocephalus eds come dark shales, often black, with
ferruginous bands and concretions. Sometimes, however, the nodules
are of white limestone, and the shales are locally sandy, and associated
with sandstones, but the beds appear to preserve their lithological
characters in general thoroughout Cutch. The chief palzontological pecu-
liarity of this subdivision is the extreme abundance of a 7erebratula,
considered by Sowerby a variety of the cretaceous 7. d¢plicata. Planu-
late Ammonttes (Perisphinctes) are also very common. The shales not
220 GEOLOGY OF INDIAmMARINE JURASSIC ROCKS. (Chap. IX,

unfrequently contain remains of plants, but no distinct impressions have
been found.!

The next subdivision, in ascending order, is a very thin band, some-
times only 20 to 30 feet thick, of whitish or grey shale, with bands
of limestone, which are generally white, but occasionally yellowish or
brown. Usually this band may be recognised easily by its colour and by
its presence beneath the Dhosa oolite. The most characteristic fossil
is Ammonttes (Peltoceras) athleta and, in north-western Cutch, the shell
of this mollusc is usually changed into black calcspar.

The uppermost Chéri subdivision, or Dhosa oolite, is the most character-
istic of all both lithologically and palzontologically. It is of no great thick-
ness, though more developed than the athleta beds, and consists of grey,
reddish, or brown oolite, sometimes sandy and oftennodular. Cephalopoda
are extremely abundant. And it abounds in many places in a Terebratula
closely allied to the cretaceous 7. se//a, and referred to that species as
a variety by Sowerby.

The Chari beds are exposed in several places in Cutch, but they no-
where occupy a large area. They are found resting upon Patcham beds
in the southern part of Patcham and Kharir, and in Kakindiya and
Gdngta, two small islands south-east of Kharir, forming only the axis of a
quaquaversal anticlinal on the latter, but none are exposed in Bela or
Chorar, though a small area exists in the extreme north of Wdgad. In
these outcrops the subdivisions are less well marked than to the south-
ward, and the two characteristic Terebratulz have not been noticed. In
the mainland of Cutch, the Chdri group occupies two series of inliers. One
of these series is scattered at intervals along the northern anticlinal tange,
The rocks appear at three places west.north-west of Chari, again around
Kira hill, near Chari, the typical Jocality, they extend nearly twelve miles
from east to west around the Patcham beds of Juria hill, north of Bhdj, and
are found in two more outcrops farther east around Halaman hill, where
they extend more than six miles, and they again appear a mile farther
east. Another series of outcrops occurs-in the Charwér range, south
of Bhij. Here the Chari beds are brought up at intervals ican the
southern side of the great fault; they are greatly disturbed and cut up by
cross faults, but the different bands can be easily recognised,—the Dhosa
oolite with Terebratula sella, var., the white athleta beds, and the band
with 7. dipd¢cata, var., being always conspicuous.

The cephalopodous fauna of the Chari group comprises a hundred and
twelve species, of which thirty-seven areEuropean. The relations between
those found in the different subdivisions and the corresponding Kelloway

1In Dr. Stoliczka’s field notes he mentions | the lower groups, probably from the Patcham
having at one locality found fragments of | teds, cemented together in the rock at this
quartz and of a limestone derived from one of | horizon. This may indicate an unconformity,
Chap. IX.) KATROL GROUP. aa

and lower Oxford groups in Europe, are the following, according to Dr,
Waagen :—

In the lowest Chari subdivision, or macrocephalus beds thirty-one
Cephalopoda have been found, 7v2z. two species of Belemnites, three of
Nautilus and twenty-six of Ammonites (Phylloceras 2, including A,
‘disputabilis, Lytoceras 1, Oppelia 1, Harpoceras 1, Stephanoceras 13, in-
cluding A. macrocephaius, and Perisphinctes 8). Three are common to this
subdivision and the upper Patcham beds, whilst none are known to range
into higher strata, Sixteen species, or rather more than one-half, are
found in Europe, all, except two, belonging exclusively to the beds with
A. macrocephalus (Lower Kelloway).

In the next subdivision, the dark shales with Zerebratula biplicata,
twenty-seven Cephalopoda are found, viz. three Belemnttes, one Nautilus,
one Ancyloceras, and the remainder Ammonttes (Phylloceras 2, Oppelia 3,
Har poceras 5, Stephanoceras 1, and Pertsphinctes 11). Six of these range
into higher beds, whilst seven are European, and of these latter five are
only found in the beds with 4. anceps (Middle Kelloway).

The Ath/eta beds have yielded twenty species, three Belemnites and
seventeen Ammonites (Phylloceras 1, Amaltheus 2, Oppelia 2, Harpoceras
2, Pelioceras 1, Aspidoceras 2,and Perisphinctes 7) ; five of these are com-
mon to the next lower subdivision, and two to the Dhosa Oolite. Eight
are European, six being peculiar to the zone of A. athleta (Upper
Kelloway).

In-the Dhosa ovlite there are thirty-four Cepha/opoda, vis. four Belem-
nites, one Nautilus, and twenty-nine Ammonites (Phylloceras 2, Harpo-
ceras 3, Peltoceras 5, Aspidoceras 4, Stephanoceras 8, and Perisphinctes 9).
Four of these range into higher and three into lower beds. Eight are found
in Europe, the most important being Am. (Aspidoceras) perarmatus, and
seven of these belong exclusively to the zones of 4. (Amaltheus) lambertd
and A. (Amalth,.) cordatus (Lower Oxford). Other fossils, especially
Lerebratula sella, are abundant in this group.

The Katrol group, which rests upon the uppermost subdivision of the
Chari beds, is of considerable thickness. It consists of sandstones of
various kinds, white, brown, pinkish grey, etc., and shales usually grey or
reddish, but sometimes very dark coloured, like those of the Am. anceps
zone. Ferruginous nodules and concretions sometimes occur in the shales
which prevail towards the base of the group, the upper portion being
chiefly sandstones. On the whole, however, shales predominate.

These beds form two belts ir Cutch proper. The first occurs in the
anticlinal along the Rann and extends for nearly eighty miles, surrounding the
222 GEOLOGY OF INDIA—MARINE JURASSIC ROCKS. (Chap. IX,

inliers of the Patcham and Chdéri groups, and extending to a considerable
distance beyond them. ‘The exposure of Katrol rocks varies in breadth
being, where broadest, nearly ten miles wide. The second belt is in the
Charwdr range, south of the great fault, This tract is about thirty-five miles
from east to west, but nowhere more than two miles broad. Besides this the
greater part of Wdgad is occupied by beds apparently belonging to the
same group. ‘The rocks are very similar in mineral character, consisting
of a coarse and fine grey, pinkish and white sandstones above, and grey
or yellowish shales below, but the Cephalopoda found are almost all dis-
tinct, and appear to indicate a lower horizon. From their development
around the town of Kantkot, these Wdgad beds have received the name
of Kantkot sandstone.

The Cephalopoda of this Kantkot sandstone are nineteen in number,
four Belemnites and fifteen Ammonites (Phylloceras 1. Aspidoceras 2,
Stephanoceras 5, Pertsphinctes 7). Four of these Am. (Aspidoceras)
perarmatus,A. (Stephanoceras) maya, fissus, and opts, are also found inthe
Dhosa oolite of the Chari beds, whilst only one species, Belemnites
grantianus (B. kantkotenszs), is common to the Kantkot bed and the
Katrol group in Cutch proper. Thus the Kantkot beds appear by their
cephalopodous fauna allied more closely to the uppermost Chéari beds than
to the Katrol group. Three species only of the Kantkot Cephalopoda
are European, A. (Asp.) perarmatus, A. (Per.) plicatilis, and A (Per.)
martelli, and only one of these, the last, is limited to a single zone, that of
A. (Pelt) transversarius (Upper Oxford) in Europe, the other two
ranging lower. Several forms are, however, allied to upper oxfordian
species,

The Katrol group proper has yielded twenty-six species of Cephalopoda,
four Belemnites and twenty-two Ammonites (Phylloceras 2, Lytoceras 1,
Haploceras 2, Oppelia 4, besides an Aptychus, Harpoceras 1, Aspidoceras
5, Perisphinctes 7). Only one ofthese species, Bel. prantianus, is found
with certainty in any other group in Cutch. Four species are found in
Europe, all belonging to the beds of the Kimmeridge group, with 4. (Asp)
acanthicus. By far the most characteristic and abundant of the Cepha-
lopoda, is a non-canaliculate Belemnite, B. katrolensis. ‘The commonest
Ammonites are A. (Oppelia) kachhensis, A. (Per.) poltingert, A. (Fer.)
katrolensts, aud A. (Per.) torquatus. :

Imperfect plant remains are common in the Katrol group, as they are
in many of the lower beds of Cutch, but in one instance near the village of
Narha, as has already been mentioned in the description of the Gondwana
system several remains of plants, whose relations have already been dis-
cussed on a previous page! were found by Mr. Wynne, in shales inter-
stratified with the Katrol beds and distinctly inferior in position to some
of the marine bands of the group.

1 Supra, p. 189.
Chap. IX.) UMIA GROUP. 223

The Umia group derives its name from a small village in western Cutch,
rather more than fifty miles north-west of Bhij. Taken as a whole, this
group appears to equal in development all the other jurassic beds together,
being, according to Mr. Wynne’s estimate, upwards of 3,000 feet thick. It
is the equivalent of the upper jurassic group of Mr. Wynne’s Memoir. As
a rule, it consists of sandstones of various kinds, and more or less sandy
shales. The sandstones are very often soft and white or pale brown,
sometimes variegated, and very generally distinguished by thin bands of
hard black or brown ferruginous grit. Occasionally the sandstones are
variegated with pink, red, and brown, they are often very argillaceous and
tend to decompose into a loose sandy soil, which covers and conceals the
rocks over a great part of the country. Ina few instances carbonaceous
shale occurs, and in one locality a thin seam of bright jetty coal. A few
thin hard bands of sandstones are met with, some being so hard.as to be
almost a quartzite. There is a marked resemblance in the beds of this
-group to some of the upper Gondwana strata of Central India: there are
the same soft argillaceous sandstones and sandy shales and the same -hard
ferruginous gritty bands.

Towards the base of the Umia group there is a thick band of calca-
reous conglomerate, hard and grey, sometimes ferruginous, associated with
sandstones and shales. In this conglomerate and in some associated beds
marine fossils are numerous, Throughout all the rest of the group plant
remains are common, but they are not often sufficiently well preserved to
be identified. Marine fossils are very rare, but Zrigonia smeet, the most
typical fossil of the group, has been found in places, as near ‘Vigori, forty
miles north-west of Bhuj, in beds near the top of the group and well above
the horizon at which most of the plant fossils have been obtained.

The beds of the Umia group are covered unconformably by the Deccan
traps and by tertiary rocks, except in one place, where they underlie
the upper neocomian (aptien) beds of Ukra hill in north-western Cutch.?

The surface occupied by the rocks of the Umia group corresponds in
magnitude with the thickness of the formation, and embraces nearly, if not
quite, half of the jurassic area in Cutch. In Cutch proper these beds ex-
tend throughout the province from the western extremity near Lakhpat to
the eastern end beyond Bachao, forming a great plain south of the irregi-
lar range of hills along the edge of the Rann. They also extend round each
end of the range, especially to the eastward, where the bottom Umia beds
extend north of the hills about twenty miles along the edge of the Rann.
The main belt of Umia beds is from eight to twelve miles across on an
average. ‘These rocks lap round the western end of the Chdrwdr range,
where the great east and west fault to which the range is due appears to
die out, and they cover another plain, nearly fifty miles in length from east

1 Infra, p. 286.
224 GEOLOGY OF INDIA—-MARINE JURASSIC ROCKS.

(Chap. IX,

to west.and about eight miles broad, south of the Charw4r range,
also form the western portion of Wagad.

The plant remains of the Umia group and their relations have already
been described in the chapter relating to the Gondwana system. It was
there shown that twenty-seven species had been identified, of which ten
are either common to the lower oolitic beds of Yorkshire or represented
by very closely allied forms. Bearing in mind that the plant beds overlie
the portion of the group which has furnished Cephalopoda, it is remarkable
to find that the latter exhibit a very decided upper oolitic facies. They
are eleven in number,! viz.—

They

A. (Perisphinctes) bleicheri.
A. (Per.) occultefurcatus.
A, (Per.) eudichotomuse

A, (Per.) frequens.

A. (Per.) denseplicatus.

Belemnites grantianus (kantkotensis).
Belemnites, 2 sp. indet.

Am. (Haploceras), cf. tomephorus.

A. (Aspidoceras) wynned.

A. (Perisphinctes), cf. suprajurensis,

Of these eleven species, one (Belemnites grantianus var. kantkotensts)
is found in lower beds in Cutch, and the two other forms of Belemnites
are closely allied to the Katrol species B. claviger and B. katz olensis,
and may be identical. All the eight A mmontzées are restricted in Cutch to
the Umia group, and two of them (A. ¢omephorusand A. eudichotomus) are
tithonian species, found in the uppermost jurassic beds of southern Europe,
whilst A. dleicher? and A, suprajurensts are found in the Portland strata
of northern France, and A. occultefurcatus is barely distinguishable from
another Portland species, A. (Perisphinctes) boidint, The connection
between the Cephalopoda of the Umia group and the forms found in the
uppermost jurassic beds of Europe is consequently very marked, and
Dr. Waagen states that the same marked similarity exists between the
lamellibranchiate bivalves of the same beds in the two regions.’

The Cephalopoda are, however, rare and exceptional in the Umia group,
and they form by no means so important a portion of the fauna as in the
other groups. The commonest Umia fossils are two species of Tvigonia
(7. smee? and 7. ventricosa), the latter being also found in the neocomian
rocks of South Africa, whilst a very closely allied form (7. ¢tuberculifera)
occurs in cretaceous beds in Southern India. The occurrence of these
Trigonté in upper Gondwana strata near Raéjamahendri has already

' Nine, according to Dr.Waagen (Pal. Indica,
series ix, 225, 232), but he appears to have
overlooked two forms—Belemnites kantkotensis
(grantianus), stated at page 4 to have been
found in Umia beds, and the specimens from
the same group doubtfully referred to B. cla-
viger on p. 7. These very trifling and unim-
portant oversights arenot noticed in order to

call attention to a trivial error, but because’
the relations of the Umia group are of consid-
erable importance and have been disputed,
In consequence of the great importance of this
group, the evidence upon which its relations
to the upper jurassic beds of Europe are based.
is g'ven in full.
* Pal. Indica, series ix, 225, (1875).
   

Trigonia smeci, Sow. X 4. Pholadomya angulata, Sow, X %.

  

STEER TRIER LUMI rt eR Aa

 

Belemnites grantianus, d'Orb X %

 

 

A, (Perisphinctes) pottingeri, Sow. A, (Aspidoceras) perarmatus, Sow

JURASSIC FOSSILS.

Caleutta Phototype Co,
Chap. TX] CORRESPONDENCE WITH EUROPEAN HORIZONS, (225

most characteristic beds being whitish or greyish sandstone, very fine and
been noticed. Other forms cf common
Occurrence in the Umia group are
species of Trigonia allied to T. vau,
clavellata and pibbosa, Astarte major,
a Gervillia, a peculiar Gryphza, in-
termediate in form between G. di/a-
tata and G, vesicularts, Goniomya, etc.
Some of these range into lower groups
also. A portion of the jaw of a Ple-
Stosaurus, also obtained from these
beds, is said to be undistinguishable
from that of P. arcuatus, Owen, from
the lower lias of England.?

 

Fig. 14.—Trigonia ventricosa, Kraurs, na-
tural size. -

Owing to the circumstance that, with the exception of the Cephalo-
poda, the large collections of fossils made in Cutch by Messrs Wynne,
Fedden, and Stoliczka have not hitherto been examined and compared, the
distribution of many of the most characteristic species has not been defi-
nitely ascertained. Amongst the forms which are most abundantly pre-
served in the lower groups of the Cutch jurassic series are species of Pleu-
rotomaria, Pholadsmya granosa, Ph. angulata, Ph. tnornata, Corbula
lyrata, C. pectinata, Nucula cunetformis, Cucullea virgata, Trigonia
costata and Ostrea marshit,

On the next page is given a table which shows the general result of
Dr. Waagen’s examination of the jurassic Cephalopoda found in Cutch.
The correspondence, not only with the European jurassic rocks as a
whole, but with the different groups into which they are divided, is re-
markable, and greater than is known in any other Indian formations, the
only other series of Indian rocks of which the fauna has been sufficiently
' examined . to justify the comparison, the cretaceous series of Southern
India, showing much less close agreement in the distribution of the fauna,
and especially of the Cephalopoda,with the corresponding groups in Europe.
The only remarkable instance in which the Cephalopoda of the Cutch juras-
sics difter from their representatives in the jurassic rocks of Europe, is in
the prevalence in the Indian area of the macrocephali at a higher horizon
than in Europe. In Cutch they abound in the Dhosa oolite and Kantkot
sandstone, the other Cephalopoda of which are of Oxford, and in the
latter case of upper Oxford types, whilst in Europe they are not known
above the base of the Kelloway group. As will, however, be shown in
the next chapter, some of the cretaceous forms of Ammonites found in,
Southern India show a remarkable resemblance to the jurassic forms of

1 Lydekker, Records, XXII, 50, (1889).

Q
226° GEOLOGY OF INDIA—MARINE JURASSIC ROCKS. [Ohap. IX.

macrocephali, and in this instance they are associated with species allied
to Ammonites characteristic of even older European deposits.

 

 

wo )eoy| 2
So | |P3l ee] ee
= —o| me) 2
BS Bh fy | S| ae
e8(ee)] ss Es|es
Name of Group. | NAME OF SUBDIVISION, Bs | *3 Ba 3 3 REMAKhS.
s °° ps e
es ne Ors ao o™m .
as /22#/2o/25/88
@8) 82/82/82) a8
3 oP) a8 | a8 | Bow
& wn n on oO

 

o|.. | 2] 4 | Two of the European species
occur in Portland beds of
northern France and two in
Tithonian beds of southern
Europe.

Uma . . | Marine beds ej 11

 

Katrol beds proper .; 27 | 26] 1 1 | 4 |All the four European species
belong to the zone of Am.

KaTrROL : acanthicus (Kimmeridge).

 

Kantkot beds -; 19 | 14 1 4| 3 ane
Dhosa Oolite .'34| 27] 4}| 3] 8: Seven characteristic of the
(Terebratula sella zone of A. transversarius
beds). " (Lower Oxford) of Europe.
Cnart . <j Athleta beds .| 20113} 2] § | 8 | Six characteristic of the zone
of A. athleta in Europe.
Anceps beds, with | 27 | 21 6 | ass 7 | Five of the seven species found
Ter. biplicata also in Europe are peculiar

to the beds with A. anceps.

|

|

| Macrocephalus beds. ; 31 | 2" | .. | 3 | 16 | Fourteen of the sixteen exclu-
sively found in beds with A.
macrocephalus in Europe.

Upper . : .| 8 vee [LOF2| satis
PaTCHAM J Pee
¢ Lower . a esd satel age We ccieee lh Rees

 

 

 

 

 

 

The occurrence of jurassic rocks in the desert tract to the north of the
Rann of Cutch has been known for many years. A few species were obtained
from the country immediately north of the Rann by Sir H. Pottinger, but
the only tracts which have yet been explored by a geologist lie further
north, near Balmer, in Jaisalmer and in Bikaner. Five distinct groups of
rocks have been recognised and named as follows :—

5. Abur beds.

4- Parihar sandstones.
3. Bedésar group.

2. Jaisalmer limestones.
-1, Balmer sandstones.

The Balmer rocks consist of sandstones, grits, and conglomerates, the
Chap. IX, ] WESTERN RAJPUTANA. 227

compact, and a still finer rock approaching a compact shale, white, but
veined and blotched with purple. ‘These beds must attain a considerable
thickness, but only the lowest are well exposed, the upper strata being
probably softer. The lower members of the group are well seen at Balmer
itself, where they rest upon the Malani volcanic rocks, and in some hills
near Naosir, about thirty miles farther east. Fragmentary plant remains
are common, but nothing sufficiently well preserved for determination has
been found, and no remains of animals have been detected in the beds.

East and south-east of Jaisalmer, beneath the marine jurassic beds of
the next group, a considerable thickness of white, grey, and trown sand-
stones is exposed, interstratified with numerous bands of hard black and
brown ferruginous sandstone and grit. Towards the base are some soft
argillaceous sandstones, streaked and blotched with purple, and closely
resembling the Balmer beds, except that they are less hard. These
rocks probably belong to the Balmer group. They havea lithological
resemblance to the Umia group of Cutch and to some of the Gondwana
beds of the Central Provinces. The only fossils found, except fragments
of leaves, were some pieces of dicotyledonous fossil wood

The sandstones and limestones of the Jaisalmer group rest upon the beds
last described, and consist of thick bands of compact buff and light brown
limestone interstratified with grey, brown, and blackish sandstone, with some
conglomerate. The limestone forms conspicuous scarps close to the town
of Jaisalmer, and it is highly fossiliferous, containing amongst other species
Terebratula biplicata,T. intermedia, Pholad omya granosa, Corbula lyrata,
C. pectinata, Trigonia costata, Nucula cunerformis, Pecten lens, and
Nautilus kumagunensis. Ammonites (Stephanoceras) jissus has been
obtained from the neighbourhood, but very possibly from a different horizon,
for in Cutch it belongs to the Dhosa oolite and the Kantkot sandstone
(both Oxford), whilst Nautilus kumagunensts is only found at a lower
horizon in the beds with Am. macrocephalus at the base of the Chéri group.
Terebratula biplizata in Cutch is chiefly characteristic of a rather higher
horizon than that of the macrocephalus beds. There can, howe ver, be but
little hesitation in referring the Jaisalmer limestones to the age of the Chari
group.

The Jaisalmer limestones are overlaid by a group of purplish and red-
dish sandstones, with thin layers of black vitreous ferruginous sandstone.
Some beds of a red calcareous sandstone contain fossils which have not
been determined. Some of them closely resemble forms from the Katro]
group of Cutch. These sandstones have been distinguished as the Bedésar
group! and are overlaid by the Parihar group of soft, white felspathic
sandstones, which weather into a clean sugary sand, largely composed of

1R.D. Oldham, Records, XIX, 158, (1886).
228 GEOLOGY OF INDIA—MARINE JURASSIC ROCKS. [Chap. IX.

subangular fragments of transparent quartz. The uppermost beds of the
jurassics are sandstones, shales and limestones, among which there is one
very conspicuous fossiliferous band, known lécally as Abur, a name which
has been applied to the village where it is quarried, and formerly referred to
in the Survey publications as the ammonite bed of Kuchri.? It ts a thin bed
of buff coloured limestone, weathering red, and abounding in yeblow colour-
ed ammonites of three or four species. None can be safely identified with
any Cutch species, though one formis very near A. (Stephanoceras) opts,
which is common to the Dhosa oolite and Kantkot sandstone of Cutch.

Jurassic beds are again found in the western half of the Salt
range and in its trans-Indus continuation, where the lower beds
rest upon the triassic rocks, and consist of sandstones of varying colour,
succeeded in ascending order by Hmestones, clays, and soft white sand-
stones, then come bands of hematite, several feet in thickness, and
thinner layers of golden oolite, precisely similar to the rock of Cutch,
the upper portion of the group consisting of coarse brown sandstones,
yellow marls, white sandstone and hard grey limestone bands. The
sandstones are often conglomeratic. and the limestones are most largely
developed to the westward. Small layers and patches of bright jetty
coal occur in places towards the base of the group and west of the Indus
near Kélabdgh, but there is nothing like a seam of coal. The patches of
coal appear to be merely carbonised fragments of drift wood.

The Salt range jurassic beds are not found east of the neighbourhood
of Naushahra. ‘They begin to appear a little further west than the triassic
ceratite strata and, increasing much in thickness, continue to the Indus.
West of that river the same rocks re-appear in the Maiddni (Chichdli)
hills, where they are well developed and more fossiliferous than in the Salt
range. They are well seen in the Chich4li pass, and extend round the curve
of the range further to the southward than the Productus limestone
does, disappearing beneath the tertiary rocks about six miles south of
Mulakhel,’ but are again found in the Shaikh Budin hills.

Until the fossils of the Salt range jurassic beds are examined in detail,
it is not possible to say exactly what members of the jurassic series are
represented, Dr, Waagen has shown that there is a close connection
between the Salt range oolitic beds and those of Cutch,® but that the Spiti
shales of the Himalayas contain a very different fauna. The Kelloway
portion of the Chari beds is distinctly represented in the Punjab, and some
of the higher jurassic groups also, Cephalopoda are scarce, except west
of the Indus, where Ammonites and Belemnites occur rather more

1 W. T. Blanford, Records, X, 16, (1877). (1859).
7 Fleming, Four. As. Soc, Beng., XXII, 278, | 4 Pal. Indica, series ix, 236, (1875).
Chap. IX-} SPITI SHALES. 229

abundantly, especially in the fine section of jurassic beds expesed in the
Chichdli pass.

The jurassic rocks of the Himalayas have long been known by their
fossils, which had acquired a sacred character and become objects of com-
merce. They are represented in the central Himalayas by a series of
dark grey and black shales, known as the Spiti! shales, whch contain
numerous calcareous concretions, a large proportion of which are formed
round fossils. The thickness of these shales is about 300 feet, and their
original extension has been very much restricted by the great denudation
they have undergone.

The Spiti shales are said to lie conformably on the underlying lime-
stones, ‘which were regarded by Dr. Stoliczka as lias, and the presence of
passage beds, in the shape of a small thickness?of clayey slates, is indicated,
In view of the probability that the beds regarded as lias are in reality
uppermost trias, and of the abrupt change of lithological character which
takes place at the base of the Spiti shales everywhere except in Spiti, it
seems more natural to suppose that there is an unconformity which has not
yet been recognised.

The Spiti shales are overlaid by about 600 feet of a light yellow silicious
sandstone, known as the Giumal sandstone, which was regarded by
Dr. Stoliczka? as upper jurassic in age, but has more recently been classed
as neocomian.®

Jurassic rocks are known to occur north of Nepal, characteristic fossils
having been brought from that direction by traders, and the Spiti shales have
been recognised to the north of the Karakoram range in one direction, and
in Hazd4ra inthe other. In Hazdra they are perfectly typical and are recog-
nisable both lithologically and palzontologically. In the Sirban mountain
they rest unconformably on a surface of triassic limestones, eroded and
pierced by boring molluscs. They are conformably succeeded by sandy and
calcareous beds, abruptly overlaid, but with no observed unconformity, by
a sandstone containing a lower cretaceous fauna,

Further south the Spiti shales have not been recognised with certainty,
but in the Suldimdn range some black shales, overlaid by sandstones, are
found below the cretaceous limestone of the Takht-i-Suldiman‘ and a similar
section is said to be observable in the eastern termination of the Safed
Koh,® but the identification lacks the support of fossil evidence. In the
southern part of the Hazdra district jurassic rocks are represented by a

1B. Stoliczka, Memoirs, V, 85, (1865). (1891).
3 Memcirs, V, 113, (1865). : * Records, XVII, 184, (1884).
*C. L. Griesbach, Memoirs, XXIII, 80, | 5 Records, XXV, 81, (1£¢2).
R 2
230 GEOLOGY OF INDIA—MARINE JuRASSIC ROCKS. (Chap. IX.

band,’ composed almost entirely of 7vivonda ventricosa, associated with
layers containing Ammonrtes, Gryphza and Belemnites, whose age relative
to the Spiti shales is undetermined.

This will be the most convenient place to mention certain fossiliferous
beds, underlying the nummulitics of the outliers in south-western Garhwél,
whose chief interest lies in the fact that they contain the only pretertiary
organic remains that have been recognised inthe outer Himalayas south of
the main snowy range. They were originally discovered by Mr. Med-
licott * in the Tal valley immediately east of the Ganges and have conse-
quently been frequently referred to as the Tal beds,

They are described by Mr, Middlemiss as grits or quartzites, frequently
calcareous and passing into limestone in places. The fossils are mostly
fragmentary, but among them he considered that corals, belemnites
lamellibranchs and gasteropods were represented, the whole indicating a
probably jurassic age, Subsequent critical examination, however, failed to
discover any specimens determinable with sufficient accuracy to indicate
the age of the rocks. Judging from their stratigraphical position they are
probably mesozoic, but beyond this nothing can be decided.

1 A. B. Wynne, Records, XII, 125, (1879). 8 Records, XVII, 161, (1884); XVIII, 73,
2 Memoirs, Ill, pt. ii, 69, (1864). (1885).
CHAPTER X.

MARINE CRETACEOUS ROCKS OF THE INDIAN PENINSULA.

Cretaceous rocks of Southern India— Relations to cretaceous rocks of Assam and South
Africa—Cretaceous rocks of the lower Narbadd valley—Relations to cretaceous of Europe—-
Contrast to cretaceous of Southern’ India—An Indo-African land connection—Doubtfully
Cretaceous sandstones of the Narbada valley and of Kathidwar.

The occurrence of cretaceous rocks in Southern India was first ob-
served in 1840 by Mr. Kaye of the Madras Civil Service, who, in company
with Mr. Brooke Cunliffe and others, collected a large series of fossils,
which were examined by Professor E. Forbes. The rocks near Pondi-
cherri had, however, some years before attracted the notice of Mons, E.
Chevalier, but no account of them was published until after the appearance
of Mr. Kaye’s description. A collection of fossils from the neighbour-
hood of Pondicherri was examined by A. D’Orbigny, and referred to an
upper cretaceous age. Forbes, on the other hand, referred the beds of
Trichinopoli and Viruddhdchalam to the age of the upper greensand or
gault, and the Pendicherri beds to the neocomian. It was shown by Mr. H.
F, Blanford that beds of two ages exist near Pondicherri, and he, follow-
ing Professor Forbes, considered the lower of these or Valudayur beds
neocomian and older than any of the Trichinopoli rocks, but the thorough
examination of all the Southern Indian fossils by Dr. Stoliczka has proved
that the general homotaxis is middle and upper cretaceous, and that the
neocomian and oolitic forms, which led to a portion of the beds being
originally classed as lower cretaceous, are less numerous than the middle
cretaceous species with which they are associated. It was also found
that the fauna of the Valudayur beds had more species than was at first
supposed in common with the lowest group of the Trichinopoli area, and
the two were consequently considered identical. ‘The Cephalopoda of the
lower beds comprise several species found in the gault of Europe, and the
number was at first supposed to be larger than it proved on subsequent
closer investigation, but as there are scarcely any representatives of gault
forms amongst the very numerous and beautifully preserved Gasteropoda
and Lamellibranchiata (Pelecypoda), the whole of the Southern Indian
beds were finally referred by Dr. Stoliczka to an age not older than the
232 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. ([Chap. X.
upper greensand of England (cenomanian), and ranging thence to the
upper chalk (senonian).

The rocks of cretaceous age in Southern India! occupy, with relation
to older and newer formations, a very similar position to that of the out-
crops of upper Gondwana beds farther to the northward. The cretaceous
beds occur in the great plain which extends along the Coromandel coast
from the north of the Goddvari to Cape Comorin. They rest to the
west upon the ‘gneiss, or occasionally upon small patches of the upper
Gondwana (R4jmahdl) beds, they have a low dip to the eastward,
and are covered up on the east by pleistocene beds, known as Cuddalore
sandstones, and by the alluvium of the sea coast. The cretaceous beds are
exposed at the surface in three detached areas, separated from each other
by the alluvial deposits. of the Penner and Vellar rivers. The southern
and largest of these areas, between the Vellar and Coleroon rivers, is in
the Trichinopoli district, and known as the Trichinopoli area. North
of Vellar ar two much smaller exposures near Viruddhdchalam and Pon-
dicherri respectively, and named from those towns.

The Trichinopoli area extends about twenty-five miles from north to
south, and for about the same breadth where widest, but it is very irregular
in form. South of the Coleroon (the principal outlet of the river Cauvery)
no cretaceous beds have been traced, and the southern boundary of the
cretaceous area north of the Coleroon is chiefly formed by gneiss, To the
northward the cretaceous rocks disappear beneath the alluvium of the Vellar
river and re-appear north of the river at Viruddhdchalam forming the
Viruddhdchalam area, in which only the highest cretaceous group is ex-
posed, and even this is only visible at very few points. It occupies a tract
of country about fifteen miles long from north-north-east to south-south-
west by about five broad, with gneiss to the west and tertiary Cuddalore
sandstone to the east. There is a second break in the rocks at the Penner
river, and alluvium extends to the neighbourhood of Pondicherri, causing
an interval of about twenty-five miles in the belt of cretaceous rocks before
they re-appear near Valudayur, ten miles west by north from Pondicherri.
Here they occupy a small tract of country about twelve miles long from
north-east to south-west, by six miles broad, and only separated from the
sea on the east by a band of Cuddalore sandstones two to four miles wide.
To the west is a narrower strip of Cuddalore sandstone, beyond which the
country consists of gneiss.

1 For a complete description of the geology
by Mr. H. F. Blanford, see Memoirs, 1V, pp.
4-217, (1862). The fossils are described and
figured in four volumes, comprising Series
i, iii, v, vi, and viii, (1861-73) of the ‘‘Paleonto-
logia Indica,” all by Dr. F. Stoliczka, with

the exception of the Belemnites and Nautili,
which are‘by Mr. H. F. Blanford. Some addi-
tional notes on the Cephalopoda are published
in the Records, I, 32, (1868), and on the fossils
generally, by Mr. R. B, Foote, in Records,
XII, 159, (1879).
Chap. X.] SOUTHERN INDIA. 233

In all three areas there appears to be a low dip to the east, the lowest
beds appearing at the western boundary and higher groups succeeding in
regular order to the eastward. Many of the dips seen in the rocks are,
however, deceptive, being due te oblique lamination or false bedding, which
prevails extensively throughout the series and especially in the southern
portion of the Trichinopoli area. In the Viruddhdchalam and Pendicherri
areas the rocks are ill seen, and the dips are less distinct, but there
‘appears every probability that the same low dip prevails in the Pondicherri
or Valudayur area; the direction js, however, south-east rather than east.

The series is divided into three groups named, in descending order
Ariyalir, Trichinopoli, and Utatir. The following table taken from the
Palzontologia Indica, exhibits Dr. Stoliczka’s final. views as to the
representation by these groups of the European cretaceous subdivisions :

 

 

South India. Eugland, France. Germany,

 

 

 

Zone of Nautifus danicus and| Upperchalk | Senonian .| Ober Quader.
Ammonites ootacodensis, Ostrea
a ae tr pectinata, and O. urgulate, Gry-
. phea vesicularis, Imoceramus
cripsii, Crania ignabergensis.
“ Zone of Ammonites peramplus, Pho-| Lower chalk | Turonian ./ Mittel Qua-
TRICHINO- ladompa caudata, Modiola typica, | der.
roLt GRoue. Ostrea dituviana, Rhynconella
. compressa, .
Zone of Ammonites rostratus and | Chalk marl | Cenomanian |Unter Quader
UTATEOR rotomagensis, Inoceramus labia-| and upper| or Tourtia, | Unterer Qua~
Group tus, Exogyra suborbiculata (Gry-| greensand. dersandstein,
; 2 phea columba), and Terebratula and Unterer
depressa, Pliner.

 

 

 

 

i

The Utatér group derives its name from a large village twenty miles
north-north-east of Trichinopoli. The beds composing the group are chiefly
argillaceous ; fine silts, calcareous shales, and sandy clays, frequently con-
cretionary and more or less tinted with ochreous matter, prevail through-
out the group, and in the southern portion of the area constitute almost
the entire bulk of the deposit. North of the parallel of Utattr, limestone
bands become intercalated in the lower or western part of the group and
sands, grits, and conglomerates in the upper or eastern part, these changes
in mineral character being accompanied by a great enrichment of the

: Pal, Indica, series viii, 1V, Introduction, | of misprints, the proof of the original was prd-
p. ti, (1873). As there are several slight errors, | bably not corrected by Dr. Stoliczka,

4
234 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [Chap. Xx,

fauna in the first case and an impoverishment in the latter. Conglomerates
are of very rare occurrence in the lower beds. Gypsum occurs in most of
the argillaceous strata, and is to a certain extent characteristic of the
group. The dips are often irregular, and apparently due to the original
deposition of the beds on shelving banks. This irregularity of dip renders
it impossible to form any trustworthy estimate of the thickness attained by
the group as a whole; it may, however, be roughly estimated as probably
not less than 1,000 feet,

At the base of the Utatuir group there are, in several places, large
masses of coral reef limestone, resting sometimes on the upper Gondwana
plant beds, more frequently on the gneiss, and occasionally on the lowest
beds of the Utatur group itself. The rock is a nearly pure pale coloured
limestone, compact and homogeneous, but often with a flaggy structure,
and frequently irregularly banded with white streaks, which, on weathered
surfaces, exhibit the corals of which they are composed, The mass of the
rock also sometimes abounds in corals, but more frequently no organic
structure can be traced, In lithological character this rock precisely
resembles the coral reef limestone of the present day, as described by
Darwin, Dana, Jukes, and other observers.

The usual position of this limestone is at the base of the Utatir
group, resting upon older rocks. ‘The coral reefs appear to have been
frequently exposed to denudation during the deposition of the later
Utatur beds, amongst which, in places, calcareous bands are found, ap-
parently derived from the waste of the reefs. The coral limestone now
remains in the form of smal{ isolated patches, scattered along the
western and southern margins of the cretaceous beds. In one locality,
however, close to the village of Caligudi, on the southern boundary
of the cretaceous area and twenty miles north-east of Trichinopoli, by far
the largest outcrop of the limestone in the area occurs at the base of
the Trichinopoli group. This outcrop is of considerable breadth, and
extends, with one or two breaks, for about six miles. From an examina-
tion of all the circumstances, however, it has been satisfactorily ascer-
tained that this outcrop also belongs to the Utatér group, and that the
Trichinopoli group rests unconformably upon it.

The coral reefs appear to have been scattered over the sea bottom in
shallow water, and probably along the coast, at the commencement of
the period during which the crétaceous deposits of Southern India were
formed, The remaining beds of the Utatér group were probably
deposited in water of moderate depth, and some of them appear to
have accumulated on submarine banks, possibly formed in tidal channels.
Hence the false bedding so prevalent in the rocks. The coarser con-
stituents of the rocks to the northward appear to indicate that the
current which brought the sediment flowed from that direction, and the
Chap. X. ] UTATOR GROUP. 233
occurrence of littoral forms of mollusca in greater abundance throughout
the northern parts of the area may be accounted for in the same manner.
The beds in the southern portion of the Utattr area appear to heve been
formed of fine silt, deposited in a bay where the force of the current was
less than to the northward, and the fossils which occur are mostly the
temains of pelagic animals, such as Belemnites, a few Ammonttes,
chiefly of Cristat/ group, or else peculiar forms of Vermetide (Tubul-
ostium atsiotdeam and T, callosum), which probably lived in the mud.
The Ammonites and Nautilz, which are numerous to the northward, are
scarce in the southern portion of the area. Cycadeaceous fossil wood,
sometimes bored by Yeredo and other Pholadidz, abounds in certain
parts of the group. On the whole, there appears every reason to believe
that the Utatir beds were formed in the neighbourhood of a coast line.

The distribution of the Utatar beds in the Trichinopoli district is very
simple. They form the western portion of the cretaceous area through-
out, their outcrop being in general from three to five miles broad, except
to the northward, where it diminishes in consequence of the beds being
overlapped by those of the next group, till, at the village of Olapddi in
the northern portion of the tract, the breadth of the Utatir outcrop does
not exceed half a mile. At the extreme northern point of the area, both
the Utattér and Trichinopoli groups are completely overlapped by the
uppermost subdivision.

The Utatir beds are not represented in the Viruddhdchalam area, but
they re-appear, as already mentioned, near Pondicherri. Here the beds
formerly classed as the Valudayur group, and considered neocomian by
Forbes, but which were shown by Stoliczka to contain several species of
fossils common to the Utatér group, consist chiefly, like the strata near
Utatir, of argillaceous beds, sandy shales, and sands, with occasional bands
of limestone and calcareous concretionary nodules. Conglomerates occasion-
ally occur amongst the lowest beds seen, but the most characteristic band
is composed of dark grey, compaet limestone in large nodules, sometimes
highly fossiliferous, Baculites vagina being the commonest fossil.

The area occupied by the Utattr or Valudayur beds near Pondicherri
extends from Valudayur for about nine miles to the north-east and is
about four miles broad. The beds are not seen to rest upon any older
formation. The country north and south is covered with alluvium. To the
eastward the Utatur beds disappear beneath the Ariyaldr group, and to
the westward beneath the Cuddalore sandstones of Tiruvakarai (Trivicary).
The beds to the westward appear to be the lowest, and there is a dip to
the eastward.

The fauna of the Utatdr group is very rich, no less than 297 species of
Invertebrata having been described from it. It has yielded an especially
large number of Cephalopoda, 109 species, 95 of which have not been met
236 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS, [ Chap. X,

with in the Trichinopoli or Ariyalér group. Of these 109 species, 27 are
known to occur in Europe, or elsewhere out of India, and although the
majority are distinctly and characteristically middle cretaceous forms, three
are, in Europe, neocomian species, viz. Nautilus neocomtensts, Ammonites
velledz, and A.-rouyanus, whilst no less than nine are found in the gault,
several of the latter ranging, however, into the upper greensand (ceno-
manian). Amongst the forms which are not European, the most remark-
able are three species belonging to the section of Ammonites known as
globost, which, amongst European rocks, are especially characteristic of
the triassic period. A very large proportion of the Cephalopoda were
collected in the neighhourhood of two villages, Odiam and Maravatir, on
the road from Perambalir to Arialdir, and about twelve miles north-east
of Utatir.

The Gasteropoda comprise only 43 species, a number far in-
ferior to that found in each of the other groups, and the majority of
these are littoral forms. The Lamellibranchiata (Pelecypoda) are 79
in number, the Brachtopoda g, Echinodermata to, and Corals 42, with one
species of sponge and one annelid. The forms found also in other coun-
tries belong almost without exception tothe upper greensand (cenomanian)
or higher groups, thus presenting a remarkable difference from the Cepha-
lopoda, in which gault forms are so largely represented. The only fossils
of much importance, besides the AZo//usca, are the corals, which, from the
prevalence of reefs at the base of the group, are superbly represented, no
less than 42 species, belonging to 23 genera, being known to occur.

The Trichinopoli, or middle group, of the Southern Indian cretaceous
series derives its name from the district of Trichinopoli, to which it is, so
far as present exploration extends, entirely restricted. To the south it
censists chiefly of sands and clays, very irtegularly bedded, with a few
bands of limestone and some conglomerates, and it differs lithologically
enly in one important respect, which will be described presently, from the
Utatir group. North of the parallel of Utatdr regular bands of shell
limestones become intercalated in the lower beds of the deposit and, to
the northward, the whole group is composed of regularly stratified alterna-
tions of sand, sandy clays, and shales, with bands of skell limestone,
calcareous grit and conglomerate. :

A peculiarity by which both the Ariyalér and Trichinopoli beds in
the southern part of the cretaceous area are distinguished from the
Utatiir is the occurrence of granite pebbles:in considerable quantity
in the gravels and conglomerates of the two former, whilst none are
found in the lower subdivision. In the Utatir group the materials of
the few conglomeratic or gravelly beds which occur are derived either
from the gneiss or from the coral reef limestone, whilst in the two
Ckap. X.] TRICHINOPOLI GROUP. 237

upper groups corglomerates are more frequently met with, and loose
masses of anigdeatined gravel and beds of rolled pebbles, alist entirely
composed of granitic materials and resembling the shingle of a sea
beach, are of common occurrence. The source of the granite pebbles
was evidently the broad belt of granitic rocks which forms the southern
boundary of the cretaceous area, dividing it from the alluvium of the Cauvery
throughout the greater portion of its extent, and the necessary inference
is, that this band of rock was in all probability beneath the sea during the
deposition of the Utatir beds, and that it was elevated above the water
in the interval between the Utatdr and Trichinopoli ages.

The Trichinopoli beds are, even more characteristically than the
Utatirs, the littoral deposits of a shallow sea. This is proved, not only
Ly the frequent occurrence of coarse sediment and the great irregularity of
the deposits in part of the area, but by the abundance of fossil wood, almost
exclusively exogenous and apparently cycadeaceous. Trunks of trees
are met with of great size, as much as three feet in diameter and sixty
feet in length, much of the wood being perforated by boring mollusca.

The shell limestone of Garudamangalam, east of Utatdr, and other
places isa very fine, hard, bluish grey, translucent rock, usually abounding
in beautifully preserved shells, both Gasteropoda and Lamellibranchiata,
which retain their original polish, and occasionally even the colouration of
their surfaces. This rock, known as Trichinopoli marble, is largely
quarried for ornamental purposes, and has yielded a considerable propor-
tion of the fossils found inthe group. The limestone occasionally contains
pebbles of granite or fragments of fossil wood, either of which is sufficient
to distinguish it, even when it is unfossiliferous, from the Utatdr lime.
stones,

The beds of the Trichinopoli group are unconformable to the Utatirs,
upon which they rest throughout the greater part of the area, the evidence
of unconformity not being confined to overlap, but depending chiefly upon
the proof afforded, by the rocks at the southern edge of the area, that the
Utatér beds had been disturbed and faulted, probably at the period of
upheaval of the granitic band already mentioned, before the deposition of
the Trichinopuli formation. Elsewhere also, the Trichinopoli beds rest
in places upon a denuded surface of Utatirs. There is also a great
change in the fauna. In the southern portion of their range the Trichino-
poli beds rest partly upon the coral reefs, which have been already shown
to be some of the lowest beds of Utatir age, and partly on the metamor-
phics, a considerable portion of the boundary being formed by the
granitoid rocks so frequently mentioned already.

The present group, like the Utatdr, is so irregularly bedded, and the
dips seen are so frequently those of original deposition, that no trust-
worthy estimate of the thickness can be formed. The general inclination
238 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. -[ Chap. xX,

is to the eastward, the average breadth of the outcrop is nearly the same
as that of the Utatar beds, and the same minimum thickness, v7z. 1,000
feet, may be assumed; the general dip of the bedding in the more regular-
ly stratified portion of the group to the northward is, however, lower
than in the underlying group, averaging about 6°. The beds thin out
greatly to the northward, and are at length completely overlapped by the
Ariyalirs.

It has already been stated that the Trichinopoli group is confined, so
far as is at present known, to the Trichinopoli area. Within that area it
forms a belt east of that formed by the Utatir group, and extending
similarly from south-south-west to north-north-east. The Trichinopoli
outcrop is, however, broader in the southern half of the area, where it is
about four miles across, than in the northern half, where it is in no place
more than two miles wide. It thins out and disappears completely about
two miles south of the place where the Utatirs are similarly overlapped
by the Ariyalir beds. Along the southern boundary of the Utatar area,
ceveral outliers of ‘richinopoli beds are found, resting partly on the Utatirs
and partly on the gneiss, and occasionally overlying the faulted boundary
between the two formations. ‘These small outliers, one of which, south of
Tirupatur, forms the south-western corner of the whole area, are composed
of coarse sands and conglomerates, usually unfossiliferous, but occasionally
containing Chemuzitzia undosa and other characteristic Trichinopoli fossils,
and the materials of which they are formed are derived chiefly fiom the
metamorphic rocks, but partly from the denudation of the Utatur beds.

The fauna of the Trichinopoli group, although not quite so rich as
that of the Utatdr beds, affords a full illustration of the life existing at the
period, 186 species of /uvertebrata having been described from these
beds by Dr. Stoliczka. The Cephalopoda are comparatively poorly
developed, only 23 species having been detected, and of these but.10, of
which four are European, are in India peculiar to the group. All the
Cephalopoda identified belong to the two genera Nautilus and Ammo-
uttes, the non-discoid Ammonitoid genera, such as Anisoceras, Scaphites,
Turrilites, etc., so largely represented in the Utattr group, as well as the
Belemnites, so abundant in the lower subdivision, being apparently want-
ing in the Trichinopoli beds. The Rotomagenses Ammonites, so charac-
teristic of the lowest cretaceous subdivision in Southern India, are also
wanting in the higher groups, with one doubtful exception. A few forms,
usually associated with older strata, still survive, however, such as Ammo-
nites menu, belonging to the Armazz (a jurassic group), 4. koluturenszs of
the Macrocephalus group, allied to such oolitic species as A. macrecephalus
and A. herveyt, and A. theobaldianus, one ot the Planulatz allied to
upper jurassic forms, such as A, éiplex, Most of the types found are,
however, characteristically upper cretaceous.
Chap. X.] ARIYALOR GROUP. 239

On the other hand, Gasteropoda, comprising 86 species, are much more
abundant than in the Utatur group, Lamedl‘branchiata, comprising 66
species, being rather less numerous. There are but 5 Arachiopoda and 6
corals, whilst no Echinodermata have been recognised. The Gasteropoda
include several siplhonostomate genera, rare in the older rocks, and not
found in the Utattr beds, the number increasing greatly in the next higher
subdivision, that of Ariyalir. The whole fauna exhibits a mixture of upper
and middie cretaceous forms, and appears fairly to represent the lower
chalk of England or the turonian of continental geologists.

The name of the highest group of the Southern Indian cretaceous series
is derived fromthe town of Ariyalir, which is situated nearly in the middle
of the comparatively large expanse of Ariyalir beds in the Trichinopoli
district. The country occupied by the beds of this group is much covered
with cotton soil, and sections are even rarer than in the two lower cre-
taceous subdivisions.

The Ariyalir beds are more sandy than the two lower groups and more
uniformly bedded, the beds being thick and homogeneous, consisting
principally of white unfossiliferous sands and grey argillaceous sands, with
casts of small fossils. Beds of calcareous grit and nodular calcareous
shales are found towards the base, and again in the upper portion of the
group, constituting two highly fossiliferous zones, separated by a con-
siderable thickness of deposits, in which fossils are rare or wanting,
although some interesting remains of a Megalosaurus were found in one
of the beds. A band of flints is associated with the uppermost beds.
There is a marked difference between the fossils of the upper and lower
zones in Trichinopoli, and it appears very probable that further examina-
tion of the rocks, now that the fossils have been compared and determined,
would justify the separation of this group into two—a_ probability
which was pointed out by Mr. H. F. Blanford at the time of the original
survey, although not shown on the map nor applied in the discrimination
of the fossils, because of the doubts which remained as to the distinction
of the two subdivisions in the Pondicherri area, where the fossils of
both upper and lower Ariyaldr beds appear to occur together. Conglo-
merates are of rare occurrence in the Ariyalir group, though a coarse
bed is found in places near the base, and there is but little irregularity
in the bedding, except close to the southern boundary. The constituents
of the Ariyaldr beds were derived chiefly from the metamorphic rocks,
amongst others from the granitic band to the southward, but a portion
of the sediment must have been furnished by the waste of some of the
older cretaceous groups, probably the Utatirs.

The above description of the lithological characters is principally
taken from the beds near Ariyalir, but it is also to a great extent
240 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [Chap. X,

applicable to the racks seen near Viruddhd4chalam and Pondicherri.
In both localities the Ariyalir deposits are chiefly represented by sands or
sandy clays, and by beds of arenaceous limestone or calcareous sandstone
at the base of the group. The strata appear to thin out to the northward,
and it is far from clear whether the uppermost fossiliferous zone ex-
tends in that direction, although some of its characteristic fossils, such
as Nautilus dantcus, occur abundantly near Pondicherri. It has not,
however, been hitherto found practicable to determine whether a distinct
upper zone exists near Pondicherri or whether representatives of the
upper fauna occur in beds of lower horizon than those in which the same
species are found near Ariyalir.

There is consequently some obscurity concerning the relations of the
beds belonging tothe Ariyaltr group amongst themselves, and this difficulty
is complicated by the circumstance that there is in many places an ap-
parent passage from the Trichinopoli group into the Ariyalur beds, the
rocks being similar in mineral character near the junction, and the fossils
being chiefly forms which appear to range from one group into the other.
It is highly probable that further examination of the ground, which, as
has been already noticed, is so much concealed by superficial accumula-
tions that the different groups can frequently only be traced by their fossils,
would show either that the number of groups or of palzontological zones
must be increased, or else, in some cases, that fossils supposed to have been
procured from the Trichinopoli group, have really been derived from the
Ariyalir, and vice versd.

The area occupied by the Ariyalir beds in the eastern portion of the
Trichinopoli tract amounts to about 200 square miles, or more than that
covered by both the other subdivisions together, the outcrop where broad-
est, near Ariyaldr, is about sixteen miles wide and extends for twenty-six
miles from north to south.

The Ariyalir beds also occupy the greater portion of a tract sixteen
miles long by five miles broad near Viruddhdchalam, and another about
twelve miles long from south-west to north-east, by two miles broad, west
of Pondicherri, whilst a very small exposure of them occurs close to the
coast ten miles north of Pondicherri, and another still smaller three miles
farther north.

The lowest fossiliferous zone is found resting upon the Trichinopoli
beds, throughout the western portion of the Ariyalur area in the Trichinopoli
district, and the same zone appears to be also represented-in the Virud-
dhdchalam and Pondicherri exposures, The great bulk of the outcrop in
all three tracts appears to consist of the thick sands, with but few deter-
minable fossils, forming the middle portion of the formation, whilst the
upper fossiliferous beds are only seen northof Ariyalir, near the villages of
Sainthoray, Niniyur, and other places farther north, in the long strip of
Chap. X.] ARIYALOR GROUP. 241

cretaceous rocks which forms the north-eastern extremity of the Trichino-
poli area.

Although the thickness of the Ariyaldr group can be estimated with a
nearer approach to probability than in the case of the two lower creta-
ceous formations, the estimate is still far from accurate. The dip of the
beds is very low, rarely exceeding two or three degrees, the general incli-
nation being north-east, and the whole of the beds in all probability do not
exceed 1,000 feet in Trichinopoli. Near Viruddhdchalam they appear to
be very thin, and in the neighbourhood of Pondicherri they are too obscurely
exposed for any estimate of their thickness to be attempted. There is an
apparent diminution of thickness to the northward, as in the other groups,
but this attenuation appears to be greatest near Viruddhdchalam, and takes
place less rapidly farther north, even if the beds are not thicker in that
direction.

The Ariyalir beds, as has been already stated, frequently appear to
pass into the Trichinopoli group at their base. They overlap the lower
groups however, both to the north and south, and there is, in places, an
appearance of unconformity where they rest upon the Trichinopoli beds,
nor is it easy to understand the very rapid diminution in the thickness of
the latter to the northward without supposing that they had been partially
denuded in pre-Ariyalir times.

As was noticed in the description of the Utatir group, the Ariyaldr
beds rest upon the Utatirs for a distance of rather more than two miles
in the northern part of the Trichinopoli area, and still farther north the
former were deposited directly on the gneiss. They also rest on the
gneiss throughout the whole breadth of their outcrop in the south of the
Trichinopoli area, and in the Viruddhdchalam cretaceous tract, whilst in
the neighbourhood of Pondicherri they are deposited to the eastward on
the Valudayur representatives of the Utatér group, and to the westward
no beds are seen beneath them, the Cuddalore sandstones covering the
boundary completely. Throughout the Trichinopoli and Viruddhdchalam
areas the Ariyalur beds disappear to the eastward beneath the Cuddalore
sandstones, which are unconformable to the cretaceous beds, and the latter
are covered up by alluvial deposits, intervening between the three areas,
in the valleys of the Vellar and Penner rivers, and also to the north of
the Pondicherri area.

The Ariyaldr beds appear to have been chiefly deposited in a tranquil
sea of small depth, although the deposits are less characteristically Jit-
toral than those of the Trichinopoli group, and the evidence of the

neighbourhood of land afforded by the occurrence of fossil wood is less
abundant.

The invertebrate fauna of the Ariyalir group exceeds in richness even
that of the Utatur beds, no less than 365 species having been detected
242 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [(hap. xX,

in the uppermost subdivision of the cretaceous rocks of Southern India.
The Cephalopoda comprise 36-species, Gasteropoda 138, Lamellibran-
chiata 117, Brachtopoda 12, Bryosod 23, Echtnodermata 26, Anthozoa
10, Foraminifera 1, and Vermes 2. It is highly probable that this large
number may be due partly to the circumstance that the Ariyalur deposits
comprise two groups differing somewhat in age. The lower fossiliferous
beds, from which the bulk of the fossils have been procured, correspond
very fairly with the senonian beds of France and the upper chalk with
flints of England. From this horizon all the Cepialopoda found in the
formation have been derived, with the exception of Nautilus danicus,
which was only observed in the upper beds of Niniyur, etc., in the Tri-
chinopoli area, although some specimens were obtained, apparently from.
a lower horizon, near Pondicherri. The fauna of these upper beds will
be noticed separately; the following remarks apply to the remainder of
the group.

In the Ariyalir beds, as in the lower subdivisions, there are some forms
of Cephalopoda which are in Europe characteristic of older beds.
These comprise two gault species of Nautilus, N. bouchardianus and
N. clementinus, Ammonites menu, found also in the lower groups, and
belonging to the jurassic section of armati, A. velledz, a lower and
middle cretaceous form in Europe, two macrocephali, A. deccanensis, and
‘A. arrialoorensis, and one of the Planulati,-A. theobaldianus. Very few
older forms occur in the other classes of mollusca, and the great majority
of the species common to Europe are found in the upper cretaceous beds
of England, France, and Germany.

The most striking peculiarity of the Ariyaltir fauna is the great abun-
dance of | Gasteropoda, and especially of the carnivorous prosobranchiate
forms, which, as is well known, appear to replace the Cephalopoda of the
older periods in tertiary and recent seas. Several genera not previous-
ly known from cretaceous beds have been detected in the Ariyaltic group,
and the Cypraide and Volutide are especially well represented. The
Lamellibranchiata are also very numerous, whilst all the Bryozoa and
the great majority of the A¢hinodermata hitherto found in the cretaceous
beds of Southern India have been obtained from the highest subdivision.
Lower forms of animals are but poorly represented. Amongst the Verte-
brata the only important species is a Megalosaurus,! of which a tooth
was found in the middle beds of the deposit, together with a number of
bones, which, however, could not be extracted ina sufficiently perfect
state for determination, The tooth closely resembles that of AZ. duck-
fandi, found in the Stonesfield slate and Portland odlites. of England, and
the occurrence of this genus in the upper cretaceous beds of India is of

1 Memoirs, \V, 128, 139: 186; —Records, X, 41, (1877).
   
    
    

Hemiaster similis, Q’Orbo;

 

Pecten quinquecostatus, Sow.

Aporrhais securifera, Forbes.

Aveliana ampla, Stol.

Tnoceramus simplex, 5tol,

 

 

 

 

 

Ammonites rotomagensis, Defrance,

Ammonites planulatus, Sow,
Caleutta Phototype Co,

CRETACEOUS FOSSILS.
 

ap” ae -

  

 

 

 

 

whan ae Ee ah aya aga Set seine
« :
a > ¥
wet 4 fod Be a, hae ”
cee
: nie gdh \
# PUTTAR As is etn Sh bo
. ailigh wel ge. g

Ps si a
ade ae et

Chee ee
Chap. X.] ARIYALOR FAUNA, 243

peculiar interest, because it only ranges from the lias to the wealden
in Europe. In this instance, as in several others, the land fauna appears
to have differed more from that which inhabited distant parts of the earth
than the marine fauna did.

The fauna of the uppermost Ariyaltir beds found at Niniyur and other
places to the north-east of Ariyaltr, comprises very few species which are
found in the lower portion of the group. Some of the fossils found most
abundantly, such as Nautilus danicus and Orbttoddes Saujast, are charac-
teristic of the uppermost cretaceous deposits of Maestrich, Aix la Chapelle,
and the Danish Island of Rugen (Danien of D’Orbigny). No other
Cephalopod except Nautilus danicus occurs in the Niniyur beds, whilst the
characteristically mesozoic genera /noceramus, Radtolites, Trigonita, Tri-
gonoarca, and Leptomaria, which are abundantly represented in the lower
portion of the Ariyalér group, are excirely wanting in the uppermost fossili-
ferous zone, where the only important mesozoic genus is Nerinza. On the
other hand, however, no typically tertiary forms make their appearance
except carnivorous Gasteropoda, and these are not more numerous in
proportion than in’the lower zone, although some additional forms are
represented.

Besides the fossils characteristic of each group, there are a few species
which are found throughout the whole series. The most important of these
are Nautilus huxleyanus, Ammonites planulatus (cenomanian, gault),
Ampullina bulbiformis (turonian, senonian), Gyrodes pansus, Solariella
radtatula (senonian), Vola guinguecostata (upper and middle cretaceous),
Ammonites menu, Forbes, is also supposed to be found in all three sub-
divisions, although there is some doubt about the Utatir beds, and a rare
Luctua, L. (Myrtea) arcotina, has also been procured from all the groups.
Some of these fossils, although found throughout the series, are especially
characteristic of one subdivision, as in the cases of Mautzlus huxleyanus
and So/ariella radiatula.

A large number of forms are common totwo groups. ‘The table on
page 245 exhibits the number of each class of /nvertedrata found in
the different formations, and the propertion found also in Europe, or
- common to two or more groups. The Vertebrata are represented by 17
species of fishes and one saurian, but the remains are of the most frag-
mentary. description, consisting in most cases of single teeth, and it is not
certain from which group some of the specimens were originally derived.

Adding these few vertebrata to the numbers given in the table, we
have a grand total of nearly 800 species of animals from the cretaceous
deposits of Southern India. Much time was devoted to the collection of the
fossils, and their exhaustive examination by Dr. Stoliczka has furnished the
best evidence extant for the correlation of any Indian fossil fauna with that

Ss
244 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [Chap. X,

of European rocks of corresponding age. Of the whole invertebrata 16°36
per cent. consist of forms known to occur in cretaceous beds in Europe.
The great majority of these are middle or upper cretaceous (cenomanian
to senonian), but there are amongst the Cepfalopoda several forms
which, in Europe, have only been found in lower beds (neocomian and gault),
whilst a few are representatives of European jurassic forms, and three
species of Ammonites belong to a triassic section of the genus. The
general facies of the cephalopodous fauna found in the lowest group, that
of Utatdr, approximates to that of the European gault, but nearly all the
species of the other classes of mollus:a found in the same beds belong to
a higher horizon, cenomanian (upper greensand), or even higher.

The whole of the cretaceous rocks of Southern India appear to have
been formed in shallow water, in the neighbourhood of a coast line, and
it is possible that the relative elevations of the country have undergone.
but little change since cretaceous times. Then, as now, there was higher
ground to the westward, and the ancient coast line appears to have
been approximately parallel to the present, although farther to the west.
We have thus in the cretaceous formation a confirmation of the evidence,
already afforded by the lower mesozoic deposits, that the Indian peninsula
is a land area of great antiquity.

Amongst the descriptions by Dr. Stoliczka, of the cretaceous fossils
from Southern India, the following species of Lamellibranchiata are in-
cluded.from Sripermatir, twenty-five miles west-south-west of Madras,
already mentioned as the typical locality for a group of the upper
Gondwana series :—

Sphariola, sp. indet. | Pseudomonotis fallactost.
Hippagus emilianus. a5 inops.
Yoldia obtusata. Lima oldhamiana.
Trigonoarca galdrina., Pecten arcotensis.

Two of these, Yoldia obtusata and Trigonoarca galdrina, are also
found in the Ariyalir group of the Trichinopoli district. but the identifica-
tion of the Sripermatir species referred to the Yoldza is slightly open
to doubt. Trigonoarca galdrina is, however, a well marked form, and
it belongs to a characteristically cretaceous genus,

The specimens were collected by the late Mr. Charles Oldham before
the country was properly examined, and there appears some slight doubt
as to the precise beds from which they were obtained. Some of the
specimens were from Sripermattr itself, others from Rajah’s Choultry.
The only cretaceous fossils found by Mr. Foote, who mapped the country
in the Sripermatér neighbourhood, occurred in water-worn blocks of grey
or greenish grey gritty sandstone, resting loosely on the surface of jurassic
Ohap. X.] PALEONTOLOGICAL RELATIONS. 248

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 ‘adoung 2 2. Qa A FT 8 DO Hm ix
“ Ul ose = puNoy = [B30], Sy
. "Solas sn0a9e}9I9 UeIpUl g 6 faa?” ¢ mot NS
y e yinog ur saivods [ejol mw MON w
is
8 =
<= © *sdnos 9314} Shere 4s ke oo
s = NB 0} uoww0d saladg . : :
8 dod
“sdnol ‘Oo - ~ “4 al
~ so japreany = pue yodou : : : : : 2
x “$Y T 07 uous saadg ‘ fe .
=
” *sduoid appedtry pue So Ss . : . : fe ‘ =
S = |rgve3¢q 03 uowuios saisadg ne on ee
<<
S o | ‘sdnost yodouyouy pue T Be : : : : : : . oo
ce ™ 1 yOIBIQ, OF uOUTUIND satzadS 2 = - ‘: = © :
4 ses
2 a ‘adomy ul ste ae a = SS - ea
g punoz g uwinyjod ur saisadg
8
‘Oo an
3 Ps _ terpuy uy g, e2@se? 8 8 8 gS 8
x dnoig nye stay 0} 1e1jud9g a) ed * se]
i x |'dnoid snypedisy ut punoy ‘Se ee aes a s a g if oS
+” saisads yO Jaquinu yejo) - a
=
4 © ‘adoing ut ose SO See ie : : er Sey as : a
= punosz $ uwnyod ut saisadg * : 2 * *
a
S
Q oy “erpuy uy duoiz Se ae gel Ne tay &S
(odours, 0} serpnday ot
%
~
8 “duos oo © WH re) ‘9
x + yodouryo1y «af ~=punos an Oo 0 < = i : 2
oe saads Jo Jaquinu 18307, : z
.
% o ‘advoing ul ose re 20 2 ae a 2
= punoj z umnjod ur satoadg
N ~
‘ a a
w a eipuy ur a 2 oS 2 3 Burg + oO
8 dnos3 pyeyp, 0} JeINIVg : a
&
> 1
3 7 *dnoi3 3nje379 Ul punos Si 2 £ OF mB + in a
a saivods jo Jaquinu [eyo "
"es
S eos an ar . . . °
*
4
S
Q iim cae Get. his ee BE . 3
: é
= °
= . e . . ° . . . ° * ol
dg
& ~
s
a e . . . ‘ . . . . .
°
~
i . . ee . is . e . ° .
& s .
im "5 s
© « 2 « . os ‘ £
es ea ee 2 § 2
a 2 ro ° o a noe
oS 2 OS og Do © .o 8
pad Oo am ° oO w } NR = =
w uw mS so 8 co) € 3° wh
= 2 De ry ° © 3h se c a
a & a | ~ ° w
o w 3 £ c e oO < a 9
» n ia
Vv oH meme «=

$3
246 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [Chap, X.

beds near Sripermatir.! The origin of these boulders could not be traced,
and the fossils cannot now be found; amongst the forms obtained were
four or five species of Ammonites, some Belemnites, etc.

As will be presently shown, there is a great difference between the
fauna of the cretaceous rocks in Southern India and that of the deposits
of similar age on the Narbadd, but on the other hand many of the fossils
of the Trichinopol: area are found in the cretaceous rocks of the Khasi hills,
to the north-east of Bengal, between Assam and Sylhet. So many species
indeed are common to the Trichinopoli and Khdsi deposits that it is prob-
able that the two regions were part of the same marine area. The creta-
ceous rocks of the Khdsi hills are almost unquestionably identical with
those extending throughout the hill ranges south of Assam and the same
strata are probably represented in Arakan. The description of these rocks -
will be best deferred to a subsequent chapter,? but the paleontological
results may be noticed here.

In a small collection of fossils Dr. Stoliczka*® recognised the following
species. The highest fossiliferous band, about 200 feet below the edge
of the cliff at Maosmai, a coarse sardy limestone, contains small Lamel-
libranchiata, a Cellepora, and echinoderms; a finer rock is principally
made up of an Astrocenza, allied to A. decaphylla.

From about the middle of the series, above Mahddeo, ina stream under
Laisophlang, in a soft, ochreous, glauconitic sandstone these fossils were
found :—

 

Nautilus (? N. elegans). Phasianella.
Nautilus, with a central siphon; Turritella.
fragments. Euspira,
U.T. A Ammonites planulatus. Dentalium.
U. » dispar. Faniva, near F. fleuriausiana.
U. 55 orbignyanus. Exozyra matherontana.
A. » ? pacificus. Spondylus striatus.
U~, Anisoceras indicum. Modiola typica.
U. +3 Subcompressum. T. Cardita jaquinoti (orbieularis) .
U. = A. Baculites, near B. vagina, Cardium.
T. A. Alaria papilionacea. Terebratula, near T. carnea.
T. A. Rostellaria palliata, Turbinolia,
T. Gosavia indtea. Hemiaster.
T. A. Cerithium inauguratum. _foloster
T. Tritonidea requieniana, Brissus.
T. Hemifusus cinctus.

The facies of this group rather resembles that of the Utatdr beds of
Southern India.

1 Memoirs, X, 61, (1873). the lists the letters U,T, A prefixed signify that
2 Infra, p. 298. the species is foundin the Utattr, Trichinopoli,
3 Stoliczka, Memoirs, VII, 181, (1869) ; in | and Ariyaltr groups respectively.
Chap. X.) RELATIONS TO SOUTH AFRICAN CRETACEOUS. 247

From the well known fossil locality about two miles from Tharia, on the
fourth crosscut taken by the footpath between the zigzags of the road

to Cherra Punji, or the first below the Devil’s Bridge, the following were
named :—

 

A. Nautilus levigatus T. A. Gyvodes pansus.
U, A. Baculites vagina. A. Gibbula granulosa,
A Cyprea globulina. A. Nerita divaricata.
» prlulosa. A. Euptycha larvata.
T. A. Rostellarta palliata, A Actéon curculio.
T. A. Alaria tegulata. Pecten septemplicatus.
T. »  glandina, Fanira quadricostata,
A. Lyria crassicostata. A. Gryphea vesteularis.
A, Volutilithes septemcostata Spondylus striatus.
T. Tritonidea requientana. Pecten, near P, rugosus.
T.  Lathtrus veussit. Inoceramus
A. Pseudoliva subcostata. T.  Rhynconella compressa.
A. Turritella pondicherriensis. Terebratula, sp., probably 7. bipli-
T.A. multistriata,- cata and T. carnea.
U. A? Mitreola citharina. Ananchytes ) several species, but
A. Euspira lirata. ; faiatince from any de-
Brissus scribed.

Nearly all the fossils of this list occur also in the Ariyalir group of
Southern India, but there are a number of species in the Tharia beds
which appear to be peculiar, and most of them new. It may be worth re-
cording that the observer who collected these fossils considered the latter
locality to be lower in the series than the former, which would be remark-
able, if true, seeing that the Ariyaldr group is newer than the Utattr.

Before quitting the subject of the Trichinopoli and Assam cretaceous
beds, it is necessary to notice the very remarkable resemblance between
a portion of their fauna and the species found in certain strata in South
Africa! In the description of the Gondwana system, and again in the ac-
count of the upper jurassic beds of Cutch, the remarkable affinities between
Indian fossil plants and animals and the forms found in South African beds
were repeatedly noticed, and there is a similar connection between the
cretaceous formations in the two regions. In some deposits found resting
upon the Karoo beds on the coast of Natal, 22 out of 35 species of mollusca
and echinodermata collected and specifically identified, are identical with
forms found in the cretaceous beds of Southern India, the majority being
Trichinopoli species. Amongst the South African fossils are some of
the commonest and most characteristic fossils of the Southern Indian creta-
ceous deposits, namely Ammonites gardeni (Ariyalir), A. kayed (Utatur)

'C.L. Griesbach, Quart. Four. Geol. Soc. | described by Buily, Quart. Four, Geol, See;
XXVII, 60, (1871. Some of the fossils were | XI, 454, (1855).
248 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. (Chap. X,

Antsoceras rugatum (Utatir), Pugnellus uncatus (Trichinopoli), /as-
stolaria 1tgtda (Trichinopoli), Chemniteta undosa (Trichinopoli), Eu-
chrysalts gigantea (Trichinopoli and Ariyalir), Solarze//a radiatula (ail
three groups), Ave/lana ampla (Trichinopoli), Zurrétella multistriata
(Trichinopoli and Ariyalir), Pecten (Vola) guinguecostatus (all three
groups) and Cardium hillanum (Trichinopoligroup). There is also some
slight indication of a representation of the different Indian zones.

From the cretaceous rocks of Madagascar six species of cretaceous
fossils were examined by Mr. R. B. Newton! in 1889, of which three, AJlec-
tryonia (Ostrea) ungulata, A. pectinata and Gryphea vesicularis are also
found in the Ariyaldr group, the other three species being typical neo-
comian belemnites, from a different locality and evidently from rocks of an
older date.

‘The South African beds are clearly coast or shallow water deposits
like those of India; the great similarity of forms certainly suggests
continuity of coast line between the two regions, and thus supports the
.view that the land connection between South Africa and India, already
shown to have existed in both the lower and upper Gondwana periods,
was continued into cretaceous times. It is very surprising to compare
the middle cretaceous fauna of Southern India with that of the distant
beds of Natal, and then with the widely differing forms found in beds of the
same age in central India and southern Arabia,

The marine cretaceous formations found in the western portion of the
Narbadd valley have been commonly known as Bagh beds, from the
town of Bagh, situated about go miles west by south of Indore and 35
miles west-south-west of Dhdr. The town is not on cretaceous rocks,
though they are well developed in the neighbourhood. The occurrence of
cretaceous fossils near Bagh was discovered by Colonel Keatinge? in 1856,
but the existence of fossiliferous limestone inthis part of the Narbad4
valley had been known for a long time, although the exact locality had
not been ascertained. The circumstance that blocks of limestone, con-
taining fragments of Bryozoa and cther fossils, had been employed in
building the houses of Mdndogarh, a city now in ruins about twenty miles
south of Dhar, first attracted attention, and it was mainly owing to an
ingenious and happy suggestion by Dr. Carter? that attention was attract-
ed to the neighbourhood of Bagh, where limestone had been observed in
1818 by Captain Dangerfield.

: Quart, Four. Geol. Soc., XLV, 333, (1889). | 2 Four, Bom, Br. Roy, As. Soe., V, 238, (1857).
Four. As. Soc, Beng., XXVIU, 116, (1858). Geological papers on Western India, p. 685.
Chap. X.] BAGH BEDS, 249
The cretaceous rocks of the lower Narbada valley! occur chiefly along
the edge of the Deccan traps, and intervene between the latter and the
metamorphic rocks. West of Bagh the outcrop of the cretaceous beds
may be traced with a few interruptions to the neighbourhood of Baroda.
East of Bagh they only occur in places around the inliers of older rocks,

As a general rule, the Bagh beds are composed of a calcareous rock
above and of sandstone below, but the character of each portion of the
formation varies, Commencing tothe eastward, the first place where
marine cretaceous beds are krown to occur is in the neighbourhood of
Bdrwai, on the Narbad4, nearly due south of Indore. Here some con-
glomerates, more or less calcareous, and sandstones containing marine
shells, represent the cretaceous formation, and in one place are seen to
be distinctly unconformable to an outlier of Mahddeva conglomerate
belonging to the upper Gondwana series. From the neighbourhood of
Bdrwai the whole Narbadé valley is composed of trap for nearly 50 miles
to the westward. Lower rocks re-appear near Mdandogarh, between
which place and Bagh the cretaceous beds are found, forming a narrow
fringe to the traps, around several inliers of Bij4war and metamorphic
rock. ;

The Bagh fossiliferous beds are divided into three zones, all calcareous
underlaid by a variable thickness of conglomeratic and sandy beds. The
fossiliferous zones are known as (1) the nodular limestone, (2) the Deloa
and Chirdkhdn marl, (3) the coralline limestone.

The nodular limestone group is the most extensive of the three, being
found in all the exposures, but at the eastern and westernmost outcrops
the peculiar nodular character is wanting. It is an argillaceous, whitish,
compact and generaily nodular limestone.

The Deola and Chirdkhdn marl is a soft band, never more than 10 feet
thick, which is chiefly interesting as having yielded the majority of the
fossils. The coralline limestone is the rock of which Méandogath is built
it is yellow or red in colour, and consists chiefly of small fragments of
Bryozoa shells, etc, The freshly broken surface has a somewhat granular
mottled appearance, and the fossils are not conspicuous, except on the
weathered surface.

The two upper groups do not extend so far westwards as the lowest,
the most western known exposure of the coralline limestone being at Umridli,
near Ali.

The total thickness of these groups united does not exceed 60 or 70 feet
at its maximum, and they thin out to the northwards, attaining their great-
est thickness along the southern limit of the exposures.

The fossil fauna obtained from these beds is an interesting one, though

1The description of these rocks is based | 163, (1869), and P. N. Bose, Memoirs, XXI,
ptincipally on W. T. Blanford, Memoirs, Vi, \ 35, (1884).
250

GEOLOGY OF INDIA—PENINSULAR CRETACEJUS.

[Ohap. X.

ameunting to only forty forms in all. Of these the late Prof. Duncan

determined the following :1—

 

 

 

Coral-
ee Deola tae
Ime- | marl, | lime-
stone, stone,
LAMELLIBRANCHIATA—
Neithea alba* - * ) P. quinquecostata in Europe (gault
to lower chalk); Africa, Utattr,
Pecten quadricostatus % Trichinopwli and Arialtr.
BRACHIOPODA—
Rhynconella depressa x Europe (lower greensand).
Bryozoa—
Escharina, sp. *
Eschara, sp. x
f&cHINODERMATA—
Cidaris namadicus x % | Lebanon.
Salenia traasii : 3 x ais
Cyphosoma cenomanensis * % | Europe (cenomanian).
Orthopsis indicus . x ses
Echinobrissus goybeti : x .. | Lebanon.
Nucleolites similis ale * x Europe (chloritic marl).
Hemiaster cenomanensis ¥% xX x Lebanon, Europe (cenomanian),
H. similis * x % | Europe (cenomanian).
VERMES—
Vincularia,sp. . . *
Serpula plexus e . *
CORALLIA ~
Thamnastreaa decipiens x Europe (neocomian to gosau).

 

 

 

To which Mr. Bose? has added the following :—

CE PHALOPODA—
Ammonites guadeloupi . ‘
GasTERKOPODA—
Fulguraria elongata.
Lyria granulosa . - :
Fasciolaria rigida . ° .
Triton, sp. . 7 . .
Natica, sp. . .
Cerithium, sp. . F
Turritella, sp. 4
LAMELLIBRANCHIATA—
Ostrea leymeriit. ‘< ‘
O. arcotensis c .
O.sp. . . ; . .
Radula obliquistriata . 5

Plicatula multicostata

Modiola archiaci . ?
Inocera nus concentricus ‘
I. coguandianus . 3

I. multiplicatus  . ‘ :

 

KK ROKK OK Ok

oe

 

* OK

2k Ok OK OK OK :

K:

2 ook:

 

 

Europe, Trichinopoli.

Europe (cemonanian and senonian) ;
Trichinopoli.

Africa ; Utatér and Trichinopoli.

Trichinopoli.

Ariyaldr, Europe (neocomian).

Ariyalac.
Trichinopoli.
Europe (neocomiar).
Europe (gault).
Europe, (gault).
Trichinopoli.

 

 

1 Quart. Four. Geol. Soc., XX, 349, (1865) ;

Records, XX, 81, (1887).

2 This and the following species were regarded
by the late Dr. F, Stoliczka as but doubtfully dis-

tinct from each other and from P. (Vola) quingue-
costata. ;

8 Memoirs, XXI, 37, 40, 43, (1884).
Chap. x PALZZONTOLOGICAL RELATIONS.

 

 

 

 

 

 

 

251
oral-
Nodular] teola nine
stone, | ™arl. stone.
Inoceramus, sp. ; . >| X a
Pinna laticostata . 5 . ‘ bi x Ariyaldr, Europe (neocomian).
Area securis x is
Cardium altum % x Utatitr.
C. hillanum . ss % Trichinopoli.
Venus, sp. - * x
Panopeaarcuuta . x ie Europe (neocomian).
BRACHIUPODA—
Rhynconella plicatiloides x % =| Trichinopoli, Aryalér,
R., sp. + . ‘ x :
Potyzoa— :
Ceriopora dispar . ‘ j | * i x % =| Ariyalir.
|

 

 

In the most recently published account of these beds an attempt has
been made,! to correlate the three zones with the three great groups of
Southern India and to the European groups ranging from albian (gault)
to senonian (chalk), To say nothing of the improbability of the greater
portion of the cretaceous period being represented by some 50 feet of
fossiliferous beds, there is no paleontological evidence in favour of the
supposition. Seven at least of the species of the lowest bed range into
the Deola and Chirdkhdn marl, and the same number of species are
found common to this and the coralline limestone. Three species (Cerzo-
pora dispar, Hemiaster cenomanensts, and H. stmzlis) range through all
three beds. Considering the limited number of species found in the upper
and lowermost beds the palzontological resemblances are greater than
would be consistent with a range of time from cenomanian to senonian.
Nor do the paleontological relations of the fauna to those of other parts
of the world ‘bear out the supposition. None of those forms which have
been determined by the late Professor Duncan are found in Europe in beds
of other than cenomanian or turonian age, and of the forms identified by
Mr. Bose with European species of albian and cenomanian age, an identifi-
cation confessedly rough, at least half are forms whose specific identity or
the reverse is difficult for any one but a practised paleontologist to deter-
mine. Omitting those species whose specific identity is open to doubt and
confining ourselves to those forms which have been determined by Profes-
sor Duncan, all those which are found in Europe occur there in beds of
cenomanian (upper green sand) age and most are characteristic of it,
consequently the cretaceous rocks of the Narbadd valley must closely
correspond to the Utatur group of Southern India.

In contrast to the relation between the cretaceous faunas of South
Africa and SouthernIndia may be noticed the divergence between the

1 Memoirs, XXI, 48, (1884).
252 GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. (Chap. X,

latter and that of the Narbad4 valley. Amcng the fossils identified by Pro.
fessor Duncan, Necthez alpina and Pecten quadricostatus, were regarded
by Dr Stoliczka as doubtfully distinct from P. gucnguecostatus found in the
Utatér group, but with this exception no other species is found in the creta-
ceous of Southern India, and even in this case the identification is one on
which palzontologists are not thoroughly agreed. Thamuastrea decipt-
ens is replaced by a closely allied form 7. Azevroglyphica, and the two He.
mtasters by remotely allied forms. The more recent additions to the
fauna made by Mr Bosc have yielded thirteen species apparently identical
with Southern Indian forms, but it is probable that this number will be
reduced when the fossils come to be more critically examined, and five of
the thirteen are wide ranging species found also in Europe or South Africa.
Even accepting the identifications, this number out of a total of forty
distinct forms is a uch smaller proportion than obtains in the case of the
south African and Trichinopoli cretaceous beds,

Another contrast between the Madras and Narbadé valley cretaceous
beds is the very small proportion of European forms found in the former
and the large proportion in the latter. Of the eight species of Echino-
dermata four are also found in Furope and two more in the Lebanon, and
of other orders Nerthea alpina, Pecten guadricostatus and Jhamnastraea
decipiens are found in Europe. To these Mr. Bose has added eight species
also found in Europe. In the intervening area, two small exposures of
cretaceous rocks are knawn at Ras Fartak and Ras Gharwen on the
south-east coast of Arabia, from which small collections, comprising but
thirteen species have been examined.! Yet, three of these are also found
in the Bagh beds, and no less than ten in Europe. It is evident from this
that there must have been tolerably free communication between the seas
in which these different exposures were deposited, and that they belong
to one of the great marine provinces of the cretaceous epoch,

With the South African cretaceous fauna the relationship of the Bagh
fauna is of the slightest, only two species, Pecten quinguecostatus and
Lyria granulosa, being common to the two regions. Of these the former
is a species of almost world-wide range, and the latter is also found in the
cretaceous of Southern India.

The contrast between the faunas of Narbadd and South Indian creta-
ceous is consequently as strongly marked as the relationships of the foriner
to the European fauna and of the latter to that South Africa. Yet the dis-
tance which separates the two Indian exposures is but 750 miles, only half
the distance which separates the lower Narbadd exposure from Arabian
localities, and a much smaller fraction of the distance to the European
localities or between Trichinopoli and South Africa.

'P, M. Duncan, Quart. Four. Geol. Soc., {| by Dr. Carter, Four., Bombay Br. Roy. As. Soc.

XXI, 349, (1865). The Arabian localities were | IV, 71, (1853), and ‘Geological Papers on
originally described, and the fossils collected, | Western India,” p. 603, (1857'.
Chap. xX.) NIMAR SANDSTONE. 253

These apparently anomalous relationships and divergences between the
cretaceous faunas are easily explicable by, and are indeed proof positive of,
the supposition that dry land stretched continuously from India to Africa
during the cretaceous period and formed a barrier between two distinct
marine provinces.

Besides the fossiliferous beds of the lower Narbdda'valley already noticed,
a lower division has been described under the name of the Nimér sand-
stone, whose age is open to doubt. The frequent presence of a band of
conglomerates and-sandstones underlying the nodular limestone has already
been noticed, In fact, it is only at or near Kachaoda, in the Man valley, that
its absence is recorded Asarule, the thickness is small in the eastern
exposures, but they begin to thicken rapidly south-westwards of Bagh,

In the Hatni valley they have a thickness of nearly 200 feet ; near the
deserted city of Ali they are 590 feet thick, and in the inlier south of
Kawant they amount to over 700 feet. With all its variation in thickness
this sandstone preserves the same general type, of conglomeratic beds and
conglomerates at the base, overlaid by fine grained sandstones and shaly
beds,

The age of this sandstone is doubtful. It has been regarded as lower
cretaceous, and in favour of this supposition there is the absence of any.
observed unconformity between it and the beds it underlies. Mr. Bose
classed it as neocomian on the strength of an oyster bed, composed of a
species which was identified with the European O, /eymeriz, but even if the
identification were correct the evidence is uot sufficient to establish the age
of the bed, and there is some doubt whether the dyster band is really
conformable to the underlying sandstone or not.

_ The sandstone contains no recognisable fossils, the uppermost beds
show crustacean andannelid tracks on their surface; some indeterminable
fragments of bone were found and fragments of carbonised driftwood are
said to occur. In this, as well as in lithological facies, the Nimdr sand-
stone agrees with the Mahddevas of the Dhar forest area. In Kathidwdr
there is the same absence of beds intervening between the upper Gond-
wdnas and the ecretaceous as would be implied by a Mahadeva age for the
Nimédr sandstones, but the absence of any observed unconformity, the
constancy of the nodular limestone beds, and the fact that the thickening
of the sandstones is in the same direction as that of the upper beds lend
some support to the supposition that they are cretaceous. In the absence
of more conclusive evidence their true age must remain a matter of doubt,

In Kathidwér a series of sandstone beds known as the Wadhwédn sand-

1 Memoirs, XX1, 23, (1884). | 2 Memoirs, XXI, preface to pt. ii, p. vii, (1884.)
|

aba GEOLOGY OF INDIA—PENINSULAR CRETACEOUS. [Chap. X,

stones are found between the Urmia group and the overlying Deccan
traps.| They are composed of brick red or dull reddish brown sandstone
with some argillaceous beds, at the top of which are in places cherty beds or
thin bands of limestone, recalling the rocks of Bagh. In the neighbourhvod
of Wadhwan, there is a thin band of drab coloured, tough, sometimes gritty
or chalcedonic limestone, containing marine fossils, chiefly Bryozoa, a
few small corals and a portion of a flattened, keeled ammonite, resembling
the cretaceous A. guadaloupae, in a matted mass of broken indetermin-
able shells. In other localities Ostrea and Natica were found, but no [ossils
sufficiently well preserved to determine the precise age of the beds. The
general facies of the fossils, as well as the lithological character of the rock
they are preserved in, suggest the correlation ef these beds with the creta-
ceous of Bagh, and their direct superposition on the Wadhwén sandstones
corresponds to the relation between the B&égh beds and the Niméar sand-
stones, whose probably upper Gondwana age has already been indicated.
The relation of the Wadhwan sandstones to the overlying trap appears to
be very similar to that of the Bagh cretaceous, there being a distinct erosion
unconformity.

1 F, Fedden, Memoirs, XXI, 87, (1884).
CHAPTER XI.

DECCAN TRAP.

Extent—Nomenclature— Petrology—Igneous formations— Sedimentary beds—Lameté group —
Infra and inter-trappeans of Rajamahendri—Inter-trappeans of the main area— Subaérial

origin of the trays—Foci of eruption—-Age of the Deccan trap—Probable conditions
during their formation.

In the last chapter the cretaceous rocks of the Peninsula were described,
and the present should, in the ordinary course, be devoted to the rocks of
the same age in the extra-peninsular area. There are, however, two very
good reasons for departing from this course, the first being the intimate
relation which subsists between the cretaceous and tertiary rocks of a
large portion of extra-peninsular India, necessitating their being dealt
with together; the other is the presence, immediately above the creta-
ceous beds in the Peninsula, of a series of volcanic rocks, forming one of the
most prominent and widely spread of all the rock systems found in the Pen-
insula. ,

In superficial area the Deccan traps are only exceeded, within the
limits of peninsular India south of the Indo-Gangetic plain, by the meta-
morphic seri¢s and, although the traps are far inferior in thickness to the
Vindhyan and Gondwana formations, their remarkable horizontality, through-
out a great part of the region covered by them, enables them to conceal
all older rocks. Some faint idea of the extensive area occupied by this
formation may be gained from the fact that the railway from Bombay to
Nagpur, 5r9- miles long, never leaves the volcanic rocks until it is close
to the NAgpur station, and that the traps extend without a break from
the sea coast at Bombay to Amarkantak at the head of. the Narbada, and
from near Belg4um to north of Goona__ Even this extent, great as it is, by
no mean represents the whole area originally occupied by the formation ;
for outliers are found east of Amarkantak as far as Jamird Pat in Sargija,
to the south-east a small outcrop occurs close to RA4jdmahendri, whilst to
the westward the series is well developed in Kathiawaér and Cutch, and is
even believed to be represented, though only by two very thin bands west
of Kotri, in Sind. We have, therefore, proof of the existence of this vol-

canic formation throughout nearly ten degrees of latitude and sixteen of
 

256 GEOLOGY OF INDIA—DECCAN TRAP. [Ohap. XI.

longitude, whilst the area covered in the Peninsula of India can be little
less than 200,000 square miles. It is probable that the limits mentioned
very nearly correspond to the original boundaries of the-volcanic rocks,
because the high level laterite, which rests conformably upon the upper-
most traps of the Deccan, is found to the southward, eastward and north-
ward, resting on rocks older than the volcanic series, and if, as will be
shown to be probable in a later chapter, this laterite was formed at a date
shortly subsequent to the cessation of the igneous outbursts, it may be
inferred that the lava flows never extended to the localities (such as
Gwalior, Rewé, etc.) in which the laterite is found resting immediately
upon Vindhyan, transition, or metamorphic rocks,

In adopting the name of Deccan! trap for this great volcanic forma-
tion, the Geological Survey has been guided partly by old usage, partly
by the circumstance that the term ‘ trap’ was originally applied to similar
horizontally stratified lava flows. Some geologists have condemned the
term on account of the loose manner in which it has been used for a great
variety of igneous rocks, but it is difficult to replace it, and in the present
case, at all events, it is employed in a well defined sense.

In consequence of its geological structure, the volcanic region of cen-
tral and western India is distinguished by marked peculiarities of scenery,
and the characters of the surface are widely different from those found in
other parts of the Indian Peninsula. Great undulating plains, divided from

 

Fig. 15.—Hill composed of Deccan trap, near Harangaon, north of Nimdwar, Narbadé valley.

each other by flat topped ranges of hills, occupy the greater portion of the
country and the hillsides are marked by conspicuous terraces, often

1Tt is scarcely necessary to state that the | Indian Peninsula which is south of the Vin-
Deccan (Dakshin) comprises that part of the | dhyan range,
Ohap. XT.] PETROLOGY. 257
traceable for great distances, and due to the outcrop of the harder basaltic
strata, or of those beds which resist best the disintegrating influences of
exposure. in some parts of the area great scarps are found, some of those
in the Sahyddri range being 4,090 feet in height, all conspicuously banded
with horizontal terraces.

The vegetation of the trap area differs no less conspicuously from
that which is found on other formations, the distinction in the dry
season being so marked that, especially when taken in connection with
the form of the surface, it enables hills and ranges of trap to be dis-
tinguished at a distance from those composed of other rocks. The
peculiarity consists in the prevalence of long grass and the paucity of
large trees,! and in the circumstance that almost all bushes and trees,
except in the damp districts near the sea, are deciduous. The result is
that the whole country presents, except where it is cultivated, a uniform
straw coloured surface, with but few spots of green to break the monotony
during the cold season, from November till March, whilst from March,
when the grass is burnt, until the commencement of the rains in June,
the black soil, black rocks, and blackened tree stems present a most re-
markable aspect of desolation. During the rainy season, however, the
country is covered with verdure, and in many parts it is very beautiful,
the contrast afforded by the black rocks only serving to bring into relief
the bright green tints of the foliage.

Throughout the trap area the prevailing rock is some form of dolerite
or basalt, but there is a large amount of variety in the characters
presented by different beds. Some are excessively compact, hard, and
homogeneous, the crystalline structure being so minute as to be de-
tected with difficulty (ananiesite), others are coarsely crystalline, and
these frequently contain olivine in considerable quantities, and one variety
is porphyritic, containing large tabular crystals of glassy felspar, white
or green in colour. Many of the basalts again are soft and earthy, evi-
dently in most cases, and probably in all, from partial decomposition.
The most striking peculiarity is, perhaps, the great prevalence of amyg-
daloid, in which the nodules, chiefly containing zeolite or agate, sometimes
form the principal part of the rock. ‘These nodules are very often coated
with glauconite (green earth), and the prevalence of this mineral is highly
characteristic. Almost throughout their range, the Deccan traps may be
recognised by the occurrence of the amygdaloidal basalts with green earth,
or of the porphyry with crystals of glassy felspar.

1The want of large trees is partly due to | for temporary cultivation of a rude kind, and
the wanton destruction to which the forests of | perhaps more than all, to the practice of

India have been exposed for ages through | annually burning the grass at the commence-
reckless cutting, to equally reckless clearing | ment of the hot season,
258 GEOLOGY OF INDIA-=DECCAN TRAP. (Chap. XI.

Exfoliating concretionary structure is common in the softer forms of
basalt, which have undergone some amount of decomposition, but it is
never seen in the hard compact beds. Frequently the hard unaltered
spheroidal cores of concentric nodules, which may easily bé mistaken for
rolled fragments, are to be found scattered over the surface of the bed,
from which they have weathered out. Columuar structure 1s less com-
mon, though it is occasionally seen, a fine example being shown in the
following woodcut. In some cases this structure has been observed in

 

Fig. 16.—Radiating basaltic columns in a dyke near Gijri, north-west of Maheswar, Narba lé
valley.

the compact basaltic flows; it is frequently seen in the lowest flow, a very
thick one, west of Hoshangdbdd, in the Narbadd valley, and in one of the
lower flows in Mdlwd, but the appearance is often confined to intrusive
dykes, as in the example illustrated. Trachytic rocks are extremely rare,
and have hitherto only been found in intrusive masses.

Beds of volcanic ash are common, so common indeed in places as to
form a very considerable proportion of the strata, and they appear to be
much more prevalent towards the upper part of the series.1 They often
differ but little in appearance from the basaltic lavas with which they are
interstratified, but, on close examination, their brecciated structure car
always be readily detected, and the blocks of scoria which they contain

1 Possibly due to the upper part of the series | lower beds on the Narbada, west of Baroda,
being chiefly preserved near the old volcanic | where remains of ancient volcanic cores also
foci. Ashes are found interstratified with the | occur.
COMPOSITION,

Chap. XL] 259
generally weather out on exposed surfaces and remain in relief, precisely
as on old volcanic cones. Magnificent examples are to be seen on most
of the higher portions of the Sahyddri or Western Ghdts and on the
high peaks around Poona, formerly used as hill forts; well marked
instances occur also in Bombay and Salsette.! Very frequently a thin bed
of ash intervenes between two basaltic flows. Occasionally pumice is
found in the ash beds, the interstices being, however, all filled up by the
same process as that by which vesicular lava has been converted into
amygdaloid. Here and there, throughout the traps, beds of red bole
occur; they are usually only a foot or two thick, but occasionally more.
Sometimes the bole contains scoriz, and in this case it frequently covers
the upper surface of a basaltic flow, into which it appears to pass. In
some instances the bole is so uniformly stratified that it has the appearance
of having been deposited from water.’

In a few instances bands of very homogeneous structure and of a pale
lilac colour, formed of an apparently argillaceous rock resembling bole
in texture and so perfectly laminated as to exactly simulate shale, have
been found interstratified with the basalts. This is especially the case
at a large hill called Pdwagarh, 2 000 feet high, near Baroda, and similar
beds are said to occur in Kathidwdr; they have also bven noticed east of
Surat. The occasional occurrence of glassy felspar crystals in these beds
and the circumstance that some of the harder basalts at times weather on
their exposed edges into a somewhat similar soft lilac rock, render it
possible that these shaly strata result from the alteration of trap. At
the same time it is far from improbable that some of them may be con-
solidated volcanic mud, composed of fine lapilli washed down and deposited
by water.

No crystallised pyroxene has been observed, except locally in some
of the ash beds, and the only felspar which occurs in distinct crystals
appears to be the form of orthoclase (gassy felspar) which is found in
the porphyritic rock already mentioned. Olivine and magnetite are com-
mon, the former occurring as translucent yellowish grains, the latter in
minute crystals, too small, as a rule, to be recognised by the naked eye

1 Amongst the best examples are the rocks
in which the Keneri caves of Salsette are

2 Sir C. Lyell has shown that bands of red
clay interstratified with the lavas of Etna

cut ; some beds on the Kamatki ghat between
Poonaand Mahdbaleshwar ; and a conspicuous
bed at the lower gateway of the fortress of
Singarh near Poona. Ash-breccias also occur
in Bombay Island at Flag-staff hill and Rai
hill, Parel, and in the neighbourhood of Sion
fort. It must not be supposed from these
examples that the rock is rare. It is found
almost throughout the trap country, but it is
much less common towards the base of the
traps.

have been formed from the crust of the lower
lava flow, decomposed into clay and then
baked and reddened by the heat of the over-
lying flow, or where “volcanic sand has been
showered down from above and washed over
the older lavas by torrents and floods ;" Phil
Trans., 1858, p. 711. Similar beds appear to
be characté ristic of subaerial lava flows ; Judd,
Quart. Four. Geol, Soc., XXX, 227, (1874).
260 GEOLCGY CF INDIA—DECCAN TRAP. (Chap. XL

but easily detected, if abundant, by the effect of the rock upon the
magnetic needle. Magnetic iron sand derived from the traps is fre-
quently found in the streams which traverse the rocks. With the tabular
felspar crystals small scales of red mica are found.

Secondary minerals of various kinds, which kave been formed since the
consolidation of the volcanic strata, are found in the greatest abundance
in some of the flows, especially in the amygdaloidal, and in some of the
more earthy and decomposed traps. These minerals uot only form the
nodules of the amygdaloid, but they are found lining cracks and hollows,
the finest crystals being always in geodes or cavities, some of which are
as much as two or three feet across, and even larger hollows lined with
crystals are said to have been found. The commonest minerals are
quartz (either crystalline or in the form of agate, bloodstone, jasper, etc.,)
and stilbite, next in abundance are apophyllite, heulandite, scolecite
(poonahlite’, laumonite and calcite; thomsonite, epistilbite, prehnite and
chabasite also occur, but they are rare. The great prevalence of glauconite
or green earth has already been noticed.

‘The crystalline quartz is occasionally, though rarely, amethystine ; it
but seldom occurs in crystals which exceed an inch in diameter, and the
larger crystals are not often transparent. The form known as trihedral
quartz, in which the terminal pyramid of each quartz crystal consists of
three planes instead of six, or in which three planes are very much more
developed than the other three, is of common occurrence. The agates
occur chiefly in geodes or nodules, large and small; many are finely
banded, and, after being coloured by heating, are cut into ornaments.’
Jasper and heliotrope or bloodstone occur chiefly in flat plates, which
appear to have been formed in cracks, and agate is sometimes met with
of apparently similar origin. Stilbite is very common, though less so
than quartz; one magnificent variety consists of large orange or salmon
coloured crystals, often two or three inches in length, usually compound or
in sheaf like aggregations, but occasionally in large flat prisms terminated
by a four sided pyramid. Apophyllite is the finest of all the Deccan trap
minerals. It generally occurs in four sided prisms with terminal planes,
a form which closely resembles the cubical crystals of the isometric sys-
tem, the double pyramid, with replacements of the secondary prismatic
faces and terminal planes, so characteristic of this mineral in other locali-
ties, being chiefly typical of small crystals in the Deccan traps. The

1 Two other mineral species besides poonah- | Your. Sci., 2nd series, XL 1:0, (1865) is stilbite,
lite have been described from the Deccan traps. | coloured in the same manner.

One of these is hislopite, Haughton, Phil. Mag., 2 Most of the stones cut fur ornamen's are
qth series, XVII, 16, (1859), which appears | e'ther procured from rivers or from the ter-

to be calcite coloured by glauconite (green | tiary grave!s derived from the denudation of
earth) and the other, syhedrite, Shephard, Am. | the traps.
Chap. XI] HORIZONTALITY OF THE TRAPS. 261

colour of the Deccan apophyllite is usually white, more rarely pink of
green, some crystals are periectly transparent, and one of the most mag-
nificent associations of minerals to be found anywhere is seen when, as
occasionally happens, perfectly clear vitreous crystals of apophyllite, of
large size, are inserted on a mass of orange stilbite. Some apophyllite
crystals are as much as three or four inches across. The other minerals
are less deserving of notice, but very beautiful long acicular crystals of
scolecite with exquisitely formed pyramidal terminations are of occasional
occurrence, and fine crystals of white heulandite are not unfrequent. The
glauconite is usually amorphous, but occasionally forms an aggregate of
crystalline scales, and a massive mineral, which, if not green earth, is
closely akin both in appearance and compusition, occasionally occupies
small cavities completely.

One of the most remarkable characters of the Deccan traps is their
persistent flatness or near approach to horizontality throughout the greater
portion of their area. This is conspicuous throughout the Sahyddri range,
over the whole of the Bombay Deccan, from Khdndesh to Belg4um and
Sholdpur, throughout southern Berar and the north-western portion of
the Haiderdabdd territory, in many parts of the Sdtpura range between the
Narbadd and Té4pti, and on the Mdlwd plateau north of the Narbada.
Where exceptions occur, as in the western Sdtpura and Rajpipla hills
and along the coast near Bombay, the disturbance is shown to be of later
date from its affecting contemporaneous or newer beds of sedimentary
origin. ‘The only departure from absolute horizontality to be seen in the
lava flows of the Deccan is frequently no more than may be due to
the lenticular form of the beds, but usually there is a very low dip
discernible, seldom exceeding 1°, and fairly constant over large areas.
This circumstance tends to show that even this small amount of inclina-
tion may be due to disturbance, because if the dips represented the origi-
nal angle at which the lava flows were consolidated, they would be found
to radiate from the original volcanic vents. Nothing of the kind has,
however, been traced.

The separate lava flows are, as a rule, of no great thickness. The
average in the two sections of the Bhor and Thal Ghats, measured on the
railway lines, is apparently 64 and 87 feet respectively, but really less,
because the distinction between the flows can in most cases only be re-
cognised by lithological characters, and where, as must frequently be the
case, two or more beds of similar appearance and cumposition occur toge-
ther, they must often be confounded and measured as one. Many of the
more amygdaloidal beds appear to be made up of several smaller flows from
six to ten feet thick, distinguished by being highly amygdaloidal above,
less so in the middle, and traversed towards the base by long cylindrical

T2
262 GEOLOGY OF INDIA—DECCAN TRAP.

[Chap. XL

vettical pipes filled with zeolite.’ But even supposing that these appa-
rent distinctions are accidental, some well marked crystalline flows in each
section do not exceed 15 feet in thickness.

Hitherto only the igneous portion of the Deccan series has been de-
scribed, but volcanic rocks, although they form the great mass of the for-
mation, do not compose it exclusively, for sedimentary bands, frequently
fossiliferous, have been found in several places interstratified with the
lava flows, and have become widely known and described as intertrap-
pean beds. There is also found in many places, at the base of the whole
series, a small group of limestones, sandstones and clays, known as the
Lametd group, from its occurrence at Lametd Ghat, on the Narbadd, near
Jabalpur.

The intertrappean beds have been found in two distinct portions of the
Deccan series, first close to the base, throughout the greater portion of the
enormously extensive circuit of the volcanic area, and, secondly, in the
highest portion of the traps, only known to occur close to the coast in
Bombay Island and the immediate neighbourhood.’ A rough classification
of the whole series is presented in the following sections :—

Approximate
thickness in feet.3

1. Upper traps, with numerous beds of volcanic ash and the

intertrappean sedimentary deposits of Bombay 1,500
2. Middle traps, ash beds numerous above but less frequent
- towards the base, no sedimentary beds known + 4,000
3. Lower traps. with intertrappeans of Nagpur, Narbadd
valley, etc., volcanic ash of rare occurrence or wanting . 500
4. Lametd or infratrappean group . . . . . 20 to 100

The whole thickness, as will be shown presently, is probably consider-
ably greater than 6,000 feet in the neighbourhood of Bombay, but the
rocks gradually thin out in other directions, At Bombay the upper limit of
the scries is not seen. It is highly probable that near Surat and Baroda
the trap may have been even thicker than near Bombay, but the uppet
portions have been greatly denuded, and it is extremely difficult here,
as in most other places, to estimate the thickness with any accuracy.
In Cutch the traps are about 2,500 feet thick, whilst in Sind they have

1 Bearing in mind that amygdaloidal basalt

must have been originally vesicular lava, and
that what are now nodules of quartz or zeolite
were originally air or steam bubbles, it is easy
to understand that the upper portion of a lava
flow, having been more vesicular originally
than the lower portion, would be characterised
by a prevalence of amygdaloid. The verti-
cal tubes must also have been originally filled
with air or vapour, perhaps expelled from the

underlying stratum by the heated mass flow-
ing over it. :

? The reasons for considering the Bombay
traps higher in the series than the others will
be explained subsequently.

5 The thickness given is little more than a
guess, except in the case of the lower traps
and Lametds. The other figures are minimum
estimates of the vertical extent of the series,
where fairly developed.
Chap. XI] LAMETS GROUP, 263

dwindled down to two bands at different horizons, each less than roo feet
thick. Throughout the greater portion of their area, no higher beds,
except laterite or post-tertiary deposits, are found resting upon them, and
it is impossible to form any accurate estimate of their original develop-
ment. In the extreme south of the volcanic area, near Belg4um, their
thickness has been estimated by Mr. Foote to be 2,000 to 2,500 feet,
On the plateau of Amarkantak, at the eastern extremity of their main
area, they are about 500 feet thick, but farther east in the outlier on the
Mdin Pat in Sargija, not more than 300 to 4oo, whilst to the south-east
near Rdj4mahendri they are represented by a thin outlier, in which from
100 to 200 feet of basalt may be exposed,

Before proceedirg further it will be necessary to give a fuller descrip-
tion of the sedimentary formations, and in accordance with the system
adopted throughout this work, the Lametd gruup as the lowest will first
receive attention.! Formerly this group was supposed to be a representa-
tive of the Mahddeva group of the Gondwdna system, but further examina-
tion has shown that the Mahddevas are much more ancient, and that the
Lametd beds are so closely asscciated with the lowest trap that they must
be considered as part of the same series. The origin of the name has
already been mentioned, and it has been stated that the group consists of
limestones, sandstones and clays. The limestones are the most charac-
teristic and persistent beds, they frequently occur alone, and they form the
upper portion of the group when other beds are associated with them.
Occasionally the limestone is pure, but it is commonly full of sand and
small pebbles, so as to form a calcareous grit rather than a limestone, and
as arule it contains an abundance of masses, sometimes irregular, some-
times more or less lenticular in form, of segregated chert. Some of the small
pebbles frequently consist of red jasper, the occurrence of which is very
characteristic. This gritty limestone, with its included chert nodules, is
found over a very extensive tract of country in the Central Provinces, and
appears to be rarely absent throughout any large area in which the base of
the traps is exposed.

The bed which, after the limestone, is most commonly found in the
Lametd group, is a rather fine porous earthy sandstone, usually of a
greenish colour. The clays are red or green, and are very frequently
sandy or marly; sometimes they contain nodular carbonate of lime.
They are of local occurrence and appear but rarely to extend over any
considerable area. All these beds pass into each other; the limestone
is not unfrequently merely the sandstone cemented by carbonate of lime,

1 For details, see Quart. Four, Geol. Soc., | VI, 216, (1869); IX, 315, (1872); XII, 87,
XVI, 154, (1860); Memoirs, II, 196, (1860); | (1877) ; Records, V, 88, 115, (1872).
264 GEOLUGY OF INDIA—DECCAN TRAP. [ Chap. XL

the marls are an argillaceous form of the limestone, and, except where the
characteristic grilty limestone is the sole representative of the formation,
there is, as a rule, a frequent change of character in the beds, beth hori-
zontally and vertically. his is usually the case where the thickness
exceeds 20 or 30 feet, but where the group is only represented by a thin
band, either the gritty limestone or the earthy greenish sandstone is com-
monly found alone.

The Lametd group is quite unconformable to all the various older
formations upcn which it rests, from the metamorphics to the Jabalpur
group. As arule, the lowest flows of trap are conformable to the in-
fratrappean beds, but in some instances distinct unconformity has been
detected, especially in one case near Jabalpur,' and it is highly pro-
bable that closer examination would show that such cases are common,
and that in many localities where Lametds are wanting their absence is
due to denudation in pre-trappean times. At the same time the denuda-
tion appears to have been local, not general, patches occurring here and
there, whilst in the intervals between them the trap rests upon a forma-
tion older than Lametd, but at such an elevation as to show that the
absence of the intratrappean bed is not due to the ground having been
above the water in which the Lametds were deposited. It is impossible
that the Lametds cin ever have been co-extensive with the base of the
trap, because the surface on which the latter rests is extremely uneven,
and many portions of it must have been above the level at which the infra-
trappean beds were deposited. It will, however, be necessary to recur
to this subject, when discussing the relations of the trap series as a whole
to the older formations.

It is unnicessary to give a list of localities at which the Lametd group
has been observed. It is principally developed in the Central Provinces,
around Nagpur, Jabalpur, etc. It has not been found in the southern
Marathd country, but elsewhere along the boundary of the volcanic area
from the Godavari valley to Bhopal and Indore, it is rarely absent over
any considerable area. As a rule, owing to its small vertical development,
it only covers small portions of the surface, and it usually forms a ‘narrow
fringe to the trap country. In the western Narbada valley it has been
recognised and described as lying unconformably on the cretaccous Bagh
beds.?

The Lameta group is, as a rule, singularly unfossiliferous, the princi-
pal fossils which have been found in it, consisting of some bones of a large
Dinosaurian reptile, 7tanosaurus indicus, allied to Pelorosaurus of the
wealden and Cetvosaurus of the Bath oolite. These fossils occur near
Jabalpur, and similar bones, together with coprolites and some chelonian

1 Records, V, 115, (1872). 3 Lydexker, Records, X, 38, (1877).
2 Bose, Memoirs, XXI, 46, (1884).
Chap. X1.] FOSSILS OF THE LAMETA GROUP. 265

remains, were found at Pisdura about eight miles north of Warord in the
Ch4nd4 district.1_ In the last named locality some of the characteristic
fresh water mollusca of the intertrappean beds, suchas Physa prinsepii, are
associated with the bones, and the same shells have also been found in beds
at the base of the trap in one or two other localities; for instance, a Palu-
dina, apparently identical with P. deccanensis, an intertrappean fossil, was
found by Mr, Hislop at Ndgpur,? Melanta and Corbicula have been met
with in infratrappean beds near Ellichpur in Berar,’ and Physa prinsepid in
a similar position at Tedihal, 15 miles north-north-east of Kalddgi in the
southern Maratha country. But it is by no means clear, in those loca-
lities, where fresh water shells are found in beds beneath the trap,
with the exception of Ndgpur, that an intertrappean bed has not over-
lapped the edge of the underlying lava flow, so as to rest upon an older
rock, which may be either Lametd or any other more ancient formation
and in the particular case of Pisdura, where all the fossils are found
scattered on the surface of a field consisting of red Lametd clay, there is
always a possibility that Physa prinsepit and similar fossils may have come
from some small unnoticed intertrappean band, concealed beneath the deep
surface soil, At the same time it is by no means improbable that the
Physa and other shells are really derived at Pisdura from the Lameta
beds, and that this group consequently is not much older than the volcanic
beds which overlie it.

The only other noteworthy occurrence of fossils in the Lametd group
is that of some fish remains at Dongargaon, six miles east by south, and
Dhamni, nine miles east by north, of Warord.§ The species have not been
described; one of the fish found was considered by Sir P. Egerton allied
to the Sphyvrenodus (a cycloid acanthopterygian) of the London clay,
but according to Mr. Smith Woodward, only differs from Belonostomus,
an upper cretaceous genus, in the prominence of the vertical foldings
of the teeth.

Leaving the question of the mode of, origin of the Lametd group
to be discussed hereafter, and deferring for the moment the description
of some beds with marine fossils found at the base of the traps near Raja-
mahendri, the next group which requires notice is that comprising the
fresh water beds interstratified with the lower traps in many parts of
India, and especially in parts of the Central Provinces, northern Haiderdbdd,
Berdr, and the states north of the Narbad4 valley. Throughout these tracts
cf country, and beyond them almost throughout the great trap area, there

' Quart. Four. Geol. Soc., XVI, 163, (1860) ; 4 Memoirs, XII, 193 (1876).
Memoirs, X11, 88, (1877). 5 Quart. Four. Geol. Soc, XVI, 162,
2 Quart. Four. Geol. Soc., XVI, 167, (1260). | ( 860°.
3 Memoirs, V1, 283, (1869). ® Records, XXIII, 24, (* 890).
266 GEOLOGY OF INDIA—DECCAN TRAP. [Chap. XI.

are found heré and there, near the base of the volcanic formations,
and in no case so far as has hitherto been recorded, at a greater height
than from three to five hundred feet above the base, thin bands of chert,
limestone, shale or clay, often abounding in fossils of fresh water or
terrestrial origin.

Perhaps the most common form of the intertrappean bands, or that
which is most conspicuous, is a compact, blackish, cherty rock, a kind of
lydian stone It is clear that this rock has been originally a silt, and has
been hardened, either by the outpouring of igneous rock over it or by
chemical infiltration, the former being the more probable, because it very
frequently happens that the upper portion of the bed only is cherty, the
lower portion being a soft earthy shale. Other forms of intertrappean
bands are a dark or pale grey limestone, often earthy and impure, but
rarely gritty, like the characteristic Lametd bed. Not unfrequently the
sedimentary bed is composed of volcanic detritus, whether removed by
denudation from solid basalt, or consisting merely of the lcose products
of eruptions, such as lapilli, it is difficult to say. Red and green clays or
bole are also found, often associated with other intertrappean rocks.

As a rule, the sedimentaty beds interstratified with the lava flows
are distinguished from those underlying the whole volcanic series by
the absence of pebbles and sand, but occasicnally, though rarely, sandy
and even pebbly beds are found at some distance above the base of the
trap. In the south Mardthd country most of the intertrappean beds are
sandstones and conglomerates. One peculiar detrital form of intertrap-
pean accumulation has hitherto only been described from the country
north of the Narbadd and south of Chhota Udaipur on the banks of the
Karo, a tributary of the Hiran river.) The lower beds cf the trap series
here consist of conglomerates, sandstones, and sandy grits, sometimes
resting on a stratum of basalt, but occasionally on the Bagh cretaceous
beds, which underlie the volcanic formations. Occasionally the sandstone
or conglomerate appears to be chiefly composed of detritus derived from
the metamorphic rocks, but volcanic fragments, usually in the form of
rolled pebbles of basalt, can always be found by search, and in many
parts the bed becomes a mass of rolled volcanic fragments, often mixed
with unrolled scoria. At times, indeed, the rock is a conglomeratic ash,
in which rolled fragments of metamorphic rocks and of basalt occur to-
gether Hornblende and pyroxene crystals have been found in these
conglomeratic ashy beds, which are in some places as much as 200 feet
thick. In some instances the conglomerates appear to have accumulated
in hollows, like river beds, but in any case the abundance of rolled pebbles
and boulders of trap is important as a proof that denudation took place in
the interval between successive lava flows.

V Memoirs, V1, 327, (1866).
Chap. XI] INTERTRACPEAN BEDS. 267

With the exception of the detrital accumulations which have just been
mentioned, the intertrappean bands rarely exceed a few feet, from three to
about twenty, in thickness, and they frequently do not exceed half a foot.
In many places two or more sedimentary beds occur at different levels in
the same section, and the different bands are in some cases dissimilar in
mineral character. Thus, at Mekalgandi! Ghat in the Sichel hills, south
of the Pen Gangd river, on the old road from Nagpur to Haiderdbdd, a
locality famous as being one of the first at which the intertrappean fossils
were detected by Malcolmson, the following beds are observed in sec-
tion :—

1. Trap.

2- Cherty bed containing Unio, Cypris, etc.

3. Trap.

4. Limestone containing Cypris ana fragments of small mollusca.
5. Trap.

6. Calcareous grit, containing broken shells (Lameta),

q. Metamerphic rocks,

A single intertrappean bed can but rarely be traced for more than three
or four miles without interruption ; it then usually dies out. At the same
time it is rare to go over any large tract near the base of the traps without
finding some sedimentary bands interstratified, and occasionally one is
found to be much more extensive than usual. Thus, an instance is re-
corded by Mr. J. G. Medlicott? in Sohdgpur, east of Jabalpur, in which an
intertrappean bed was traced for nearly 25 miles.

It would take up too much space to enumerate all the localities at
which the lower sedimentary intertrappean beds have been observed.
They have been noticed in several places in the southern Maratha
country ; they are commonly found near the base of the trap flows almost
throughout the great and irregular line of boundary extending from the
Goddvari to Rajputdna, and they oceur even in small outliers, for instance,
at Mdin Pat in Sargtija; they have been detected by Mr. Rogers ® to the
westward at Dohad, about 75 miles north-east of Baroda, and still farther
west in Cutch, by Mr. Fedden* of the Geological Survey.

The abundance of fresh water and terrestrial animals and plants in
the intertrappean beds has been the principal reason for the comparatively
large amount of notice which these thin bands of rock have attracted,
The mollusca are. very abundant and are occasionally exquisitely pre-
served in the cherty layers, the commonest species being forms of PAy'sa and
Lymuea, whilst Unio, although abundant locally, is of comparatively rare
occurrence, Paludina, Valvata, and Melania are far from uncommon.

' Mucklegundy pass of Malcolmson, 3 Quart. Four. Geol. Soc,, XXVI, 122, (1870).
2 Memcirs, II, 201, (1859). 4 Menoirs, 1X, 58, 240, (1872).
268 GEOLOGY OF INDIA DECCAN TRAP. [Chap. XI.

Land shells are very seldom found, but they have been detected! in one
case at least. Entomostracous crustaceans are very nearly as common as
mollusca, all hitherto found belonging to the genus Cypris. The other
remains of animals hitherto detected have consisted of insects, fishes, and
reptiles, all of which are fragmentary. Plant remains abound, but leaves
are rare, seeds and fragments of wood being more common and the most
abundant vegetable fossils are the seed vessels of Characez, of which one
species has been described under the name of Chara malcolmsont.

The plants have not been described, with the exception of the Chara.
Those collected near Nagpur are said by Mr. Hislop to comprise about
fifty species of fruits and seeds, twelve of leaves, and five kinds of woods,
the only forms mentioned are endogens and angiospermous exogens. The
relations of the fossils will be discussed in the sequel, together with
the fauna of the other intertrappean deposits.

The whole of the mollusca and crustacea are fresh water forms; no
marine species have been detected associated with them, except in the
beds near Rdj4mahendri, of which a description will’ be given in the
next paragraph. The insects and plants, with the exception of Chara
afresh water form, are of terrestrial origin. The general prevalence of
the pulmoniferous mollusca Physa and Lymnea appears to indicate that the
water was shallow, as these forms live partly at the surface. Cypris, too,
is commonly found in shallow marshes.

The outcrops of trap near Rdjdmahendri are so remote from any
other exposure of the Deccan volcanic series, being about 210 miles
distant from the nearest point of the great Deccan area north-west
of Sironchd, that some doubt would remain as to the identification, despite
tLe similarity of mineral character, had not some of the typical fresh water
fossils of the Deccan intertrappean beds been discovered in the Rdja-
mahendri area. The Rdjdmahendri outcrops occur on both banks of the
Godavari,’ and consist of an interrupted narrow band of volcanic rocks,
chiefly earthy ‘dolerite and amygdaloid of the usual character, extending
altogether for about 35 miles from east- north-east to west-south-west.
Traps are seen at Kdteruon the left bank of the Goddvari just north of
Rajdmahendri itself, and extend rather more than ten miles to the east-
north-east, resting upon metamorphic rocks whenever lower beds are

1 Memoirs., Il, 213, (1859); several forms | and collections of the fossils were made by
were referred to the terrestrial genus Acha- | Lieutenant Stoddart and Sir W. Elliot, and
tina. Some similar fossils from a French | described by Mr. Hislop, Quart Four. Geol:
deposit had been placed in the same genus, but | Soc., XVI, 161, 176, (1860). The infratrappean
it appears more probable that the Indian shells | band was first noticed by Dr. King, Records,
are of fresh water origin and belong to Lymnea | VII, 159, (1874). See also Memoirs, XVI,

or to some allied type. 324, (1880); Quart. Four. Geol, Soc. X, art
2 The intertrappean beds were discovered (1854). ) i F i » X, 471,

originally by General Cullen and Dr. Benza,
Chap. XL] RAJAMAHENDRI OUTCROPS. 269

seen. On the right bank the volcanic rocks appear in two areas, divided

by small alluvial valley: the larger extends for about ten miles to the
westward from Pungadi, 7 miles west of Rajamahendri, and the smaller
occurs a few miles still farther west. In these outcrops the beds of
the volcanic series rest upon the Rdjmahdl rocks of the Ellore region. In
both cases the strata overlying the trap are Cuddalore sandstones
and all the beds alike have a low dip to south or south-east. The whole
thickness of the volcanic series at this locality, as already mentioned,
nowhere appears to exceed about 200 feet, and in places it is not more
than 100,

At the base of the traps, and intervening between the basalt flows and
the underlying jurassic sandstone, about 50 feet of sandstone, white,
yellowish, or greenish in colour, are exposed. near the village of Didkur, 12
miles west of Raj4mahendri. The upper portion is calcareous, and on the
top there is a band, about six inches to two feet thick, of sandy limestone
abounding in marine fossils, the most abundant of which is a 7urrdtella,
apparently identical with 7. d/spassa of the cretaceous Ariyaltr group. If
not identical, the two species are very closely allied. A Nausz/us, about
fifteen Gasteropoda, and eleven Lamellibranchiata accompany the Turri-
tella, but not a single species, except ZTurritella dispassa, has been
recognised as identical either with the cretaceous beds of southern India
or with the eocene fossils of the nummulitic group. The collections have
not, however, been sufficiently compared to enable the species to be de-
termined with certainty. Only one single species, too, Cardita variabilis,
has been recognised as occurring also in the overlying intertrappean bed.
Although the whole facies is tertiary, there 1s a remarkable absence of
characteristic genera,! and the chief distinction from the cretaceous fauna
of the upper beds in Southern India is simply the want of any marked
cretaceous form ‘The fauna is distincly marine.

It is difficult to say whether this bed should be referred to the Lametd
group or not. The mineral character is similar, but all known Lametd
outcrops are so distant that the identification is somewhat doubtful.
The distinctions between the fossils of the B4gh beds and those of the
infratrappeans of Dédktr and Pungadi appear too great to te attributed
solely to the existence of a land barrier between the two areas; it is
difficult to suppose that the two tormations can be of the same geolo-
gical age, and the Bagh beds are probably more ancient than the
Pungadi infratrappeans, ‘The balance of evidence is rather in favour of
referring the latter to cretaceous times than to tertiary, and they may be

considered of intermediate age, as will be shown to be probably the case with
the Lametas.

' Amongst the genera identified are Rostel-
laria several forms of Musicida, a Volutilithes
near the tertiary V. forulosa,Natica, Tw ritella,

Dentalium, Cytherea or ailied genera (three

sp.), Cardita (four sp.), Corbis, Pectunculus,
Cucullea and Ostrea.
270 GEOLOGY OF INDIA—DECCAN TRAP. [Ohap. XL

Upon the fossiliferous limestone described in the last paragraphs ‘a flow
of basalt is superposed, varying in thickness from about 30 to about
too feet. There is an appearance of slight unconformity where the volcanic
rock rests upon the sedimentary bed, the surface of the latter being slightly
uneven, as if denuded, and the upper fossiliferous infratrappean zone is
occasionally wanting. The variation in thickness of the basalt stratum
may be due to its having been poured out upon an uneven surface, but it
is not quite clear whether this unevenness was due to disturbance of the
sedimentary beds before the outburst of the traps. That the denudation
of the underlying formations can have been only partial is shown by the
fact that they may be traced between three and four miles, the upper
portion alone being locally absent,

On the left bank of the Godavari, near Rdjdmahendri itself, the in.
fratrappean band has not been observed. The thickness of the lower
flow of basalt cannot be clearly ascertained, but it is not less than 4o
or 50 feet, and is probably more. Above this lower flow on both banks of
the Godavari there is found a sedimentary band, twelve to fourteen feet
thick at Kdteru, where it only extends for about half a mile, and about two
to four feet thick in the Pungadi direction, where it has been traced for
about ten miles. The intertrappean bed consists of limestone and marl,
portions of which abound in fossils. Numerous quarries, which have been
opened near both Pungadi and Kéateru, have afforded good opportunities
for obtaining fossils, which are difficult to extract from the argillaceous
limestone when it is first quarried but weather out on exposure. About
30 or 40 feet above the fossiliferous limestone of Kdteru, another sedi-
mentary bed, consisting of yellow calcareous shale, is seen in one place,
It is very thin, and no fossils have been found in it.

The most marked feature of this fauna is its distinctly estuarine char-
acter! Tympanotonus, Pirenella, Certthidea and Potamides are all brackish
water forms. A/ydrodéa is an estuarine genus, and the fossil called Hemz-
toma closely resembles a species of Acmeza found living in creeks in the
deltas of Indian rivers. The shell described as Cerithium multiforme ap-
pears tobe a Zympanotonus or Pirenella,; C. leithi has the characteristic form
and sculpture of a Cer¢thidea, and C. stoddardi is, at least, as much allied to
Potamides as to Cerithium proper. Some of the shells referred to Cytherea
agree best with the typical forms of the genus (C. meretrix), many species
of which abound in backwaters and at the mouths of rivers,and Mr. Hislop
has remarked the similarity between Corbula ofdhaméand a Brazilian specics
belonging to the estuarine genus Azara. There is a complete absence of
pelagic shells such as the Cephalopoda, no Echinodermata or corals are
found, and, above all, four species Physa prinsepil, Lymnea subulata,
Paludina normatlis and Corbicula ingens are characteristically fresh water

* Fora list of the species that have been obtained see Memoirs, XVI, 233, (1880).
Chap. XT] AGE OF THE RAJA MAIIENDRI BEDS. 27t

forms; the first three of these are of comparatively rare occurrence,
but the Cordzcula is common, and the last named may perhaps have lived
in brackish water, as its near ally Cyrena-does at the present day, whilst
the purely fresh water shells were washed down by rivers, this view being
quite in accordance with the theory that the intercrappean beds of
R4jdmahendri were deposited in brackish water, which was supplied
with fresh water by streams, but was also in communication with the
sea.

The mollusca, however, cannot be considered as very characteristic of
age. They were compared by Mr. Hislop with the nummulitic fauna of
western India, but, as he points out, no forms appear to be identical, and
although Natica dolium, Turritella affinis and an unnamed Certthium
found in the tertiaries of Sind and Cutch resemble ™. stoddardi, T. pre-
longa and C. stoddardt, the intertrappean forms are more closely allied to
the cretaceous NV. (Afammilla) carnatica, T, elicita and Cerithtum vagans
than to the eocene species mentioned, while other forms might easily be
shown to be affined to those occurring in the cretaceous rocks of Southern
India. Inthe case of Turritella prelonga and T. elicita the affinity is
very great. The shell called Vicarya fusiformts appears not to be really
congeneric with V. verneuzll7, the type of the genus,? and the latter has
now been found to be miocene, not eocene. On the whole, it may be safely
asserted that no tertiary alliances of any value have been detected amongst
the intertrappean RAjAmahendri fossils, and that their relations are rather
with the upper cretaceous rocks of Southern India, although the connection
is not strong.

In the islands of Bombay and Salsette, and probably farther north on
the same line of coast, the traps have an inclination of from 5° to 10°
to the westward. The islands are separated from each other and from the
mainland to the north by tidal creeks and alluvial flats, whilst the expanse
of water forming Bombay harbour lies between them and the mainland to
the eastward. In the islands of the harbour, and on the hills between
Thdna and Kaly4n north of the harbour, the same westwardly dip is dis-
played, but further to the eastward, from Kaly4n to the Sahyddri range,
the traps are horizontal.

About 2,000 feet of horizontal beds are exposed on the flanks of
Matherdn hill, and a still greater thickness farther to the east in the hills
near the Bhor Ghat and close to the Great Indian Peninsula Railway line
between Bombay and Poona, but it is impossible to say how far the lowest
strata, exposed at the base of the hills, are above the bottom of the series,
as no lower beds than the traps are seen. Owing to the numerous breaks in

1 When Mr. Hislop wrote, the South Indian | # This was pointed out by Mr. H. M. Jenkins,
cretaceous fossils had not been described. Quart. Four. Geol, Soc. KX, £8, (1864).
272 CEOLOGY OF INDIA—DECCAN TRAP. (Chap. XL

the section, it is difficult, without closer measurements than have hitherto
been made, to estimate the precise thickness of the rocks dipping to the
westward near Bombay, but taking the average dip at 5°, the whole thick-
ness would be nearly 7,000 feet, This is a minimum estimate, as the
average dip is probably higher and the thickness consequently greater.
From 1,200 to 1,500 feet of rock are exposed in Bombay island, so that it
is evident that the lowest beds seen on the island are higher in the series
than the highest flows seen .on the Sahyddri mountains to the eastward,
although some of the higher portions of the range are 4,000 feet above
the sea.

The intertrappeans of Bombay are entirely confined, so far as is known,
to these higher beds, no sedimentary rocks having hitherto been found
amongst the middle portions of the Deccan trap series and it is manifest
that the Bombay fresh water beds belong to a very different horizon from
that to which the intertrappeans of Ndgpur and the Narbadé valley must
be assigned. The most important bed is that which underlies the basalt
of Malabar hill and Worlee hill, forming the broken ridge along the
western or sea face of the island; this stratum is consequently imme-
diately beneath the highest lava flow known to occur anywhere through-
out the trap area, for the rocks, as already stated, dip to the west, and no
beds higher than those of Bombay have been discovered. It must, however,
not be forgotten that the coast north and south of Bombay has not hitherto
been examined with sufficient care to make it quite certain that no higher
beds occur,

This intertrappean bed on the east side of Malabar hill is more than one
hundred feet thick in places, and consists principally of soft grey, greyish
blue, brown, and brownish yellow earthy shales, with occasional harder
bands, some of which are black and carbonacecus. The greater portion
of the bed is evidently formed of volcanic detritus, whether lapilli washed
down by water, or sand produced by the disintegration of lava flows,
it is difficult to say, very possibly both may have contributed to the |
formation of the rock, At the top of the deposit the shale occasionally
becomes hardened and silicious, as if by the action of the overlying
basalt. The black carbonaceous shale is locally highly bituminous and
sometimes contains small layers of a coaly substance and fragments of
mineral resin. Impressions of vegetables abound, although they are tut
seldom well preserved, and remains of animals are common, the best known
being skeletons of small frogs and carapaces of Cyprides.

Besides this thick sedimentary band, several thinner beds have been
found at lower horizons amongst the lava flows and ash beds of Bombay
island. They are, however, very thin and, except one which is seen in
the quarries of Nowrot! hill south of Mazagaon, they are difficult to detect ;
indeed, the circun:stance of their occurrence has only become known through
Chap. X11

the careful scrutiny of local geologists, who, living in the town, could take
advantage of any excavations for buildings, tanks, roads, etc.. to examine
the strata exposed. According to Dr, Buist there are five or six sediment-
ary beds below the thick band of Malabar hill, but fossils have only been
found in that exposed at Nowroji hill, where Cyprides occur. All these
bands consist of shaly beds.!

The fossils found at Bombay are tolerably numerous, but hitherto only
the Vertebrata appear to have received more than a superficial notice. ‘The
remains of a fresh-water tortoise, Yyaraspis lec:ht (Testudo lerthi, Carter)
belonging to the Amydda, and of a trog, Rana pusilla? considered by
Dr. Stoliczka an Oxyglossus,* have been found, the latter in abundance,
while some bones of a larger frog have been obtained. The Archropoda are
represented by three species of Cyprzs, one of which, C. (cylindrica), is also
found in the intertrappean deposits of the Deccan; another species has
been called C. semtmarginata by Dr. Carter, the third is unnamed. C. semz-
margtnata is the most generally diffused, but the other forms also occur in
great numbers. Only fragments of insects have been found. Mollusca
are rare, and the few specimens hitherto procured have been in poor con-
dition, they have been referred to A/elania and Pupa, but with some doubt,
and none of the characteristic Deccan forms have been detected, The
plant remains comprise stems, leaves, seeds, and perhaps roots, but very
little has been determined, except that endogens and angiospermous exo-
gens are represented.

The life represented by the species named is clearly that of a shallow
marsh. The frogs occur in large numbers, and their bodies have evidently
been deposited near the spot where they died, as the whole skeleton
is found perfect. In some cases, as was noticed by Dr. Stoliczka, the
skeleton has been dragged along the surface of the shale in which it is
imbedded, and he suggests with great probability that this was done by
wind. The tortoise is a marsh or river form, the nearest living ally, accord-
ing to Dr. Gray,* being a genus found in fresh water in South America.

INTERTRAPPEANS OF BOMBAY ISLAND, 273

After the description of the various sedimentary formations intercalated

! For fuller description of these beds see Carter,
Four. Bo.. Br. Roy, As..Soc., 1V, 161, (1853), and
Geological Papers on Western India, p. 128;
Buist, Trans, Bo. Geogr. Soc., X, 195, (1852)

Wynne, Memoirs, V, 193,(1866); V1, 385, (1869).

It must not be forgotten that Dr. Carter’s views
as to the relations of the sedimentary beds
differ essentially from those stated in text,
with which all other observers agree.

2Owen, Quart. Four. Geol. Sve., IL), 224,
(1847).
3 Memoirs, VI, 387, (1867). Dr. Sroliczka

shows that the form agrees well with Oxyglos-
sus and with no other known existing genus.
At the same time, as some of the principal
characters by which genera of frogs are dis-
tinguished are not preserved in the skeleton,
the Bombay frog may have differed greatly
from recert Oxyglossi, From the species of
true Rana it is distinguished by the want of
vomerine teeth, the large head, and short
hinder limbs.

4 Ann, Mag. Nat. Hist , 4th series, VUI, 339,

(1871).
274 GEOLOGY OF INDIA—DECCAN TRAP. (Chap. XT,

with the traps or underlying them, the next point for consideration is the
mode of origin of the trap rocks themselves. Their volcanic character is
sufficiently proved by their composition. Precisely similar rocks occur
amongst the lavas poured out from recent volcanoes, whilst nothing of the
same kind has ever been known to be deposited from water. But the first
difficulty which arises, and it is one. of very great importance, is to account
for the persistent horizontality of the beds. Two observers certainly,
Jacquemont! and Adolph Schlagintweit,* have considered that the traps are
unstratified, but after the evidence already mentioned as to the differences
in mineral character between successive bands, the frequent occurrence of
vesicular structure on the upper surface of flows, the presence in abundance
of beds of volcanic ash, and the repeated interstratification in the same
localities of sedimentary layers, it is unnecessary to refute this view. A
much more common opinion, and one which has been supported by numer-
ous excellent geologists, from Newbold downwards, is that the Deccan
traps are of subaqueous origin, and it is necessary to show why this opinion
is untenable.

In all cases of subaqueous eruptions the ejected masses consist of
substances very similar to the lava, ashes, scoriz and lapilli of ordinary
subaerial volcanic outbursts, but these materials being thrown out into the
water are reduced by the sudden cooling to the condition of a fine
powder, which is dispersed and deposited in layers in the same manner
as ordinary detritus, so as to form what are known as stratified tuffs,
With these tuffs ordinary marine deposits are necessarily intercalated,
and both these and the tuffs are usually fossiliferous, the very destruction
of life in the waters of the sea, caused by the heat and gases which are
evolved during eruptions, encouraging the preservation of those portions
of the organism which are not liable to destruction from the temperature
of boiling water. or the process of decomposition. Now, the volcanic
ashes, already described as occurring in great abundance amongst the
higher beds of the Deccan traps, are not, as a rule, stratified in the
manner in which beds deposited from water would be. Although they
occur in strata, intercalated with basaltic lava flows, these ash beds them-
selves have no internal lamination, except in a few rare instances in
which they are chiefly composed of bole, and may have been formed in the
small pools of fresh water so common in volcanic areas. Above all, not a
trace of a marine organism has ever been found in any ash bed, or in any
rock intercalated with the traps, except in the intertrappean and infratrap-
pean formations of Rajdmahendri, where the lava has evidently been

' Voyage dans I'Inde,” 4°, Paris, 1848, IIT, | engaged in the Magnetic Survey of Indiz,”
504, etc. No. I, p. 6.—Reisen in Indien und Hochasien.
2 Report of the Proceedings of the Officers | Vol. I, p. 141, (1869).
   
   

Unio deceanensis, Sow X *

    
   

Lymnawa telankhediensis

var perexcuminete, Hisl .

Physa prinsepi var. elongata,

Hisl.

   

Vicarya fusiformis,
Hisl.

Turritelle prelonga,Hiel

 

Centhium stoddaidi, Hisl.

Cardiba varia bilis, Hisl.

Lithe graphed & Printed at. Geological Survey Office.

iINTERTRAPPEAN FOSSILS.
Chap. XI] SUBAERIAL ORIGIN. 275

poured out on the coast. It may be thought that the prevalence of
volcanic conditions would destroy all life in the sea, and thus the absence
of marine fossils in the traps may be explained, but even if this view were
conceded, and it is entirely opposed to all that is known of recent
submarine volcanic action, there must have been a great destruction of
lite at the commencement of the volcanic epoch, and some traces of the
animals destroyed should have been preserved,

The evidence afforded by the characters of the traps and the absence cf
marine fossils is, therefore, opposed to the hypothesis of a submarine
origin, and the relations of the lowest lava flows to the underlying rocks
are strongly antagonistic to the idea that the volcanic outbursts were
subaqueous. The surface of the older rocks upon which the traps rest is
in many parts extremely uneven, the basalt filling great valleys, some-
times as much as 1,000 feet in depth, whose form shows that they were
excavated by subaerial erosion. Admirable examples are seen between
Bhop4l and Hoshangadbdd, where the Deccan traps rest upon an extremely
uneven surface of Vindhyan rocks. It is true that this uneven surface
might have been formed above the sea and then depressed beneath the
water, but in that case we should expect to find aqueous deposits of con-
siderable thickness at the base of the volcanic cocks, as periods of depres-
sion are always favourable for the accumulation of sediment. It is pre-
cisely in this uneven ground that no deposits whatever are found at the
base of the traps and the general absence of any infratrappean deposit
has been noticed in the south Mardthd country, where also the surface
upon which the traps rest is very irregular.

Where the underlying formation consists of the cretaceous Bagh beds,
these are, as a rule, apparently conformable to the volcanic series, and it
might be thought that in this tract of country the traps were submarine.
But every here and there a spot is found where the cretaceous rocks are
wanting, and where the level of the infratrappean surface shows that
their absence is due to denudation. In some cases where the Bdgh beds
are not more than 30 or 40 feet thick, the denudation which has removed
them has only extended over a small area, and has scarcely affected the
harder rocks beneath, and from the small area, often only a few yards wide,
over which the cretaceous rocks have been removed, it is evident that
the denuding agent was subaerial erosion. It has also been ascertained
that the Bagh beds had been locally disturbed to a small extent, besides
having suffered from denudation, before the commencement of the
volcanic outbursts,

Lastly, the circumstance that, with the single exception of the estuarine
intertrappean band of Rdéjdmahendri, every fossiliferous sedimentary bed

1 Memoirs, V1, 240, 242, etc., (1867). 1 # Memoirs, VI, 212, 300, 313, ete., (1867).
U
276 GEOLOGY OF INDIA—DECCAN TRAP.

(Chap. XL

intercalated with the Deccan traps is unmistakably of fresh water origin
is a conclusive proof that all those lava flows which are associated with
such sedimentary beds are not submarine. We have thus not only a
complete absence of all proof of submarine origin,* but clear and
unmistakable evidence that the traps were in great part of subaerial
formation.

Another favourite idea with many writers, and especially with Mr. Hislop
and Dr Carter, has been that the lower traps were poured out ina vast, but
shallow, fresh water lake extending throughout the area over whieh the
‘intertrappean limestone formation’ extends.* ‘This hypothesis involves
the existence of a lake of enormous size, several hundreds of miles in length
and breadth, but shallow throughout. It appears more probable that the
lakes in which the Lametd group and the intertrappean beds were de-
posited were of moderate size, and that they were formed by unequal
elevation of different parts of the area, prior to the volcanic outbursts, or
by the obstruction of the drainage of the country by lava flows. The
lake or lakes in which the Lameté beds were formed may have been more
extensive, but it has already been shown that single sedimentary bands
intercalated in the traps, can rarely be traced for more than three or four
miles, and the character of the fauna, in the intertrappean formations,
both of Central India and of Bombay, is in favour of the animals of which
the remains are found having inhabited shallow marshes rather than deep

lalzes,

1It may appear to many geologists that an
unnecessary amount of space and argument
has been devoted to proving a very clear pro-
‘position, viz, that the Deccan traps are
subaerial. The reason for giving the arguments
at length is that a different view has been
expressed by many geologists. A reference
to the Quart. Four. Geol. Soc, XXX, 225,
(1874), will show that the arguments used by
Professor Judd to prove the subaerial origin of
the voicanic rocks in the west of Scotland and
north of Ireland, are precisely the same in
many cases as those mentioned above. These
views had been urged in the case of the Indian
Tocks (A/emoirs VI, 145) some years before the
publication of Professor Judd’s papers,

2 By both the writers named the intertrappean
beds of Bombay were supposed to be iden-
tical with those of Central India, and both
were under the impression that there was but
a solitary fresh water bed which was deposit-
ed before any volcanic outbursts took place,
which was then covered up by lava flows, and
finally separated from the underlying rocks

 

and broken up by a great sheet of intrusive
basalt injected beneath it.

The geologists named would doubtless have
modified their views had they been acquainted
with all the facts now ascertained with regard
to the Deccan traps and the associated sedi-
mentary beds. The conception of a great
sheet of intrusive basalt so injected between
two formations that it always overlies the
one and underlies the other, over an area of
thousands of square miles, is quite untenable.
It is a physical impossibility that an immense
dyke should be injected horizontally for hun-
dreds of miles instead of breaking through to
the surface. Moreover, the fact that successive
sedimentary beds, as in the case at Mekalgandi
Ghat, mentioned on page 267, are often of
different mineral compos.tion, and the very fre-
quent instances in which the upper surface of a
sedimentary band is altered, whilst the lower
is unchanged, prove that both lava flows and
sedimentary intertrappean beds were regu-
larly and successively formed, one above the
cther, as they now occur.
Chap. XT] ; MODE OF ERUPTION, 277

We are thus thrown back upon our original difficulty, the horizontality
of the Deccan traps. It has been shown that this is not due to a sub-
aqueous origin, whether marine or fresh water. At the same time the
phenomenon cannot be said to have been thoroughly explained, because
no such formation is known to be in process of accumulation at the
present day. Many such masses of horizontal stratified traps are, how-
ever, found in various parts of the world, and though it is impossible,
for want of recent examples, to demonstrate the circumstances which cause
their formation in place of volcanic cones, there is abundant evidence that
such traps were a common form of volcanic accumulation in past times,
and that similar stratified lava flows were not confined to any particular
epoch, although several instances are known of about the same geological
age as that attributed to the Deccan outbursts.

Assuming, therefore, as we are justified in doing, that the horizontal
dolerites of western and central India precisely resemble modern lavas
in everything except their horizontality and the extent of area which they
have covered, it remains to be seen what evidence there is of the sources
from which this enormous accumulation of molten materials was poured
out. The original cones and craters, if any ever existed, must have been
the first portion of the volcanic area to suffer from denudation, and it is
easy to conceive that subaerial erosion, from eocene times to the present,
would have more than sufficed to remove every trace of the loose material
of which volcanoes are chiefly composed, Still, it is surprising that the
inclined beds forming the slopes of a volcanic cone should, in no single case
which has been recognised, have been preserved by being encased in
subsequent outbursts of harder materials. Possibly the tendency of great
lava streams to sweep away all loose volcanic materials may suffice, in
those cases in which large quantities of lava are poured out, to prevent
volcanic cones from forming.

When, however, we look for other evidence of the neighbourhood of
igneous outbursts, we find dykes and irregular intrusions abundant in
some localities, rare or absent in others, whilst the presence of volcanic
ash throughout a large portion of the trap area has already been noticed.
The ash beds, especially when, as usually happens, they form a coarse
volcanic breccia, containing blocks several inches in diameter, cannot
have accumulated far from volcanic vents, although they may have been
transported to a much greater distance floating on the surface of molten
lava than they could have been ejected from the volcano.

A much closer examination of the Deccan area than has hitherto been
practicable will be requisite before the distribution of dykes and ash beds
can be considered as even approximately known, So far as the country
has hitherto been examined, both appear to prevail much more largely in
the region near the coast, froni Mahdbaleshwar to the neighbourhood of

U2
GEOLOGY OF INDIA—DECCAN TRAP.

[Chap. XI.

Baroda, than in other parts of the trap area. It is, of course, very often
difficult to recognise dykes amongst rocks of precisely similar mineral
character, much closer search being needed than is requisite in order to
detect volcanic intrusions amongst sedimentary formations. It is only
where dykes are large and numerous that attention is likely to be directed
to them.

There is one tract of country in which dykes are peculiarly large and
abundant. This. is in the RA4jpipla hills, north-west of Surat. In this
country, over a considerable area, very large, parallel, or nearly parallel,
basalt dykes are found, sometimes not more than two or three hundred yards
apart, the general direction being east by north to west by south. The
traps are much disturbed, and frequently dip at considerable angles.

To the southward of the T4pti, along the line of the Sahyddri, and its
neighbourhood, in western Khdndesh, the northern Konkan, and the
intervening small native states, a tract whose geology is unknown,
it is probable that dykes may continue numerous for a considerable
distance, as their number and size in the Konkan, north-east of Bombay,
are especially noticed by Mr. G. T. Clark,’ but intrusions are far from
abundant in the lava flows exposed in the higher country east of the Ghats.
At the same time the frequent occurrence of ash beds in the higher
traps around Poona and Mahdbaleshwar sufficiently attests the neighbour-
hood of the old volcanic vents.”

North of the Réjpipla hills and of the river Narbad4, and west of
Baroda, trap dykes are not so common as in the R&jpipla hills, but in-
trusive masses occur. One of these, forming Matapenai or Karali hill,
about 14 miles south-west of Chhota Udaipur, is a mass of grey trachyte
or trachy-dolerite, containing enormous masses of granite, evidently
derived from the metamorphic rocks through which the mass, when
molten, passed on its way tothe surface. The silicious character of the
intrusion in this, and some other cases, is perhaps due to the fusion of
quartzose metamorphic rock in the basic dolerite. Another trachytic core
was noticed near the village of Padwani, 18 miles east of Broach. The
occurrence of fragments derived from the metamorphic rocks in intrusive
dykes is by no means an uncommon occurrence.

It is only natural that far better evidence of volcanic foci is to be

' Quart Four. Geol. Soc. XXV, 164, (1869). | homogeneity of the accumulations in the
3 Mr. Clark has described what he regards as | neighbourhood of the foci, as compared with
a series of vents along the course of the Barwi) the more evenly bedded traps and ashes ata

278

river above Kalydn. They are said to be crater-
like hillocks, in whose neighbourhood the traps
lie in small streamlets, crossing and overlap-
ping one another. These supposed vents lie
along the region where the traps have under-
gone a maximum of denudation, believed by
Mr. Clark to be due to the less degree of

distance ; Records, XIII, 69, (1880). The only
reason for doubting those conclusions is the
peculiar form assigned to the hillocks; as a
rule, the material filling the neck of an old
volcano is harder than the surrounding stuff,
and would not forma hollow surrounded by
a ridge, but stand out as a solid mass.
Chap. XI.] FOCI OF ERUPTION. 279
found outside the trap area, or in the inliers of older sedimentary rocks,
than amongst the lava flows themselves, and it may therefore be useful
to point out very briefly the distribution of such intrusive masses so far
as the country isknown. Commencing to the north-west, no trap dykes
have been found in Sind, where, however, the deposits of older date than
eocene cover an exceedingly small area. In Cutch intrusive masses of
basalt and dykes of large size abound throughout the jurassic rocks, and
some of the former rise into hills of considerable size.1_ In Kathidwdr the
Girndr hill is said to be formed of the dioritic core of an ancient volcano,
and intrusions are abundant throughout this district. ‘Throughout the
northern edge of the trap country in Rdjput4na, Gwalior, and Bundelkhand,
dykes are rare or wanting, but they abound in some of the areas of older
rocks exposed in the Narbadd valley, and they are especially common in
the Gondwd4na tract, south of the river, in the neighbourhood of the
Mahddeva hills, Farther to the eastward they are less numerous, but some
occur throughout the upper Son valley, and they gradually die out in
Sargtja and Pal4mau, only 200 miles west of the ground in which the older
lava flows of Rdjmahdl age are seen, and within less than 100 miles of
the Gondwana basins in the upper Damodar valley, which are traversed by
basalt dykes, probably of the same age as the Rdjmahdltraps, Passing
southwards from Jabalpur and Mandld however, there is a total absence of
volcanic intrusions amongst the Vindhyan and Gondwana formations of
Nagpur and Chandé, and none have as yet been noticed in the neighbour-
hood of the Prdnhfta and Godavari between Chdndd4 and Rdjdmahendri.
The country south of the Goddvari and north-west of Haiderdbd4d is
still imperfectly known, but in the south Mardthd country, and the
Konkan near Vengurla, the few dykes which have been observed travers-
ing the unmetamorphosed azoic strata are but doubtfully connected
with the Deccan traps. Ashes, moreover, are much less abundant in this
region, amongst the Deccan flows, than they are further north.

We have thus abundance of evidence of the former existence of
volcanic foci in Cutch, the Rdjpipla hills, and the lower Narbad4. valley,
and probably in the neighbourhood of the Sahyddri range east and north-
north-east of Bombay, whilst there is every probability that vents ex-
tended to the eastward as far as south Rewd and Sargija, but there
‘is no evidence of any having existed in the Ndgpur country or to the
south-east. Yet, as the traps are found represented at RA4jdmahendri, it
appears probable that they once extended over all the Godavari valley,
though it is quite possible that the Réjdmahendri outlier may have been
originally isolated and derived from a centre which has not been dis-
covered,

1 One of these, called Denodhar, was de- ( numerous geological works. The hill is very
scribed originally asa volcano, Geol. Trans., 2nd | probably the basaltic core of a pre-nummulitic,
series, V, 315, (1840), and the statement that | volcano, Its crateriform appearance is due to
it 1s an extinct crater has been repeated in | denudation,
280 GEOLOGY OF INDIA—DECCAN TRAP. [Chap. XI.

It is, however, very clear that the lava flows must have extended to an
enormous distance from the vents through which the molten material was
poured out, Trap dykes are rarely solitary ; they are generally abundant
in the neighbourhood of volcanic foci, and the country in the neighbourhood
of NA4gpur and Chdndé4 has been examined so closely that the improbability
of any intrusions of igneous rock having been overlooked is much greater
than in most parts of the country. It is known that the comparatively
moderate outbursts from existing volcanoes flow to great distances from
their source, while the occurrence of the Deccan traps in immense horizontal
sheets of but slight thickness, but of great horizontal extension, shows that
they must have been extremely fluid when poured out, and it is difficult
to form an accurate idea of the distances to which they may have flowed
before consolidating. Further observations are necessary before all the
sources of the great volcanic series of Western India can be said to have
been even approximately determined."

'

The question of the geological age to be assigned to the Deccan vol-
canic outbursts has been left to the last, because it was desirable to
precede it by a full statement of all the facts upon which a conclusion may
be founded. The evidence to be recapitulated is twofold, that founded on
the affinities of thé fossils found in the various intertrappean rocks, and
that derived from the relations of the stratified traps to beds above and
below them. It is, of course, clear that the traps rest upon cretaceous
beds and are overlaid by nummultitics, and the only question is whether
the lava flows are cretaceous or eocene.

The most important clue to the correlation of the volcanic rocks with
the known series of fossiliferous deposits might be expected to be obtained
from the marine beds associated with the volcanic formations at Raja-
mahendri. This, however, as has been already shown in the description of
the sedimentary beds, proves of little service. So far as is hitherto
known, the relations of both the infratrappean and intertrappean faunas
are with the cretaceous rather than with the tertiary beds, but the points
of connection, in the latter case especially, are quite insufficient to decide
the affinity of the formations,

Turning to the fresh water fauna of the intertrappean beds, the question
arises, as in the case of the Gondwdnas, of the amount of dependence to be
placed upon terrestrial animals and plants as a guide to geological age. In
the case of the Gondwdna formations it has been shown that forms

1 Sir A. Geikie (Nature, 4th Nov. 1880) has | volcanic cones. This would, toa great extent,
suggested that the Deccan traps, like the great | account for the horizontality of the traps, but

‘lava flows of the Pacific slopes of North Ame- | the ash beds must have been thrown out from
rica, were poured out from fissures, and not from | vents, round which cones would accumulate,
Chap. XL] AGE OF THE TRAPS. aha

characteristic of particular epochs in Europe occur, in a very different posi-
tion, in the geological sequence in India, and it is, therefore, necessary to be
cautious in accepting conclusions founded upon slight evidence. There is a
very marked similarity between some of the fresh water mollusca of the
Deccan intertrappeans and species found in some beds of plastic clay age
(lower eocene) occurring at Rilly-la-Montaigne in northern France,1 one
species, Physa gigantea, from the latter locality being considered by some
palzontologists identical with the Indian P. prinsepit. This identifi-
cation is, however, to say the least, extremely doubtful, and the fauna of
the Rilly beds appears more recent than that of the Deccan intertrappeans.
Other resemblances between the plants and fish of the intertrappean beds
and those of the London clay have also been indicated, and a Physa, said
to be allied to P. prinsepzt, has been found in the nummulitic rocks of the
Himdlayas, but even the generic identification in the latter case is far
from certain.?

This evidence only suffices at the most to show an approximation
between the age of the Deccan trap and the lower eocene, and is quite
insufficient to prove whether the former should be classed as upper-
most secondary or lowest tertiary. But the closest relationship of the
intertrappean fauna is with that of the Laramie group in North Ame-

rica. According to Dr. Neumayr® the following species are very closely
allied, if not identical :—

Intertrappean. Laramie,

Physa prinsepit « ‘ . « P. copet.

% 5 var. elongata. . P.disjuncta.
Acella attenuata . . ‘ « A, haldemani.
Paludina vivapai . . . - Hydrobia anthonyi.

: - U. gonionotus.

Unio carteri . « ‘ 3 ‘ f By. goloumbancluk,
Corbicula ingens . . 5 « C. cleburnt.

The Laramie group is regarded as intermediate in age between the
cretaceous and eocene—an age which would harmonise the conflicting

alliances of the fauna of the inter and infra trappean beds near Raja-
mahendri.

The relations between the traps and the underlying cretaceous beds of
the lower Narbada valley have been already described. There isa general,
though slight unconformity, due to subaerial denudation of the underlying
beds. In a very few localities the latter appear to have been disturbed

1 Mem. Soc. Geol. de France, 2nd series, | terrestrial forms.
III, 265, (1848). The genera found at Rilly-la-| 2 D’Archiac and Haime, Description des
Montaigne are Cyclas, Ancylus, Vitrina, Helix, | animaux Fossiles du groupe nummulitique de
Pupa, Clausilia, Megaspira, Bulimus, Achatina,| Inde, Paris, 1853, p. 277.
Auricuia, Cyclostoma, Paludina, Physa,| * Neu. ¥ahrb, Min. Geol., 1884, Bd. 1, p. 74;
Valvata, The majority of these genera are | Records, XVII, 87, (188 ).
282 GEOLOGY OF INDIA—DECCAN TRAP. (Chap. XI.

before the formatign of the lowest traps. Between the highest volcanic
beds and the nummulitic rocks of Surat and Broach the break appears to
be much greater ; not only do the tertiary rocks rest upon a largely denuded
surface of the traps, but they are in a great measure composed of materials
derived from the disintegration of the lava flows, the lowest tertiary beds
being frequently coarse conglomerates of rolled basalt fragments, whilst
beds, hundreds of feet in thickness, are chiefly composed of agates
derived from the traps. This, however, although it proves that great
denudation of the volcanic rocks took place during the deposition of
the nummulitic formations, does not necessarily imply a great break and
an interval of disturbance prior to the commencement of the tertiary epoch,
because the traps, being of subaerial origin, were, unlike most sedimentary
rocks, subject to erosion from the period of their formation. In this case,
however, the unconformity is distinctly marked, and appeurs to show a
great break inthe sequence. The lowest tertiary beds near Surat contain
fossils which appear to be a mixture of middle and lower eocene forms
(Kirthar and Rdnikot).

Farther to the westward, in Cutch, the rocks at the base of the tertiary
group resting upon the trap are locally conformable, and they have even
been considered ! to be partially volcanic, but, as will be shown in the next
chapter, there can be no doubt that a break, marked by unconformity,
exists between the two series. It appears most probable, too, that the
lowest tertiary beds are really composed of detritus derived from the
volcanic rocks, as all appear to be of sedimentary origin, and no instance
has been noticed of intercalation with the lava flows. The great difficulty
of distinguishing between volcanic ash and the detritus of igneous rocks
when mixed with ordinary sediments, especially where the rocks are much
decomposed, as in Cutch, is too well known to require comment. The
beds immediately resting upon the traps are of older date than the num-
mulitic limestone. The trap rests unconformably on neocomian and
jurassic beds.

In Sind the very thin representatives of the Deecan traps may, of course,
only represent a small portion of the period during which the volcanic
rocks were in process of accumulation further to the eastward. One band
rests conformably upon beds which, according to the most recent palzon-
tological investigations, are intermediate in age between the cretaceous
and lower eocene, whilst a second bed of trap is found about 700 feet
lower, interstratified with sandstones.

It will be seen, therefore, that whilst it is clear that the Deccan traps
were poured out in the interval between middle cretaceous and middle
eocene, the evidence tends to show that the lowest volcanic outbursts
were at the oldest of uppermost cretaceous age. That an immense period

! Wynne, Memoirs, 1X, 66, (1872).
Chap. XI] HISTORY OF FORMATION. 283

of time was occupied by the accumulation of the successive volcanic out.
bursts is manifest; long intervals must have elapsed between successive
flows in all those cases in which fossiliferous sedimentary beds are inter-
calated, for these intervals were sufficient to enable lakes to be formed
and stocked with life, and in other cases for rivers to cut beds in the lava
flows, and to fill up those beds with gravel and sand.

Recapitulating the whole evidence, so far as it is presented to us by
the observations hitherto made, we find that in times subsequent to
middle cretaceous, a great area of the Indian Peninsula formed part of a
land surface, very uneven and broken in parts, but to the eastward
apparently chiefly composed of extensive plains, which, by some slight
changes of level preceding the volcanic period, were converted into lakes,
There is much probability that springs charged with silica were common
either at this epoch or shortly after. The lakes had apparently been
drained, and the deposits, which had accumulated in them, had locally
been subject to denudation before the first outbursts of lava took place.
These occurred at considerable intervals, small and very shallow lakes or
marshes being formed in the meantime by the interruptions to the drain-
age produced by lava flows, or by changes of level accompanying the volca-
nic eruptions, In these lakes arich fauna of fish, mollusca, entomostra-
cous crustacea and water plants existed, whilst a varied and probably a
rich vegetation occupied the surrounding country. There is evidence of
the existence of insects and of reptiles, whether terrestrial or aquatic has
not been determined, but hitherto no remains of mammals or birds have
been found—a circumstance which by no means proves that they did not
exist, Fresh flows of lava filled up the first lakes, and covered over the
sedimentary deposits which had accumulated in the waters, but these
very flows, by damming up other lines of drainage, produced fresh lakes,
so that several alterations of lava and sedimentary beds were produced in
places. Gradually the lakes seem to have disappeared, whether the lava
flows succeeded each other so rapidly that there was no time for the
accumulation of sediment in the interval, or whether, as is more probable,
the surface had been converted into a uniform plain of basalt by the
enormous lava streams which had been poured out it is difficult to say,
but no further traces of life have hitherto been found until towards the
close of the volcanic epoch, It is possible that at the end, as at the com-
mencement, of the period, the intervals between eruptions became longer,
and the animal and vegetable life, which may have been seriously dimi-
nished or altogether driven out of the country during the rule of igneous
conditions, resumed its old position, but a great change had taken place in
the long interval, the old lacustrine fauna had died out, and the animals
and plants which now appeared in the country seem to have differed from
284 GEOLOGY OF INDIA—DECCAN TRAP. (Chap. X.

those which had formerly occupied it. Lastly, in the north-western por-
tion of the area, parts of the volcanic country were depressed beneath
the sea, and marine tertiary deposits began to be formed from the
detritus of the extinct volcanoes and their products. A great tract of the
volcanic region, however, appears to have remained almost undisturbed
to the present day, affected by subaerial erosion alone and never de-
pressed beneath the sea level though probably for a time at a lower
elevation than at present.
CHAPTER XII.

CRETACEOUS ROCKS OF THE EXTRA-PENINSULAR AREA.

Absence of break between the cretaceous and tertlary—Isolated exposures of lower cretaceous
beds—Cretaceous rocks of Sind and neighbouring areas—Suldimdn range— Afghdnist4n—
Himélayas— Assam—Burma—Doubtfully cretaceous beds of Tenasserim.

We have already seen that the great distinction between the palzozoic
and mesozoic of Europe does not hold good in India, and that the inter-
val is bridged by rock systems which include beds both of older meso-
zoic and newer palzozoic age. In the last chapter we saw that the
interval between mesozoic and cainozoic is similarly bridged, in the
Indian Peninsula, by the Deccan traps, and we will now find that in extra-
peninsular India, in Sind and Balichistén on the west, as in Assam on the
east, it is similarly impossible te separate the secondary and tertiary
eras, on any but purely paleontological grounds, as they are connected by
a continuous series of deposits ranging from cretaceous to tertiary, which
in the intervening country, is not only carried through the tertiary epoch,
but is extended into recent times,

It is not meant by this that there is on any one section a complete and
conformable sequence extending from cretaceous to recent. The sections
indeed are all imperfect, and unconformable breaks are found in all, but the
unconformity of one section is bridged by a continuous series of deposits on
another, and there is no widespread, universal break which would give a
natural line of demarcation for the separation of the rocks below from
those above them. In many ways it would, consequently, be more natural
to group the cretaceous and tertiary beds together for descriptive pur-
poses, but this would necessitate a system of description that would
obscure some important relations between the tertiary and especially the
upper tertiary beds of widely separated areas, and it will be best to take
advantage of the recognised division between tertiary and secondary, and
confine our attention for the present to the cretaceous rocks.

Before proceeding to the description of the more important and com-
plete exposures it will be well to notice some isolated occurrences of lower
cretaceous beds.

In Cutch there is a thin bed of ferruginous oolitic rock which occurs at
286 GEOLOGY OF INDIA—EXTRA-PENINSULAR CRETACEOUS. (Chap. XII.

the base of the Deccan traps forming Ukra the hill, seven miles south-east
of Lakhpat in north-western Cutch, and rests upon beds of the Umia group.
The outcrop is very ill seen, and nothing has been definitely ascertained as
to the degree of conformity between the cretaceous bed and the underlying
formation, but there appears to be no marked contrast between them.}

The following three Cephalopoda have been obtained from this locality,
Ammonttes martint, A, deshayest, Crioceras australe. The two former
of these occur in the lower greensand (Neocomian) of Europe, and are most
characteristic of the upper portion; the third has been found in cretaceous
beds of Australia, whose exact horizon is not known.

In the Chich4li pass, in the trans-Indus continuation of the Salt range,
the jurassic beds are conformably overlaid by a band of tough black sandy
clay, full of Belemnites, Ammonites, etc., among which Dr. Waagen recog-
nised the lower neocomian form Perisphinctes asterianus.*

This fossiliferous band is overlaid by a band of soft yellowish unfos-
siliferous sandstones which increases in thickness to the west. This was
at first regarded as cretaceous, but it is said to contain pebbles of alveo-
lina limestone in the westerly exposures.°

In the Sirban mountain near Abbottdbdd, the jurassic sandstones are
capped by a bed, ro to 20 feet thick, of a similar, but much harder,
sandstone, which weathers of a rusty brown colour and is commonly crowd-
ed with fossils, These include Ammonites of the groups Cristatz and
Inflati, Ancyloceras, Antsoceras, and Baculttes; Belemnites are abundant,
and the general facies of the fauna is that of the gault. ‘They are succeeded
by a group of thin bedded unfossiliferous limestones which may belong
either to the cretaceous period or to the nummulitics by which they are
overlaid.

The only locality in Sind where beds of older date than eocene have
been identified is in a range of hills running due south from the neigh-
bourhood of Sehwdn, and generally known to Europeans as the Lakhi
range,° from the small town of Lakhi near the northern extremity. South-
west of Amri, on the Indus, a number of very dark coloured hills are
seen in this range, contrasting strongly with the cliffs of grey and

! The only published account of this bed is in
the Pal, Indica, series ix, pp. 245-47, (1875).
No account of the locality was ever printed
by the discoverer, Dr. Stoliczka, and his note
books contain scarcely any details on this
particular point.

7 W. Waagen, Pal. Indica, series ix, p. 245,
1875.

9 A.B. Wynne, Memoirs, XVII, 242, (1880),

* W. Waagen, Memoirs, IX, 342, (1872).

* This range has no general name, different
portions being known by a number of local
terms, It is one of the ranges which combine
to form the Hala range of Vicary and other
writers, and the name is the less inappro-
priate in this case as there is an unimportant
pass through the chain known as the Hala
Lak. Different portions of the range are
known as Tiydn, Kara, Eri, Surjana, etc,
Chap. XII] SIND. 287

whitish nummulitic limestones behind them. These dark hills consist of
cretaceous beds, but the lowest member of the series is only exposed in
a single spot, at the base of a hill known as Barrah, lying about ten miles
south-west of Amri, The whole range here consists of three parallel
ridges, the outer and inner, composed of tertiary rocks, while the inter-
mediate one consists of cretaceous beds, faulted against the lower eocetie
strata to the eastward and dipping under them to the westward. Close
to the fault some compact and hard whitish limestone is found, the lower
portion pure; the upper portion, often containing ferruginous concre-
tions, is sandy, gritty, and forms a passage into the overlying sandstones.
The base of this limestone is not seen, the whole thickness exposed is a
little over 300 feet, and the length of the outcrop does not exceed half
a mile. The limestone is fossiliferous, containing echinoderms and
mollusca, but it is so hard and homogeneous that nothing obtained from it
can be easily recognised, except one fragment of a hippurite. This
fossil is, however, of great importance, because it shows that the white
limestone may very probably be an eastern representative of the hippuritic
limestone, so extensively developed in Persia, and found, in numerous
localities, from Teheran to east of Karman in longitude 58°, just ten
degrees west of the Lakhi range in Sind!’ The precise position of the
Persian hippuritic limestone in the cretaceous series has not been deter-
mined, but the European formation, which is very similar and probably
identical, is of the age of the lower chalk (turonian).

The sandstones resting on the hippuritic limestone occupy a consider-
able tract around Barrah hill, and extend for about three miles from north
to south. They are also seen at Jakhmari to the northward, and in one or
two other places in the neighbourhood. They are gritty and conglomeratic,
frequently calcareous, and contain a few bands of shale, usually of a red
colour, The prevailing colour on the weathered surfaces is dark brown or
purple, many of the beds being highly ferruginous, On the top of the
sandstones is a thick bed of dark coloured impure limestone, containing
oyster shells, and occasionally large bones, apparently reptilian, but none
have been found sufficiently well preserved for identification.

In one place a bed of basalt, about 40 feet thick, has been found inter-
stratified in the sandstones, and it is possible that the band may exist
elsewhere, but it has hitherto remained undetected. The position of this
bed of basalt on the face of a hill called Bor, about 13 miles north of
kK Anikot, is at an elevation of 300 or 400 feet above the base of the sand-
stones, and about twice as much beneath the main band of interbedded
trap, to be described presently.

These sandstones are overlaid by soft olive shales and sandstones,

1 W, T. Blanford, Eastern Persia, London, 1876, Il, pp. 457, 485.
288 GEOLOGY OF INDIAmEXTRA PENINSULAR CRETACEOUS. (Chap. XII.

usually of fine texture. The sandstone beds are thin, and frequently have
the appearance of containing grains of decomposed tasalt or some similar
volcanic rock, or else fine volcanic ash. A few hard bands occur, and
occasionally, but rarely, thin layers of dark olive or drab impure lime-
stone. Gypsum is of common occurrence in the shales.

The olive shales are highly fossiliferous, the commonest fossil being
Cardtta beaumonti,} a peculiar, very globose species, truncated posteriorly,
and most nearly allied to forms found in the lower and middle cretaceous
beds of Europe (Neocomian and Gault). This shell is extremely abund-
ant in one bed, about 200 or 250 feet below the tep, but is not con-
fined to this horizon. Nauti/i also occur, the commonest species close-
ly resembling MN. dabechei.
of Messrs. D’Archiac and
Haime, but differing in the
position of the siphuncle.
This form appears undis-
tinguishable from XN. bouchar-
dianus, found in-the upper
cretaceous Ariyalir beds of
Pondicherri and at a lower.

Fig. 17.—Cardi'a beaumonti, DArch, and Haime. cretaceous horizon in Europe.
A second Nautilus resembles N. subfleurtaustanus (another eocene Sind
species) in form, and is also allied to some cretaceous types. Several
Gasteropoda occur, especially forms of Rostellaria, Cyprea, Natica, and
Turritella but none are very characteristic. Two forms of Osérca are
common, one of them being allied to the tertiary O. flemingz and to the cre-
taceous O. atttelfana, but distinct from both. The only mollusc which
certainly passes into the Rdnikot beds is Corbula harpa.

In the lower part of the beds with Cardita beaumontd some amphiccelian
vertebrae were found, which Mr. Lydekker has ascertained to be crocodilian,
All amphiceelian crocodiles are mesozoic, and the present form must be one
of the latest known. So far as it is possible to form an opinion from very
fragmentary materials, the vertebre in question appear more nearly allied
to the wealden Suchosaurus than to any other form hitherto described.
It has, however, been already shown, when writing of the Gondwd4na flora,
that the distribution of Repfilia in past ages was not the same in India as
in Europe. :

Only the corals and echinoids of the Cardita beaumonti beds have as
yet been criticically examined, the former by the late Prof. P. M. Duncan, the
latter by the same palzontologist with the assistance of Mr. P. M Sladen.
The results obtained are not very definite so far as the corrclation of

 

' D’Archiac and Haime, Description des de inde, Paris, 1853, p. 253, pl. xxi
Animaux fossiles du groupe Nummulitique fig. 14.
Chap. X10) VOLCANIC BEDS IN SIND. 289

the beds is concerned. Among the corals the genera Caryophylla, Smila-
trochus and Litharea are cretaceous, but also range into and through the
tertiary, and the Smilotrochus blanfordi.very closely resembles S. incurvus
of the Italian eocene.! The echinoids are equally indefinite as regards
their relations, being neither distinctly cretaceous, nor definitely eocene,
and the general facies is such as to indicate an age intermediate between
these two periods,”

Mention has already been made of one bed of basalt intercalated
in the sandstones above the hippuritic limestone and a much more import-
ant band of the same igneous rock has been traced, resting upon the
Cardita beaumonti beds, throughout a distance of twenty-two miles from
Rdnikeot to Jakhmari, about seventeen miles south of Sehwan, wherever the
base of the Ranikot groupis exposed. The thickness of this hand of trap
is trifling, and varies from about 40 to about go feet. Apparently in some
places the whole band consists of two lava flows, similar in mineral charac-
ter except that the upper is somewhat ashy and contains scoriaceous frag-
ments ; the higher portion of each flow is amygdaloidal, and contains nodules
of quartz, chalcedi ny and calcite, and in places the nodules are surrounded
by green earth, as is so frequently the case with the Deccan traps.
Another characteristic accessory mineral, common also in the traps of
the Deccan and MAlw4, is quartz with trihedral terminations. The basaltic
trap of the Lakhi hills is apparently of subaerial origin, although it rests
con‘ormably on the marine (or estuarine) Cardita beaumontibeds. There
is nothing in the igneous bed to indicate its having consolidated otherwise
than in the air, and the structure differs altogether from that of subaqueous
volcanic tuffs.

The evidence that this band of basaltic rock is interstratified and not
intrusive is ample; throughout the whole distance the trap is found in
precisely the same position, between the lowest beds of the Rdnikot and
the highest of the Cardrta beaumonti groups, and apparently perfectly
conformable to both. The close resemblance in mineral character and
the similarity of geological position, at the base of the tertiary beds, show
that this band must in all probability be a thin representative of the great
Deccan trap formation, and the occurrence of a second bed at a lower
horizon, interstratitied with the passage beds between cretaceous and
tertiary, tends strongly to confirm the inference drawn from the relations
of the traps to the cretacecus and tertiary rocks of western India, that
the great volcanic formation must be classed as intermediate in age be-
tween those two eras.

\

In Balichistdn the section of cretaceous beds is more extensive than

" Pal. Indica, series xiv, |, pt. ii, p, 25, (1880). | Pat. Indica, series xiv, I, pt. iii, p. 28, (1882),
290 GEOLOGY OF INDIA—EXTRA-PENINSULAR CRETACEOUS. ([Qhap, XII

that seen in the one small exposure in lower Sind, and none of the hori-
zons except the uppermost have been identified in the two regions.

In the neighbourhood of Quetta and the country to the east of it the
lowest rock known is a massive limestone of great thickness, The lower
portion of this, as seen at the head of the Bolan pass and at Sariab, is of
a pale cream colour, the upper portion is more or less dark grey in
colour. Fossils are not very abundant, but in places it exhibits sections
of Hippurites, Inoceramus and corals on the weathered surface. The
exact age of this limestone is undetermined, but it is regarded as lower
cretaceous,

The massive grey limestone is succeeded by a series of dark grey or
black shaly beds, often containing an admixture of volcanic ash, overlaid by
red and green mottled shales and thin bedded limestone, capped by white
limestone. The thickness of these beds appears to be about 1,000 feet,
about 200 at the top of which consists the white limestone, but there
are great variations in thickness of both members, owing partly to the
manner in which the soft shaly beds have yielded to compression, and
partly to the removal of the upper beds by denudation previous to the
deposition of the next succeeding group. Locally the shaly beds of this
group are abundantly fossiliferous, but except a few fragments of ammo-
nites the only fossils that have yet been found are belemnites, mostly be-
longing to the section Dé/atatz, and including one very broad and flattened
form ; owing to the abundance of these fossils the group has been referred
to, in previous publications of the Survey, as the Belemnite beds!

The age of this group is clearly secondary, yet on some sections the
white limestone at its summit contains numerous specimens of Mummu-
lina and Alveolina, which are usually regarded as indicative of a tertiary
age, thus introducing an anomaly which is repeated in the next succeeding

group.

According to any local system of classification the next succeeding
group of strata would be separated from those just described and united to
the overlying Nummulitic beds, with which it is perfectly conformable, for
there is a slight but distinct unconformity at the top of the Belemnite
beds. The unconformity is unaccompanied by any recognisable want
of parallelism of stratification between the beds below and above it, but is
marked bya considerable degree of erosion and a complete change of
fauna, none of the belemnites having been found in the overlying group.

The Dunghan group as this is called, from a hill of the same name east
of Spintangi, is an important and ‘interesting one, In the neighbour-
hood of Hurnai it is essentially a limestone formation and caps the bare

' Records, XXV, 19, (1892).
Chap. XID] BALOCHISTAN, 291

=

hogbacked hills east of the Hurnai route to Quetta. To the south of the road
which connects the Spintangi railway station with Thal Chotidli, the lower
beds become argillaceous, and the argillaceous element more and more
replaces the calcareous till, in the hills east of Khattan, the group has
become essentially a shale group in which the calcareous element is quite
subordinate, A similar change takes place in the hills west of the Bolan
pass, but the country there has been less fully examined.

In the hills inhabited by the Mari tribe the lowest beds of the group-are
usually unfossiliferous, grey, green, and purplish shales, overlaid by about
1,000 feet of grey shales, many beds being so profusely fossiliferous as
to become impure limestones. Above these shales there is a band of 100
to 200 feet thick, composed principally of more or less impure sand-
stones, capped by a limestone composed almost entirely of oysters (Zxo-
gyra?), but containing also a few Nazézl¢ and other fossils. The oyster bed
is separated by some 600 feet of beds, on some sections of limestone on
others grey shale, from a peculiar band of pseudo-breccia regarded as the
base of the nummulitic series in this district.

The fauna of this group is a peculiar one; nummulites are abundant,
but associated with them are Crtoceras, Baculttes and Ammonites, while
fully half the echinoderms belong to the order LEehinocontdz, and an
oyster resembling O. car/nata is not uncommon. Cardita beaumonti was
not found, but the admixture of characteristically cretaceous forms with an
abundance of nummulites points to the group being intermediate in age
between the cretaceous and eocene periods, and consequently equivalent
to the Cardita beaumont? beds of Sind. The suggestion is supported by
the frequent occurrence of beds of impure volcanic ash, immediately above
the sandstones and oyster bed, and of ash beds and even basaltic trap
apparently interbedded with the uppermost beds of the Dunghan group in
the Bolan pass. It is natural to suppose that they represent the same
horizon as the trap above the Cardita beaumont? beds in Sind.

It is not known how far the grouping adopted in the country east of
Quetta holds good for the rest of Balichistdn, as this country has never
been geologically examined with any thoroughness. Some particulars
of observations made on his journeys through Balichistdn, have been
recorded by Dr, Cook, in’which it appears to be possible to recognise the
various rock groups mentioned above. He describes! the cretaceous
rocks as consisting of “more or less compact, fine grained, red and
white limestone, interleaved with slabs and veins of chert; the lime-
stone generally containing fine microscopic specks, and the upper part
one or two massive strata of an excessively hard limestone abounding
in Orbtioides, Orbitolina and Operculina, the lowér strata becoming

1 Trans, Med, Phys. Soc., Bombay, V1, 101, (1860).
292 GEOLOGY OF INDIA—EXTRA PENINSULAR CRETACEOUS. (Chap. XT.

argillaceous, shaly and containing (rarely) ammonites.” These beds are
underlaid by a dark blue fossiliferous limestone containing RAynconella.
There is, however, some uncertainty about the section, for near Khelat the
white limestone appears to underlie the fossiliferous shales.

There is some indication in this description of the massive limestone,
belemnite beds, and Dunghan group further east, and with one exception
the fossils recorded would accord with the identification. Ammonites,
Ceratites, Crioceras, and Belemnites have all been found in the eastern
area, and Scaphites might well accompany them, but if the identification
of Orthoceras, which has been referred to in a previous chapter,’ was
correct, it must belong to an older set of beds than cretaceous, and sug-
gests that the section is more extensive and less simple than Dr. Cook’s
descriptions would indicate.

In the Suldimdn range, west of Dera Ghazi Khan, the cretaceous rocks,
so far as they are exposed, comprise two well marked stages. The lower
consists of dark grey limestones, occasionally sandy or shaly, passing down-
wards into dark to bluish grey, often nodular, calcareous shales. The lime-
stone abounds in indistinct fossils, especially foraminiferz, and in the under-
lying shales a cephalopod belonging to the Ammonitndz, [noceramus and
two species of Exopyra resembling cretaceous forms have been found.?

These beds, of which about 1,500 feet are exposed, are overlaid by about
the same thickness of sandstones, generally white or pale coloured, brown,
greenish or purplish grey. No fossils have been found in the sandstones,
and no unconformity has been detected between them and the overly.
ing beds, while they overlie a bed of pseudo conglomeratic limestone
exactly resembling that found at the base of the eocene beds of eastern
Baldchistan,

Further north, in the neighbourhood of the Takht-i-Sul4iman, the same
pale sandstone and underlying shales and limestone are found underlaid
by some hundreds of feet of massive grey limestone, showing sections of
Inoceramus and corals on the weathered surface, which exactly resembles
the massive limestone of the Quetta neighbourhood, and is probably of
the same age. West of the Suldimdn range, in the direction of the Zhob
valley, this massive limestone is underlaid by a.great thickness of green
and grey slaty shales, intercalated with beds of sandstone and a few of
limestone, from which no fossils have as yet been obtained.

Before passing on to the cretaceous beds of Afghdnistdn, it will be
interesting to notice that the rocks just described are the source of the
petroleum of eastern Baltichistan and of the Suldimdn range. The Bali-
chist4n petroleum is a thick, black, tarry maltha, traces of which are very

' Supra, p. 143. { 2 Memoirs, XX, 217, (1883).
Chap. XII] AFGHANISTAN. B69

frequently found in the Dunghan group. At Khattan there is a natural oil
spring, where the petroleum issues along with an abundance of hot sulphur-
ous water, and for seven years past an attempt to work this oil for profit
has been in progress, but the quantity obtained has not heen sufficient to
prove remunerative. It has been supposed that the greater abundance of
the oil was in some way connected with the unusual profusicn of organic
remains in the Dunghan group at this locality, but the connection is by no

neans clear, A precisely similar oil is found in the Bolan pass near Kirta,
and in the Robdar valley, south of Bibi Nani, it issues from the limestones
below the Belemnite beds, that is to say, from rocks much older than the
Duaghan group. No rocks older than this limestone are exposed in the
Bolan pass, and it is impossible to say whether there are any profusely
fossiliferous beds underground, analogous to those of the Dunghan group
at Khattan, but the widespread presence of traces of a similar oil, even
where fossils are rare, appears to indicate that the concentration of the oil
at Khattan, Kirta, and the other localities, has no connection with the
greater or less profusion of organic remains at those spots.

In the country round Kandahar,! and between itand the Khw4ja Amrdn
range, the cretaceous system is represented by a great thickness of hard
grey limestone, usually unfossiliferous but locally containing an abundance
of Hippurites, corals, etc. Near Kandahar this limestone is underlaid by
a series of beds showing the following sections in descending order :—

3. A shaly sandstone, made up more or less of trappean material.
2. Bright green and intensely red shales with thin sandstone bands of trappean
substance.

1. Coarse and thick conglomerate, almost entirély made up of pebbles of trap and
cemented by a trappean, though calcareous matrix.

The description of these beds agrees well with certain beds seen near
Kach and Hamadun on the Hurnai route to Quetta, which were formerly
regarded as cretaceous. More recent examination has shown, however,
that they are nummulitic and consequently cannot be representative of the
Kandahar beds, unless one of the sections has been misinterpreted.?

Associated with the cretaceous limestone there are intrusive rocks,
both basic and acid, and bedded traps. The latter are basaltic in character
and overlie the limestone; very few details have been recorded, but one of
the original foci of eruption was supposed to have been recognised about
four miles west of Kandahar. These bedded traps are newer than the
trappoid conglomerates described above, which occur below the limestone:

The intrusive basic rocks are said to be fithologically similar to the
bedded ones, but they have undergone a serpentinous change, and contain

'C. L. Griesbach, Memoirs, XVIII, 42. { 2? R.D. Oldham, MS. report, (1891).
(2881). 3 Memoirs, XVIII, 52, (1881).
W 2
294 GEOLOGY OF INDIAmEXTRA PENINSULAR CRETACE JUS. (Chap. XI

veins and lumps of bright green and yellowish chrysotile where in contact
with the limestone. The intrusions are numerous and vary in size cown
to mere wafer like strings.

The acid intrusive rocks are varieties of quartzsyenite, sometimes
porphyritic in the hills crossed by the Maiwand pass west of Kandahar, which
occur in veins and dykes varying from several hundreds of feet to quarter
ofan inch in thickness, In the larger syenitic masses numerous veins of
a porphyritic rock with crystals of pink orthoclase imbedded in a fine
grained felsitic matrix are found. Similar syenitic intrusions were ob-
served at Dabrai and on the western side of the Khwdja-Amrdn pass.
The syenitic intrusions are of somewhat older date than the basaltic,
as the former are penetrated by dykes of the latter.

In Afghdn-Turkistén the cretaceous system is well developed, and covers
a large area of ground in which the older rocks only appear as inliers here
and there. The lower cretaceous beds, consisting of about 800 feet of
sandstones and shales with earthy limestones at the top, are said to be
conformably underlaid by the red neocomian grits? at the top of the
plant bearing series. The upper cretaceous is formed by about 1,800 to
2,000 feet, thickening to 4,000 in the sections near Balkh, of white thick
bedded limestone with occasional sandstone bands. No defined subdivi-
sions were recognised, but the limestones may be divided into three zones?:-=

3. Chalk with flints.
2. Concretionary earthy white or brownish white limestones, occasionally dolomitic.
1. Hard white splintery limestones.

In the north-west Himdlayas the cretaceous system, apart from the
possibly cretaceous Giumal sandstones, is represented by a few small
patches left on the tops of some of the hills in Spiti. They were named
the Chikkim series by Dr. Stoliczka, who described it as consisting of a
maximum thickness of about 500 feet of bluish or greyish white Tinfestone,
weathering white, with occasional earthy bituminous bands, overlaid by
about 200 feet of grey or darkish unfossiliferous marly shale. The lime-
stone yielded several fragments of Rudistes and numerous Foraminiferz$

Precisely similar limestones were observed further east by Mr. Griesbach
in Hundes. They exhibit no features calling for special notice, and the
fossils collected have not yet been described,

To the north hippuritic limestone has been observed in the Lokhzung
range, north of the Lintzihang plain,* and at Sanju, on the road from

1 Supra, p. 196. 3 F. Stoliczka, Memoirs, V, 116, (1865).

2 C. L. Griesbach, Records, XIX, 253, ‘Drew: Jummoo and Kashmir Territo«
(3886), ties, London, 1875, p. 343.
Chap. XID] ASShM: 295

Leh to Yarkand, Dr. Stoliczka recorded the presence of coarse grey
calcareous sandstones and chloritic marls, some beds being almost entirely
composed of the middle cretaceous Gryphexa vestculosa,'

The occurrence of cretaceous beds on the shores of the Namcho lake,
about 75 miles north of Lhasa, is proved by specimens of Omphalia trottert,
which were brought from that locality by one of the native explorers of
the Trigonometrical Survey in 1876."

In the Assam range the cretaceous rocks occur both on the plateau,
where they lie nearly horizontal, and along the southern edge, where
they are bent down to a steep dip in a monoclinal flexure. They thin
out in a marked manner to the northward on the section south of Shillong,
having a thickness of about 600 feet at the edge of the scarp, while
ten miles further north, near Surarim, there is only about 100 feet.
Still further north there are some small outliers which lie in hollows on the
surface of the Shillong quartzites marking the position of pre-cretaceous
valleys.

It is in these little primitive basins on the plateau that the cretaceous
coal is found, one of them, a tiny coal basin at Mdobehlarkdr between
Surarim and Mauphlong, having for years supplied the station of Shillong.
The mineral itself has a persistent character throughout the whole cretaceous
area. It is remarkable as being less of a true coal than is that of the over-
lying nummulitic group; the texture is compact and splintery, with a
smooth conchoidal pasture, and the coal gives a dull wooden sound when
struck, It has the additional peculiarity of containing numerous specks and
small nests of fossil resin,

The most persistent member of the cretaceous series is known as
the Cherra sandstone, about 200 feet of coarsish hard rock, unfossiliferous
except for some vague stem-like vegetable impressions, which comformably
underlies the nummulitic limestones. ‘The: next most constant member
is the basal conglomerate, whose larger components are almost all derived
from the neighbouring Shillong quartzites, and are generally subangular,
Varying in thicknes from 20 to 100 feet, it everywhere forms the base
of the series, but whether it represents a definite geological horizon is
doubtful.

In the Maobehlarkdr coal basin the basal conglomerate and the Cherra
sandstone are in contact, but at the south scarp of the plateau they are
separated by glauconitic sandstones, overlaid by a pale fine grained sand-
stone, often containing broken plant remains, and in places marine fossils.

IF, Stoliczka, Quart. Your. Geol. Soc., | Mission, Geology, p- 22, (1878).
XXX, 572, (1874); Records, VII, 50, (1874); | * Records, X, 21, (1877).
Scientific Results of the Second Yarkand
296 GEOLOGY OF INDIA—EXTRA PENINSULAR CRETACEOUS. (Chap. XIL

These beds, about 400 feet in thickness, thin out to the northward by an
original limit of deposition, and it has been noticed? that the matrix of the
basal conglomerate, at the different levels, partakes of the nature of the
corresponding horizon below Cherra, and on all the sections there is more
or less of atransition, by interstratification, between it and the particular bed
which happens to overlie it. The only point tending to cast a doubt on its
being a marginal form of the successive sandstone beds is the frequent
occurrence of carbonaceous matter in the rock immediately above, but
this is not conclusive.

Where the beds bend over, and are exposed with a high dip in the
low ground south of Tharia, the basal conglomerate is represented by a
coarse felspathic ochrey sandstone, while the overlying beds, having a
thickness of about 1,200 feet, consist of pale grey shales, locally nodular,
calcareous, or ferruginous, with some thin layers of earthy limestone or
sandstone, The whole series, besides being thicker than that exposed on
the plateau north of the uniclinal axis, is earthy im character, instead of
sandy, implying a greater distance from the margin of the sea.

The marine fossils of the cretaceous rocks of the Khasi hills have
already been mentioned,? and need not be further referred to here. No
fossils have as yet been found west of the Khdsi hills,

In the Géro hills the cretaceous attains a considerable development
as an arenaceous series, containing impertant coal seams in places. The
sandstones of the plateau are horizontal and rest on a more deeply eroded
and irregular surface than those of the Khdsi hills, At the western end
of the range the sandstones lap round the end of the Turd gneissic ridge,
and the original relations of the rocks are nowhere better seen than
here. The spur on which the station of Turd stands, some 2,000 feet
below the crest of the ridge has a midrib of gneiss, with sandstone on both
sides, through which the streams have again excavated their channels.
There is but little disturbance in this locality, and it is plain that the
ridge must have stood as it does now when these sandstones were laid
down.

East of the Khdsi hills, throughout the south-east portion of the
Jaintia hills from the neighbourhood of Jowai eastwards, cretaceous rocks
are found at the surface, horizontal or nearly so, and to the eastwards pass
conformably beneath the tertiaries near the Kapili (Kopili) river. Beyond
this we have only isolated observations; the thin bedded sandstones at the.
falls of the Kapili are beheved to be cretaceous, and typical cretaceous
coal is associated with sandstone and some hard sandy limestone resting
flatly on the gneiss in the Ndmbar and Doigrung valleys, near Goldghat.

+H. B. Mediicott, Memoirs, VIL, 571, (8860). | ® Supra, p. 247.
Chap, XII] MA-f GROUP. 297

The existence of cretaceous beds in the Arakan Yoma is only shown
by the discovery of one species of mollusc ina single locality near Ma-f
in the northern part of the Sandoway district of Arakan, ‘The species
found, Ammonites tnflatus, is a characteristic cenomanian cephalopod,
common in the Utatur beds of Southern India. The only specimen
obtained was picked up in the bed of a stream, and had evidently been
derived from some shales in the neighbourhood. No other specimens
nor other fossil of any kind could, however, be found.

What may be the extent of the cretaceous beds, and which strata should
be referred to this group, are matters on which but little trustworthy in-
formation has been obtained. Mr. Theobald was disposed to consider that
a peculiar, compact, light cream coloured, argillaceous limestone, resem-
bling indurated chalk, sometimes speckled from containing sublenticular
crystalline particles, belongs to the cretaceous system. This limestone
has been traced at intervals from near Ma-ji, about thirty miles north
of ‘Tongip (Toungoop), to the neighbourhood of Sandoway, whilst some-
what similar limestone, though not so characteristic, may be traced to
Keantali, some thirty miles farther south, The same limestone is found in
the western part of Ramri Island. Another peculiar formation isa greyish
rather earthy sandstone, with a pisolitic structure in places, due to the
presence of small globular concretions of carbonate of lime and iron, The
concretions decompose and leave small holes, which impart to the earthy
sandstone the aspect of an amygdaloidal trap. Like the limestone, this
peculiar sandstone is traced from Ma-i to near Keantali, a distance of 94
miles, and if, as appears probable, these beds are really cretaceous, for
both are closely associated with the shale from which the ammonite had
apparently been derived, the rocks of this formation may be considered as
extending at least the distance mentioned. The strata ascribed to the
cretaceous group are less hardened ard metamorphosed than the other
rocks of the Arakan Yoma; they are of great thickness, and may include
all the beds of the main range of the Yoma, as far south as Keantali. No
rocks which can be referred to the Ma-f group have been detected east of
the main Arakan range in Pegu. To the northward their range is
unknown, but a limestone resembling that of the Ma-i group was seen in
the hills east of Manipur.?

Apart from those just mentioned no rocks of cretaceous age are known
to exist in Burma, though there is a probability that they may be repre-
sented in Tenasserim. On the Lenya river,® in the extreme south of the
province, a bed of coal occurs, of very laminated structure and containing
numerous small nodules of a resinous mineral, like amber. This peculiar

\Memeirs, XIX, 223 (1833) ; Supra p. 148, [* T. Oldham, Sel. Ree. Goot. Ind., X, 48, (1856).
298 GEOLOGY OF INDIA—EXTRA PENINSULAR CRETACEOUS. (Chap. XIL

association of mineral resin is characteristic of the cretaceous coals in the
Assam hills, and it is possible that the Tenasserim mineral is of the same age,!
At the same time no paleontological evidence has been discovered, the rocks
associated with the coal are soft clays and sands, having a more recent
appearance than those accompanying the other coal seams of the Tenas-
serim province, and these other coal seams are, it is believed, not older
than eocene. The coal occurs in an irregularly developed bed, varying
from 1 to 5 feet, or rather more, in thickness, with thin layers of fine
jetty coal between bands of hard black shale, and rests on clay with
vegetable remains, and patches of jet coal, Thin coal laminz are also
found in the associated strata.

Below the rocks immediately associated with the coal are fine, whitish
earthy sandstones and indurated clay, passing into marl, with some con-
glomerates. Above the coal is a series of soft muddy sandstones, marls,
conglomerates and a few seams of carbonaceous matter. The whole may
be 600 feet thick. The dip is considerable, about 35°, and the rocks
have undergone disturbance and faulting. Nothing has been ascertained
as to the relations of the coal bearing beds to other formations, indeed
all that is known of the. Lenya river coal is the result of a hurried visit
to a locality very difficult of access. =

.
! Mr. Bose (MS. Report, 1892) regards these | larly ill adapted to geological investigation
beds as belonging to the Maulmain series, of | and the examination wag necessarily incom.
Palzeozoic age ; the country is, however, singu- | plete,
CHAPTER XIIL.

TERTIARY DEPOSITS.

(Excluding those of the Himdlayas.)

Quilon and Ratndgiri—Surat-—Sind and Baléchist4n—Cutch and Kathidwar—Afeh4nistan—
Kohd4t—Assam—Burma,

No tertiary beds are known in the Indian Peninsula except in the im-
mediate neighbourhood of the coast, and if we exclude certain unfossiliferous
sandstones, now regarded as subrecent though possibly of upper tertiary
age, they are confined to a few small exposures on the west coast, the
most southerly of which is near Quilon, in Travancore.

The earliest, and practically still the only, information published on
the occurrence of tertiary beds in Travancore is comprised in some
notes supplied by General Cullen to Dr. Carter, and published by the latter
in his ‘Summary of the Geology of India’! Beneath the laterite of the
neighbourhood of Quilon, at a depth of about 4o feet from the surface,
grey fossiliferous limestone (or dolomite according to General Cullen)
is found, partly compact and partly loose and rubbly. This limestone
is exposed beneath a laterite cliff near the coast, four or five miles north-
east of Quilon, and the same rock has been found in the neighbourhood of
the town at a depth of about 4o feet in numerous wells, many of which
were sunk or deepened by General Cullen for the purpose of ascertaining

the presence of the limestone. Further south, near Warkalli, twelve to
fourteen miles south of Quilon, the cliffs on the coast expose, beneath the

laterite, beds of brightly coloured sand and clays with bands of lignite,
abounding in fossil resin and iron pyrites, both in lumps of considerable
size. The sandy beds overlie the lignites and clays.

The limestone contains marine shells in abundance, amongst which
Dr. Carter recognised Strombus fortisi, Cassts sculpta, Voluta jugosa,
Ranella bufo, Conus catenulatus, Conus marginatus, and Cerithium rude,

1 Four. Bo. Br. Roy. As, Soc., V, 301, (1857); | the geological examination of the Cochin
and Geological Papers on Western India, Bom- | neighbourhood. The locality had been incor-
bay, 1857, pp. 740 and 743, footnote, This foot- rectly defined in the first instance, but its
note is an addition to the original summary. | existence was subsequently verified by Mr.
The very small outcrop was not found during | Logan; Records, XVII, 9, (1884).
300 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap, XIII.

besides species of several other genera resembling forms found in the
tertiary beds of Sind and Cutch, A species of Oréztolites (?) was
described by Dr. Carter as O. malabarica. All the mollusca identified
belong to species occurring also in Cutch and Sind, and, so far as is
known, in beds of later date than the nummulitic limestone. No plants
appear to have been collected from the lignite beds.

Another deposit of obscure date and origin has been found beneath
laterite at Ratnagiri (Rutnagherry) on the western coast.!| White or
blue clays with thin carbonaceous seams are found in various quarry and
well sections near the town beneath a considerable thickness of laterite,
35 feet in one case. Some of the clay is said to be sandy or gravelly.
Above the deposit is a layer of hard ironstone, about an inch thick, but
said sometimes to be thicker. As in Travancore, fruits and leaves are
found in the clay and lignite, together with mineral resin and pyrites,
No specimens of the organisms found appear to have been collected. The
beds are only a few feet thick, 27 in one section measured by Dr. de
Crespigny, and rest unconformably upon Deccan trap.

There is but little evidence to connect this deposit with the Travancore
beds, but, owing to some similarity of mineral character, the presence
of lignite in both, and the circumstance that both underlie laterite, they
have been classed together,

The tertiary rocks in Surat and Broach® are almost confined to two
tracts of country, separated from each other by the alluvium of the river
Kim, a small stream running to the sea from the Rajpipla trap area, The
southern tract is the smaller, extending about ten miles north from the Tépti
river and being about fifteen miles broad from east to west; the other
area, between the Kim and Narbadd, extends about thirty miles from
north-east to south-west, and is about twelve miles across where widest, In
both the few good exposures of rock which occur are to the eastward.

At the base of the tertiary formations, north-east of Surat, are thick
beds of ferruginous clay, assuming, where exposed, the characteristic brown
crust and pseudo-scoriaceous character of laterite, from which they differ
in no respect. These beds at first sight appear to be of volcanic origin,
an idea which is strengthened by the neighbourhood of the traps on which
they rest, but close examination has shown that they are really sediment-
ary deposiis, although composed, in all probability, of materials derived
from the disintegration and denudation of the trap. With them are inter-
stratified beds of gravel or conglomerate, containing agate pebbles derived

Carter, Four. Bo. Br. Roy, As. Soc., V, 626, | 8on, Records, IV, 44, (1371).
(1857); Geological Papers on Western India, * Fora fuller description, sce Memoirs, VI,
Bombay, 1857, p. 722; footnote; C. J. Wilkin- | 223-27 and 356-73, (1869),
Chap. XTIL.] SURAT. 301
from the traps and limestone, sometimes nearly pure, but more frequently
sandy, argillaceous, or ferruginous, and abounding in nummulites and other
fossils. The thickness of the whole can only be roughly estimated at
between 500 and 1,000 feet.

These beds are well seen on the banks of the Tdpti below Bodhdn, a
village eighteen miles east by north from Surat. They extend thence
to the northward through Tarkesar to the Kim alluvium, and again north
of the Kim to the neighbourhood of a village called Wdgalkhor, about
twenty-four miles north-north-east of Bodhdn, and seventeen east by south
of Broach. North of this they appear to be overlapped by higher beds.

The nummulitic limestones and their associates are distinctly uncon-
formable to the underlying traps, and rest upon the denuded edges of the
latter. Amongst the fossils found in the lower tertiary beds are Nummu-
lites ramondt, N. obtusa, N. exponens (or N. granulosa), Orbitoides dis.
pansa and some other species which are common in the Kirthar beds
of Sind, together with Ostrea flemingt, Rostellaria prestwicht, and
Natica longispira, which are particularly characteristic of the Rdnikot
group, and Vulsella legumen found in both. Some other fossils have been
identified with species found at a higher horizon, but the identification
appears doubtful. The nummulitic beds of Surat and Broach may safely be
classed as eocene,

Above the limestones and lateritic beds there is found a great thickness
of gravel, sometimes cemented into conglomerate, together with sandy clay
and ferruginous sandstone, often calcareous. These higher beds are poorly
exposed in the Tapti and Kim rivers, but they are well seen in the stream
which runs past Ratanpur, east of Broach. -Here tkey consist chiefly of
sandstone, gravel, and conglomerate, with occasional beds of red and white
clay and shales. The pebbles in the gravels and conglomerates consist
chiefly of agates and quartzose minerals derived from the trap, and
from some of these beds near Ratanpur, east of Broach, the agates and
carnelians are obtained which have from time immemorial supplied the
lapidaries of Cambay. At the base of the teitiary beds in this direc-
tion is a coarse conglomerate composed of large rolled fragments of basalt,
but it is uncertain whether this bed belongs to the upper tertiary group
or to the lower tertiaries, as it is not quite clear, owing to the few sections
exposed, whether the lower eocene beds are completely overlapped
to the northward, or merely represented by unfossiliferous beds of a
different mineral character. Like the underlying beds, the higher tertiary
strata have a steady dip to the westward, and the thickness of the whole
tertiary series exposed near Ratanpur appears to he between 4,000 and
5,000 feet, but this estimate is based on a very imperfect exposure of the
rocks. Of course, if, as appears possible, the lower beds are overlapped, the
whole of this thickness consists of the upper members of the series.
302 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap, XIII.

No nummulites are found in these upper tertiary beds, and the few
fossils discovered in them appear to differ from those in the nummulitic
limestones below. The commonest organic remains are valves of Badanz,
which are also abundant in the G4j (miocene) rocks of Sind. The abund-
ance of Balani and the absence of Nummulites together form strong
reasons for believing that the upper beds of Surat and Broach are of
later date than eocene.

It is far from certain whether any pliocene beds are found in eastern
Gujardt. They occur in Kéthiawdr and on Perim Island in the Gulf of
Cambay, and further search may detect them in Surat and Broach.

Tertiary deposits are found in the debateable ground west of the
Ardvallis which belongs structurally to the peninsular and stratigraptically
to the extra-peninsular area, and in the extra-peninsular area they attain
an immense development, both as regards their thickness and the area
they cover. Taken as a whole, and ignoring local breaks in the continuity
of deposition, they form a great system of deposits whose lower portion is a
marine formation while the upper consists of fresh water subaerial deposits.
The distinction is not absolute, nor can the line of demarcation between
the two types be everywhere drawn on the same horizon, yet the distinc-
tion isa real and important one. Everywhere, from Sind onthe one hand
to Burma on the other, the eocene deposits are marine and the pliocene
fresh water or subaerial, with the possible exception on of the pliocene ede
of the Irawadi valley, and wherever there is a continuous succession of
deposits, there is a gradual transition from the one type to the other. .

It will be well to commence the general description of the extra-penin-
sular tertiaries with those of Sind, as, owing to the completeness of the
section there and the abundance and excellent preservation of the fossils
of the various horizons, it may well be regarded as a type area for the rest
of India. At the same time the tertiaries of those adjoining areas will be
noticed where the Sindrock groups have been recognised with some degree
of certainty, those of Cutch and Kdthidwar, where they are less distinctly
represented, being taken separately. After noticing the tertiaries of
Afghdnistén and the western frontier, those of Assam, and finally Burma,
will be described, the consideration of the Himalayan tertiaries being
more conveniently postponed to the following chapter.

The great series of tertiary deposits of Sind has been divided into the
following groups or subdivisions, whose approximate correlation with the
European sequence is given—

Manchhar, 8,000—10,000 ft. . . Upper miocene to pliocene.
Gaj, loo—1,500 ft. ; . + Miocene.
Chap. XIII] THE RANIKOT GROUP. 303

Nari, soo— 6,000 ft. . . Upper eocene to lower miocene.
Kirtar, 6,000—9g,000 ft. . , . Eocene.
Ranfkot, 2,000 ft. ‘ ‘ . Lower eocene.

The lowest group of the Sind tertiaries, which lies with perfect con-
formity on the Cardzta beaumonti beds described in the last chapter,
derives its name from a hill fortress of Sind Amirs, situated in the Lakhi
range of hills, known as Ranikot and also as Mohan-kot, from the Mohan
stream, which traverses the fortification.1 The Rdnikot group is much more
extensively developed in Sind than the underlying cretaceous beds, for
although it is confined to lower Sind, and although its base is only seen
in the Lakhi range, north of Rdnikot, its upper strata occupy a consider-
able tract of country.

All the lower portion of the Rdnikot group, including by far the
greater portion of the beds, consists of soft sandstones, shales and clays,
often richly coloured and variegated with brown and red tints. Gypsum
is of frequent occurrence. Some of the shales are highly carbonaceous
and occasionally sufficiently pyritous to be used in the manufacture of alum.
In one instance a bed of ceal (or lignite), nearly six feet thick, was found,
and a considerable quantity of the mineral extracted.? The quality was
poor, the coal decomposed rapidly and was liable to spontaneous com-
bustion owing to the quantity of iron pyrites present, whilst the deposit
was found to bea small patch, not extending more than about 100 yards
in any direction. The only fossils found in the lower portion of the
Rénikot group, with the exception of a few fragments of bone, have been
plants, some dicotyledonous leaves, hitherto not identified, being the most
important. All the Rdnikot beds, except towards the top of the group,
have the appearance of being of fresh water, and are probably of
fluviatile origin.

A variable portion of the group, however, towards the top, consists
of highly fossiliferous limestones, often light or dark brown in colour
interstratified with sandstones, shales, clays, and ferruginous bands, These
are the lowest beds in Sind containing a distinctly tertiary marine fauna.

: The brown limestones are well developed arour.d Lynyan, east of Band
Vero and north-west of Kotri, and throughout the area of Rdnikot beds
near Jerruck and Tatta In this part of the country there appears to be a
complete passage upwards into the overlying nummulitic limestone (Kir-
thar), but in the Lakhi range the upper marine beds of the Rdnikot group
are poorly represented or wanting, and it is evident that they were removed
by denudation before the deposition of the Kirthar limestone, for the
latter is seen at Hothian pass resting upon their denuded edges.

The greatest thickness of the RAnikot group in the Lakhi range, where
alone the base of the group is visible, is about 2,000 feet, but generally

1 Memoirs, XVII, 37, (1879). l * Memoirs, V1, 13, (1869).
304 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [(hap, XITl,

the amount is rather less, about 1,500. It must, however, be recollected
that in this locality some of the upper marine beds are wanting,
and as these marine limestones and their intercalated shales, sand-
stones, etc., are 700 or 800 feet thick in places north-west of Kotri, it is
evident that the original development of the group exceeded the
2,000 feet seen in the Lakhi range.

The fossils of the RAnikot group} indicate a lower eocene age, though
cretaceous affinities are not wanting. The Naué7i7 are all connected with
cretaceous, rather than tertiary, types, a Zeredratuda is undistinguishable
from 7. subrotunda, one of the commonest upper mesozoic types. On the
other hand, the presence of nummulites, and the general aspect of the
mollusca, indicate a lower eocene age. The only fossils which have been
critically examined are the corals and echinoderms. In neither case is the
result decisive, but the corals comprise, out of a total of 50 species, 7
species identical with European eocene species and 5 closely allied to
forms found on that horizon or in slightly newer rocks.

In the Mari hills of eastern Baltchistdn the base of the strata regarded
as nummulitic is marked throughout by a peculiar pseudo-conglomerate
which has also been recognised in the G4j river in Sind on the one hand
and in the southern portion of the Suldimdn range on the other. It has
the appearance of being composed of subangular fragments of dark grey
limestone, imbedded in a limestone matrix of paler colour, both matrix
and pebbles containing numerous small nummulites, though no difference
can be traced between the forms found in the two portions of the rock.
The resemblance of this rock to a true conglomerate is especially striking
in the sections near Khattan, but the similarity of the fossils found in the
apparent pebbles and in the matrix, the comparative uniformity in
thickness of this band, which lies among fine grained shales, the absence
of any known rock from which the pebbles could be derived, and the
presence of every gradation from the most conglomerate like form to a
merely mottled limestone, all point to the structure being in some way ot
concretionary origin.*

Above the pseudo-conglomerate, which has been accepted as the base of
the tertiaries, there comes a great thickness of green and grey shales with
interbedded impure sandstones which are, as regards both their litho-
logy and stratigraphical position, the equivalent of the Ranfkot group in
Sind, but owing to the fossils not having been examined and the homo-
taxis verified, they have as yet been provisionally described as the Ghdzij
group.’ Along the outcrop of this group, from Mach in the Bolan pass to

1 A detaile1 list will be found in Memoirs, | ? W.T. Blanford, Memoirs, XX, 149, (1883) ;
XVII, 197, (1879). The corals and echinoderms | R. D. Oldham, Records, XXIII, 94, (1890).

are described in Pal. Indica, series xiv, I,] % Records, XXIII, 93, (1890).
pts. 2 and 3, (1880-36).
Chap. XIII.) KIRTHAR GROUP IN SIND. 305

Hurnai, coal seams are found near top, which attain a maximum thickness
of about three feet and have proved of great economic importance in a
country where fuel isso scarce. The distribution of the coal seams is
peculiar. Besides the localities mentioned, coal has been found north of
the Thal Chotiali plain and in the Luni Pathdn country to the east, in
every case close to the western limit of the known exposures of this group,
while to the eastwards the group ceases to be coal bearing. As the coal
seams were doubtless formed in marshes near the margin of the sea, and
as the only rocks known westwards of the present limit of the group are
either older or very much newer, it would seem that the original western
limit of deposition cannot have been far removed from the present limit
of outcrop, at any rate, in the country east of Quetta.

Further north, in the Suldimdn range, very similar shales are found
immediately overlying the upper cretaceous sandstone, but they differ from
what is seen in the southern sections in the prevalence of a red colour
throughout the greater portion of the thickness of the shales.

In both these areas the relation of the lowest tertiary to the underlying
beds is one of perfect conformity, and they form part of a continuous
system of deposits with the upper cretaceous beds, as has been mentioned
in the last chapter.

The Rdnfkot group in Sind is overlaid by the Kirthar group, so called
from the frontier range of hills of that name. Though inferior in thick-
ness to several other subdivisions of the tertiary series in Sind, this group
comprises by far the most conspicuous rock, the massive nummulitic lime-
stone which forms all thé higher ranges in Sind. It forms the crest of the
‘Kirthar range throughout, and all the higher portions of the Lakhi range,
of the Bhit range south-west of Manchhar lake, and of several smaller
ridges, and consists of a mass of limestone, varying in thickness from a
few hundred feet in lower Sind to about 1,000 or 1,200 at the Gj river,
and probably 2,000, or even 3,000, farther north. The colour is usually pale,
either white or grey, sometimes, but less frequently, dark grey, the texture
varying from hard, close, and homogeneous, breaking with a conchoidal
fracture, to soft, coarse and open. Ordinarily the nummulitic limestone
is tolerably compact but not crystalline, and chiefly composed of Fora-
minifera, especially whole or fragmentary Nummulites ; corals, echino-
derms, and molluscs also abound, but the two latter very frequently only
weather out as casts,

Throughout northern Sind, except near Rohri, no beds are seen be-
neath the Kirthar limestone. The remarkable range of low hills, surround-
ed by Indus alluvium, and extending for more than forty miles south from
Rohri, consists of nummulitic limestone, having a low dip to the westward,
and a considerable thickness of pale green gypseous clays, with a few bands
306 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap, XIIL -

of impure dark limestone and calcareous shale, is exposed benesth the lime-
stone forming the eastern scarp of the hills, on the edge of the alluvial
plain. No Foraminifera have been found in these beds, although Num-
multtes abound in the limestone immediately overlying. Several speciés of
mollusca occur, but none are characteristic, and it is far from clear whether
the green clays and their associates are merely thick bands intercalated in
the limestone, or whether they belong to a lower group. Probably these
argillaceous beds of the Rohri hills represent some of the marls, shales
and clays forming the lower portion of the upper Kirthar group on the
G4j river.

In some places west of Kotri a band of argillaceous and ferruginous
rock is found close to the base of the Kirthar group. It is mainly composed
of brown hematite, weathers into laterite and appears to be found over a
considerable area near Kotri and Jerruck. 2

It has already been mentioned that the Kirthar limestone rests uncon-
formably on the Rdnikot group in the Lakhi range. The Kirthar group
here cannot be much more than 500 or 600 feet thick, and consists en-
tirely of limestone. To the south-east, towards Kotri and Tatta, there is
no unconformity between the Rdnfkot and Kirthar groups; on the
contrary there is an almost complete passage between the two, and the
limestone of the latter becomes much split up and intercalated with shales
and sandy beds. This is even more the case further to the south-east in
Cutch, where the whole group consists of comparatively thin beds of
limestone, interstratified with shales. To the south-west the massive
limestone dies out altogether, and although it is well developed in the
southernmost extremity of the Kirthar range near Karchat, about 59
miles south of Sehwdn, it disappears entirely within a distance of
12 or 14 miles, and is entirely replaced by shaly limestones, shales,
and thick beds of sandstone in the ranges on the Hab river. Some
rather massive beds of nummuliferous dark grey limestone, very
different in character from the pale coloured Kirthar limestone, are
found west of the Hab, but their precise position in the series is not
known, and the rocks appearing from beneath the Nari group, in the
place of the Kirthar limestone, consist of shales and sandstones, with
some calcareous bands abounding in nummulites, and closely re-
sembling, both in character and in the species of Foraminifera they
contain, the nummulitic shales beneath the massive limestone on the
Gaj river,

The most characteristic fossils of the Kirthar group are Nummuilttes and
Alveolina, the extraordinary abundance of individuals rendering it usually
easy to recognise even small fragments of the rock by the organisms
preserved in it. Many of the species, and especially the Foraminifera,
are characteristically eocene, and there can be no question that the
Chap. XIII] KIRTHAR GROUP IN BALUCHISTAN. 307

nummulitic limestone of India is a continuation of the same formation
in Europe. Several species pass from the Rdnfkot beds into the Kirthar
group; indeed, the palzontological differences between the two appear
to be principally due to a change of conditions from the shallow muddy
water of the RAnfkot to the deeper clear sea of the Kirthar beds.

The result of Messrs. Duncan and Sladen’s examination of the
echinoderms does not altogether bear out the conclusions regarding the
relation between the Kirthar and Rdnikot groups expressed above. They
found no less than 63 out of 70 species being peculiar to the group, and
the horizon of the remaining 7, which are supposed to have been obtained
from the Rdnikot group, is very doubtful! Caution is, however, neces-
sary in applying the palzontological results obtained from a single order
of animals, and the conclusions based on the general palaontology and
stratigraphy of the two groups may be accepted in spite of this apparent
contradiction.

In eastern Balichistén the Kirthar limestone appears to be largely de-
veloped in the Mari hills and south of the Bolan pass, having been
given the local name of Spintangi in this region, Between Hurnai and
Quetta it has been very much reduced in thickness by denudation previous
to the deposition of the Siwdliks, The relation of the Kirthar, or Spin-
tangi, limestone to the underlying shales is one of perfect conformity by
interstratification, and there is reason to believe that to a certain extent
the Spintangi and Ghazij, or Kirthar and Ranikot, groups merely represent
different conditions of deposition and are partly of contemporaneous origin.

The Spintangi limestone has frequently a nodular structure that makes
it weather into an aggregate of rounded lumps, easily mistaken fora conglo-
merate ; so much so that three practised geologists have each recorded the
fact that, after crossing the boundary of the Siwdlik conglomerate in the
Bolan pass, they walked for some distance over the nummulitic limestone
before discovering the change, *

Another peculiarity of the group in the Mari hills is the occurrence of
thick beds of gypsum, interstratified with the clear limestones and green
shales. The thickness of these beds in the country east of Khattan is
very considerable, one bed of 50 feet thick, besides four others aggre-
gating 33, having been seen near Mamand.? Whether they were originally
deposited as gypsum, or are due to the subsequent alteration of limestone
beds, their occurrence among distinctly marine beds is not easy to account
for. On the east flank of the Suldiman range the Kirthar group does not
appear to be present in its characteristic form, but some thin beds of white
nummulitic limestone overlying the shales regarded as the probable
equivalent of the Ranikot group may represent it.

1 Pal, Indica, series xiv, I, pt.3, p. 245, | '? Reeerds, XXV, 24, (1892).
(1884).
308 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap, XIII.

In western RAjputdna two outcrops of nummulitic rocks are known,
the larger one north-west of Jaisalmer, the smaller near Koilath, thirty miles
west-south-west of Bikaner. The rocks represented are a white num-
muliferous limestone, resembling that of the Kirthar group of Sind, and
shaly beds, mostly grey and impregnated with salt, though a very fine
grained pale buff coloured fuller’s earth is also found and quarried for
export under the name of Multani mitti. In Jaisalmer a bed of ferru-
ginous lateritic rock, like that found near Kotri, is associated with the
nummulitics. The rocks of these exposures resemble thoze of the Kir-
thar group as seen east of Sukkur, and there is good reason to suppose
that they are of the same age and indicate an easterly extension of the
nummulitic sea,?

The series of tertiary rocks above the Kirthar nummulitic limestone is
superbly developed and very well seen in the hills on the frontier of upper
Sird whose culminating ridge is known as tle Kirthar. The names of
the tertiary groups overlying the nummulitic formation have consequently
been derived from places in this range, and the Nari group takes its title
frorm a stream which traverses the lower portions of the range, where
it is composed almost entirely of Nari beds, for a considerable distance,
and issues from the chills west by north of Sehwan.? ‘The present sub-
division comprises at its base the uppermost bands of limestone contain-
ing Nummulites, the species N. garansensis and WN. sublevigata
being distinct from those so commonly found in the Kirthar sub-
division, and the limestone itself is usually distinguished by its yellowish
brown colour, and by being in comparatively thin bands, interstra-
tied with shales and sandstones. Several other fossils, besides the
nummulites, differ from those in the Kirthar beds. Not unfrequently,
however, there is an apparent passage from the white or greyish white
Kirthar limestone into the yellow or brown Nari rock, and the two groups
appear always to be perfectly conformable, but no intermixture of the
characteristic species of nummulites has been detected, and the division
between the Kirthar and Nari beds can always be recognised by the fossil
evidence.

In some places the lower Nari beds consist almost entirely of brown
and yellow limestones, but more frequently the limestone bands are
subordinate, dark shales and brown, rather thinly bedded, sandstone
forming the mass of the rocks. The limestone bands are often confined
to the base of the group, and always diminish in abundance and thickness
above, although they are occasionally found as muchas 1,500 feet above the
top of the Kirthar group. ‘These shales and fine sandstones, with occasional
bands of limestone, constitute the Jower Naris, and pass gradually into

' Records, XIX, 159, (1886). | 2 Memoirs, XVII, 49, (1879).
Chap. XID] NARI GROUP. 309

the coarser, massive, thick bedded sandstones that form the greater portion
of the group, and attain a thickness of 4,000 or 5,000 feet. On the flanks
of the Kirthar range a few bands of clay, shale, or ironstone, are inter-
stratified with the sandstones, and bands of conglomerate occasionally
occur. The Nari beds in their typical form extend throughout the eastern
flank of the Kirthar range, and occupy a belt of varying width, from one
or two miles to as much as ten miles in breadth, between he underlying
Kirthar and the overlying Gaj beds.

On the western side of the Bhagotoro hills, four or five miles south of
Sehwan, there is a break in the Nari beds, and some variegated shales,
clays, and sandstones, richly tinted in parts with brown and red, which
represent the ypper Nari sandstones, rest unconformably on the denuded
edges of the lower Nari limestones and shales. The break is evidently
local. To the east of the Lakhirange the Nari bedsare entirely wanting,
and it appears very possible that they have never been deposited in this
portion of the Indus valley. From the neighbourhood of Sehwan to
Jerruck, the Manchhar beds rest with more or less unconformity on the
Kirthar, a very faint and imperfect representative of the G4j group
occasionally intervening. But west of the Lakhi range, throughout lower
Sind, the Nari beds are exposed almost wherever the base of the Gaj .
group is seen ; they increase in thickness to the westward, and the Hab
valley, from the spot where the river first forms the boundary of British
territory to the sea, consists entirely of these strata. There is, however,
no longer any such marked distinction between the subdivisions of the
tertiary series as is found in the Kirthar range. The disappearance of
the Kirthar limestone has already been mentioned, and with it the lower
Nari limestones also disappear, so that it is no longer possible to draw a
distinct line between the two groups. The two groups can still be traced,
although the dividing line between them is obscured, as the calcareous
shales, with the characteristic Kirthar nummulites below, and the mas-
sive Nari sandstones above, are still recognisable. Beds of brown
limestone, too, full of Ord¢toides papyracea or O. fortisi, occur in the Nari
beds of the Hab valley, but instead of being found at the base, they
appear in the middle of the group. Again, just as there is a difficulty in
distinguishing the Naris from the Kirthars at’ their base, so the beds
at the top of the former group can only be separated by an arbitrary line
from the overlying Gdj beds. In the Kirthar range the upper boundary
of the Nari group, although there is no unconformity, is distinct and defi-
nite, limestones with marine fossils of the Gaj group resting immediately
upon the upper Nari sandstones, But in southern Sind bands of limestones
‘or calcareous sandstone, with marine fossils, some of which are well
marked G4j species, occur in the upper part of the Nari group, whilst
limestone bands with the Nari Orbitocdes papyracea are found in the
Gdj.

x2
310 GEOLOGY OF INDIA--EXTRA HIMALAYAN TERTIARY. [Chap, XIII,

The sandstones, which form so large a portion of the Nari group
in upper Sind, have hitherto proved destitute of animal remains, but
the occasional interstratifications of shales and clays often contain
fragments of plants, and some ill marked impressions, probably due to
fucoids, have been found in the sandstones themselves. ‘here appears
a probability that these sandstones maybe of fluviatile, and not marine
origin, and although some species pass from the Mirthar, and even
from the Rdnikot, group into the lower part of the Nari group, the
fauna is chiefly distinct and marks a higher liorizon. The most
marked change is in the Foraminifera, because they are so abundant
and characteristic, whole beds of limestone towards the base of the
Nari group being entirely made up of three species, distinct from those
occurring in the Kirthar group, Vummulites garansensis, N. sublevigata
and Orditotdes papyracea, the last named frequently of large size, and
reaching two or three inches in diameter. One of these species of Num-
mulites (N. garansensis) is of importance, because it occurs in Europe,
as in Sind, in the highest strata characterised by the abundance of the
genus, those beds being at the base of the miocene. Nummulites
sublevigata is peculiar, so far as is known, to India. .

Several of the molluscs and echinoderms of the Nari beds also, such
as Siliquaria granti, Solarium affine, Venus granosa, and Clypeaster
profundus, show distinctly miocene affinities, and some of these pass up
into the Gdj group. At the same time there are so many eocene forms
present, such as Natica patula, N. sigaretina, Ostrea flabeliula, Voluta
jugosa, etc., that it is somewhat difficult to decide to which subdivision
the Nari beds should be assigned. They probably occupy an intermediate
position, corresponding to the oligocene of continental geologists, a con-
clusion which is borne out by the detailed examination of the corals and
echinoderms.

The lower ‘Nari limestone is found in its typical form as far north
as Bibi Nani in the Bolan pass, where it is overlaid by some grey
sandstones and mottled beds, which probably represent the upper Nari
of Sind. The lower Nari limestone, with Nummulites garansensts, N.
sublzvigata and Orbitordes papyracea, has not been found north of this,
but to the north of the Gandahdti hill, 20 miles east of Dera Bugti, and
along the eastern slopes of the Suldimdn range, a series of sandstones, with
subordinate bands of conglomerate and clay, occupy a position intermediate
between the upper eocene and the overlying Siwaliks. They are described
as apparently conformable to both, and were regarded by Dr. Blanford as
probably the equivalents of the upper Nari of Sind.

Resting upon the Nari group, almost throughout Sind, and forming
the base of the upper tertiary series, there is found a mass of highly
Chap. XIII. J GAJ GROUP. Bur

fossiliferous limestones and calcareous beds, usually more or less shaly,
always distinctly stratified, and easily distinguished from the limestones
of the older tertiary formations by the absence of nummulites. A superb
section of the strata forming this group is exposed on the banks of the
GAj river, from which its name is derived.!

On the eastern flanks of the Kirthar range in upper Sind, the G4j
group forms'a conspicuous ridge, the hard dark brown limestone bands
near the base of the formation resisting the action of denudation far more
than the soft sandstones of the Nari beds, and rising every here and there
into peaks of 1,c00 and 1,500 feet, or even more, scarped to the
westward and sloping to the east, Amru, the highest summit of the Géj
ridge, being 2,700 feet above the sea. Still, the limestone bands, although
so conspicuous, are subordinate, the greater part of the group consisting
of sandy shales, clays with gypsum, and sandstones towards the base.
Many of the bands of limestone appear very constant in position and may
be traced fora long distance; they are dark brown in colour as a rule,
but one bed is white and abounds in corals and small Foraminifera
(Orbttordes), whilst some of the darker bands contain Echinodermata in
large quantities,

The uppermost portion of the group is usually argillaceous, being
chiefly composed of red and olive clays with white gypsum, and these beds
pass gradually into precisely similar strata belonging to the overlying
Manchhar group. ‘Ihe< passage beds contain Corbula trigonalis, Lucina
(Diplodonta) incerta Tellina subdonacialis, Arca larkanensis, amongst
ether fossils, such as Zurriteléa angulata, and forms of Ostrea and
Placuna.

All of these have allies living in estuaries at the present day, Arca
granosa, a recent representative of A. darkanensts, being one of the com-
monest and most typical of Indian estuarine mollusca. To these estuarine
passage beds further reference will be made presently when the relations
of the Manchhar to the G&j beds are discussed.

The G&j beds at the Gdj river are very nearly 1,500 feet thick, but
they appear to be less developed to the northward in the Kirthar range,
and not to be much more than half the thickness named west of Larkhdna,
where, however, they are nearly vertical and have probably suffered from
pressure. In lower Sind, the G4j group, like the Nari, disappears to the
eastward of the Lakhi range, where it is either entirely wanting, or
else represented by a thin band at the base of the Manchhar group,
containing one of the characteristic fossils, Ostrea multicostata. There
is, however, a very large area of Gdj beds north and north-east of
Karachi, and the appearance of the formation here is somewhat different
from what it is in the Kirthar range, for the greater portion of the group

> Memoirs, XVU, 83, (1879).
312 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap, XIII,

consists of pale coloured limestones, almost horizontal or dipping at very low
angles, and forming plateaux 400 or 500 feet high, bounded by steep
scarps, which rise from the low ground of the soft Nari sandstones east of
the Hab valley. A low range of kills, formed of Gaj beds, extends to the
south-west, past the hot spring at Pir Mangho (Mugger or Manga Pir) to
the end of the promontory known as Cape Monze, and the same beds form
the low hills east and north-east of Kardchi, and furnish the materials of
which the houses of the town are mostly built. A smallisland called
Churna, in the sea west of Cape Monze, also consists of Gaj rocks, To
the northward the Gaj area of lower Sind extends with very irregular
outline to the neighbourhood of Tong and Karchat, almost due west of
Hala, and there are several outliers farther north, connecting the southern
portion of the group with the typical outcrop in the Kirthar range.
East of Kardchi, Gaj beds extend in the direction of Tatta, until they
disappear with the other tertiary rocks beneath the alluvium of the
Indus. It is quite possible that the present group, as well as the Nari,
never was deposited in the neighbourhood of Kotri and Jerruck.

It has been already stated that the Gaj beds, throughout the greater
portion of the Kirthar range, rest conforrnably upon the Nari group,
although there is a change in mineral character, and that, in lower Sind,
the passage from one group into the other is gradual, calcareous bands,
with Gaj fossils such as Ostrea multtcostata and Pecien subcorneus,
being found interstratified with the upperinest Nari sandstones. At one
place, however, near Tandra Rdhim Khan, west by north of Sehwan, the
outcrop of the G4j beds, here dipping ata high angle to the westward,
runs nearly in a straight line across the mouth of a valley, composed of a
deep synclinal of the Nari group between two anticlinal ridges of Kirthar
limestone. As the Gdj beds do not share the synclinal curve of the Naris,
it is difficult to see how the two can be conformable, but an examination
of the boundary between the two groups failed to show any clear evidence
of unconformity. There are, however, some places south of Sehwan
where the Gaj group overlaps the Nari beds and rests upon the Kirthar
limestone, but it must be recollected that the G4j group is itself overlapped
by Manchhar beds in the immediate neighbourhood.

The commonest and most characteristic fossils of this group are Ostrea
multicostata’ and Breynia carinata. There cannot be any question that
the Gaj fauna is newer than eocene ; some of the species are recent (for
instance, Dosinia pseudoargus is identical with the recent D. exasperata),
and it is probable that many others, when they are compared with recent
forms more carefully than has hitherto been done, will prove to be the

'It is not quite certain whether this | D’Archiac and Haime. There is another

species is identical with the European form, | species known by the same name and found
but it is c2rtainly the shell figured by Messrs. | in triassic beds in Europe.
Chap. XTIL J] MANCHHAR SERIES. 313

same as living species. Several genera, tuo, as Maret¢a, Breynia
Meoma, Echinodiscus Clypeaster, Cladocora, and Mycedium, are rare or
unknown in the older tertiaries, and there is almost a complete diappear-
ance of eocene forms, very few species being common even to the Nari
beds.

The only mammal yet obtained from the Gaj beds is R&duoceros siva-
lensis—a species found also in the Siwdliks,

The highest subdivision of the tertiary series in Sind was originally
named! from a large lake, the Manchhar, a few miles west of Sehwan, but
there can no longer be any doubt that in age, as well as mode of origin,
they are part of that system of pliocene fresh water deposits which, under
the name of Siwdlik, ranges round the extra-peninsular area from Sind
to Burma. It will be well, however, to retain the local name in the
description of the Sind Siwdliks.

The Manckhar series of Sind consists of clays, sandstones, and conglo-
merates, and attains a thickness of but little, if at all, less than 10.000 fect
in places on the flanks of the Kirthar range. Although it is difficult to
draw an absolute line between the subdivisions, the whole group may be
divided, wherever it is well exposed, into two portions. The lower consists
mainly of a characteristic grey sandstone, rather soft, moderately fine
grained, and composed of quartz, with some felspar and hornblende, to-
gether with red sandstones, conglomeratic beds, and, towards the base,
red, brown, and grey clays, the latter, however, being much less largely
developed than in the upper subdivision. The conglomeratic beds chiefly
contain nodules of clay and of soft sandstone, apparently derived from
beds precisely similar to those of the typical Siwdliks, but, so faras has
been observed, do not contain any fragments derived from the older tertiary
rocks, no pebbles either of the characteristic G4j limestones or of the
still more easily recognised nummulitic limestone of the Kirthars having
been noticed in the beds of the lower Manchhars, although both abound
in the upper strata of the group. These conglomeratic beds of the lower
Mancbhars are frequently ossiferous, the bones and teeth contained in
them being, however, usually isolated and fragmentary.

The upper Manchhar subdivision, where it is best seen on the flanks
of the Kirthar range west of Larkhdna, is thicker than the lower, and
consists towards the base of a great thickness of orange or brown
clays, with subordinate bands of sandstone and conglomerate. The
sandstones are usually light brown, but occasionally grey, like the charac-
teristic beds of the lower subdivision. The higher portion of this upper’
subgroup contains more sandstone and conglomerate, and the whole is

1 Memoirs, XVAl, §7, (1879).
314 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY, [Chap, XIIL

capped by a thick band of massive, coarse conglomerate, which forms a con-
spicuous ridge along the edge of the Indus alluvium throughout upper Sind.
This conglomerate contains numerous large pebbles of. nummulitic and
Gaj limestone, together with fragments of quartzite and other rocks of
unknown origin, Throughout the conglomeratic beds of the-upper Man-
chhars, pebbles of nummulitic limestone and of the brown G4j limestone
occur, showing that these older tertiary beds must have been upheaved-
and denuded in the latter Siwdlik period, although there is a complete
passage between the Gaj beds and the lower Manchhars.

There appears, however, good reason for supposing that some dis-
turbance of the older rocks took place before the deposition of the lower
portion of the Manchhargroup. To the east of the Lakhi range the Manchha,
beds, themselves dist urbed, rest unconformably on the beds of the Kirthar
group, which are vertical in many places, so that it is manifest that the
Kirthars had, in this locality, been upheaved before the deposition of the
Manchhars. The presence of the lower portion of the latter series appear-
ing to be proved by the occurrence of teeth and bones of the same mam-
mals as are found in the lower Manchhars elsewhere.

In one case a few estuarine fossils were found, near the Nari stream,
in a Manchhar bed 300 or 400 feet above the base of the group. The
only form recognised was Corbula trigonatzs, already mentioned a8 charac-
teristic of the estuarine passage beds between G4j and Manchhar. With
this exception, and that of some rolled oyster shells, possibly derived from
a lower formation, no marine or estuarine fossils have been observed in
the Manchhar beds of upper Sind, above the passage beds at the base of
the group, and there appears every reason to believe that the group is of
fluviatile origin.

In lower Sind, however, there is a very considerable intercalation of
marine or estuarine beds with the Manchhars, and this evidence of depo-
sition in salt water increases in the neighbourhood of the present coast.
Around Karachi beds of oysters, and sometimes of other marine or estu-
arine shells, are not unfrequently found interstratified with the Manchhar
group.. There is also some change in mineral character, the sandstones
becoming more argillaceous and associated in places with pale grey,
sandy clays and shales. The passage from the GAj beds is very gradual,
calcareous bands with Gdj fossils, such as Ostrea multicostata and Pecten
subcorneus, being found some distance above the base of the Manchhar
group.

Although there is no difficulty in drawing aline between Manchhar and
Gaj beds, except in the neighbourhood of the coast, everything tends to show
that there is no break in time between the two, the lower portion of the upper
group being an estuarine or fluviatile continuation of the underlying marine
beds. But the great thickness of the Manchhar group in upper Sind
 

Echinodiscus desori, Dune, and Slad.

 

Vicarya verneuili, a’Arch.

   
   

Pecten favrei, d’Arch,

Corbala trigonalis, Sow,

 

Arca larkhanensis, a’Arch.

Turritella angulata, Sow.

 

Arca kurracheensis, @’Arch. Ostrea multicostata, Desbayes.

MIOCENE FOSSILS.

Calcutta Phototype Co.
Chap. XIII. ] UPPER MANCHHAR GROUP. 315
alone would suffice to prove that a considerable period of time must have
elapsed during the deposition of this formation, and it is far from improb-
able that the lower Manchhars may be upper miocene, whilst the upper
Manchhars are pliocene.

The Manchhar beds extend along the age of the alluvium, and form a
broad fringe to the Kirthar range, throughout upper Sind from west of
Shikdrpur to the Manchhar lake, but the breadth of the outcrop varies
greatly, being as much as 14 miles where broadest west of Larkhdna, and
diminishing both to the north and south. As already noticed, the Man-
chhars are thickest just where their outcrop is widest ; the breadth of the area
occupied by them is not, however, due simply to their vertical development,
but chiefly to their forming a synclinal and anticlinal roll before disappearing
beneath the alluvial plain, whereas in other parts of the range the same
beds are exposed in a simple section, all the strata dipping to the east-
ward, To the north the section is complicated by faults, but to the south
the thickness of the Manchhar group diminishes greatly, and near Tandra
Rdhim Khan, west of Sehwan, although both upper and lower subdivisions
of the group are developed and the uppermost conglomerate is exposed, the
whole thickness of the Manchhar strata cannot be much more than about
3,000 feet. The Manchhar beds are seen west, south, and east of the
Manchbar lake ; they are well developed and occupy a large plain to the
east of the Lakhi range, and west of the nummulitic limestone tract, near
Kotri and Jerruck, they re-appear in many places in the different synclinal
valleys to the west of the Lakhi range, and they occupy a considerable tract
of country east and north-east of Kardchi. Throughout these areas in
lower Sind the rocks are not nearly so well seen as to the northward, the
soft sandstones and clays of the Manchhar group having been denuded into
undulating plains, covered and concealed in general by the pebbles and
sands derived from the comparatively hard older tertiary rocks of the
neighbouring hills, and it is far more difficult than in upper Sind to
distinguish the different portions of the group, or to form a correct idea
of the thickness of strata exposed.

The Manchhar beds extend along the edge of the sea west of Kardchi,
almost to the end of cape Monze, but no representative of this formation
is seen for a considerable distance to the westward of the cape. The
greater part of the country near the shores of Sonmiani bay consists of
alluvium, and the few exposures of rocks are older tertiary, or perhaps
cretaceous, a low cliff near the coast, north of Gadani, being apparently
composed of subrecent deposits. But west of Sonmidni bay, in the
neighbourhood of Hingl4j, a well known place of Hindu pilgrimage,
there are high hills of hard greyish white marls or clays, usually sandy
316 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap. XTIL

often highly calcareous, and occasionally intersected by veins of gypsum.
With this clay or marl, bands of shaly limestone, dark calcareous grit,
and sandstone are interstratified, usually forming but a small portion of
the mass, although their greater hardness renders them conspicuous. ‘This
marl formation extends for many hundreds of miles along the coasts of
Baldéchistén and of the Persian Gulf, forming great horizontal plateaux,
surrounded by cliffs of whitish marl or clay and capped by dark coloured
calcareous grit, at the headlands of Ras Mal4n, Ormdra, and Gwadar, Rds
Maldn especially being a table land rising abruptly to a height of 2,000
feet from the sea. ‘These remarkable rocks have been called the Mekran
group! from the name usually applied to the littoral tracts of Baluchistan.

The Mekrdn group is of shallow water marine origin, and abounds in
mollusca, echinoderms, etc., many of the species being apparently the same
as living forms. The echinoderms alone haveas yet been examined in detail,
they belong without exception to living genera, while most of the species
are very closely allied to recent forms, and one species alone is doubtfully
identical with a G4j form. The general facies of the fauna is distinctly
pliocene.?

Although there is no resemblance between the typical Manchhar beds
and the characteristic rocks of the Mekr4n group, nor, from the widely
different conditions under which the two formations must have been de-
posited, would any similarity in mineral character be probable, some of the
soft argillaceous shaly sands in the Manchhar beds near Karachi closely
resemble some similar beds in the Mekrdén group near Gwddar. As all
that is known of the geology of western Laltichistdn is the result of
brief visits to a few widely separated points, it is uncertain to what
extent the rocks of Sind extend to the westward, and whether any
representatives, of the G4j group especially, exist in that direction, but
it appears probable that the marine Mekrdn group in western Baldchistan
represents the fresh water Manchhar series of Sind.

The only fossil remains of any importance hitherto detected in the
Manchhar series are bones of mammalia, and all that have been recog-
nised belong to the lower Manchhars, the upper subdivision of the group
having hitherto furnished only a few bones, in too poor and fragmentary
a state of preservation for the species, or even the genera, to be deter-
mined. The few estuarine shells found in the lowest Manchhar beds in
upper Sind, and a portion at least of the marine fossils procured from a
similar horizon near Karachi, appear to be Gaj forms, and to indicate a
close connection between the lower Manchhars and the underlying group

2 W. T. Blanford, Records, V, 43, (1872); 2 Duncan and Sladen, Pal. Indica, series
Eastern Persia, London, 1886, II, p. 462. xiv, I, pt. iii, p. 370, (1885).
Chap. XIII. ] AGE OF THE MANCHHAR SERIES, 317
In places, and especially in the neighbourhood of the Lakhi range, silici-
fied fossil wood is found in abundance in the Manchhar beds, stems
of large trees being of common occurrence. The majority are dicotyle-
donous, but some fragments of menocotyledons are also found.

The vertebrate remains are extremely fragmentary, and chiefly consist
of single teeth and broken portions of bones,’ and the fauna is chiefly
remarkable for the prevalence of artiodactyle ungulates, allied to pigs, or
intermediate between pigs and ruminants. The majority of the genera
are extinct ; Rhinoceros, Sus, and Manis being the only living types, and
the last named has only been recognised from a single digital phalange,
so that the generic identification is far from sufficient. Both RAdnoceros
and Sas existed in miocene times, whilst Amphicyon, Anthracothertum,
Hyopotamus, and Dinotherium, which are also found in the Manchhar, are
not known to occur in Europe in beds of later date than miocene. The
genera Hemtmeryx and Stvameryx are peculiar, both being allied to the
Siwdlik Merycopotamus.

The species found also in the piiocene Siwdliks are Rhinoceros palein-
dicus, Acerothertum perimense, Chalicotherium sivalense, Sus hysudricus,
Dorcatherium majus, D. minus, Mastodon tatidens, and M. falconert;
but as the presence of these forms in the Manchhars is inferred for the
most part from fragments, the identifications are by no means quite
certain, whilst the general facies of the fauna, the absence of characteristic
living forms like Eguus, Bos, Antilope, Cervus, and Elephas, and the
presence of several extinct genera not hitherto detected in the Siwdliks,
show that the mammaliferous beds of Sind are of older age than the typi-
cal Siwdlik strata. It should be recollected, moreover, that the precise
horizon at which the Siwdlik forms are found is but rarely known with
accuracy, that some of the Siwdlik strata are as old as the lower Man-
chhars, if not older, and that a portion at least of the older types of mam-
mals are from beds low down in the Siwdlik series, None of the remarkable
series of types allied to the giraffes and Svathertum, nor of the peculiar
bovine and antilopine forms so characteristic of the Siwdlik fauna, have as
yet been found in Sind. The only ruminant detected in the Manchhar beds
is the miocene Dorcatherium, and the place of the more specialised
Pecora appears to have been occupied by the less specialised even toed
ungulates allied to the pig. While, therefore, it is probable that some
extinct types, such as Anthsacotherium afnd Hyopotamus, which are not
known in Europe above the lower miocene, existed in India at a somewhat
later period, together with species which survived till pliocene times, it is
evident that the lower portion of the Manchhar group can scarcely be
considered of later date than upper miocene. The paleontological

1 Details will be found in Mr. Lydekker’s | 225; XI, 64,71, 77, 79, (1876-78) and Pal.
papers in Records, 1X, 91, 93, 106; X, 76, 83, | Indica, series x, I—IV, passim.
318 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap, XIII,

evidence is in accordance with the geological, and both show the close
connection between the lowec Manchhar beds and the GA4j group.

Further north, in the country between Sibi and Quetta the Siwdliks
rest with a slight, but distinct, unconformity on the Spintangi limestone,
There is everywhere a perfect parallelism of stratification between the
two and a superficial appearance of conformity, but on a close examina-
tion it is seen that there are slight, though distinct, traces of erosion at the
contact, that pebbles of nummulitic limestone are found near the base of
the Siwdliks, and that there is a progressive thinning out of the nummulitic
limestone from east to west by removal of the upper beds. The lower
Siwéliks, which are conspicuous on the section in the Bugti hills further
east, are represented by greenish and grey sandstones which do not at-
tain a thickness of much over 100 or 200 feet. Above them the upper
Siwaliks consist principally of red eaithy clays, with interbedded sand-
stones, which become more and more frequent till it becomes a sandstone
formation with subsidiary bands of clay and conglomerate, the whole capped
by a great thickness of strong conglomerates.

These Siwdliks extend westwards, in the country intervening between
the Bolan and Hurnai routes to Quetta, almost to the Quetta plain, but
their original western extension has been obscured. In the valleys of
Quetta and Pishin there are some deposits of conglomerates and red clays
which have in places undergone considerable disturbance, and have been
referred to a Siwdlik age! On the accompanying map they have been
coloured as upper tertiary, but, as they appear to be distinct in age and
origin from the Siwdliks referred to above, and much more closely con-
nected with the recent deposits, their description will be deferred to a
subsequent chapter.?

In the Bolan pass, in the Gandak valley, north-east of Quetta, and
doubtless in many other valleys of these hills, there are sandstone and
conglomerate deposits which have undergone considerable disturbance,
and dip at high angles. In accordance with the custom, which separates
the deposits which have undergone considerable disturbance from the un-
disturbed recent deposits that unconformably overlie them, these have been
called Siwdlik, but they are in a manner distinct from the older Siwdliks
for the former were deposited after the valleys, in which they lie, had been
excavated, while the latter date, in this neighbourhood, from a time when
the disturbance of the strata and consequent elevation of the hills had not
commenced.

In the Bugti hills and southern portion of the Suldim4n range, the
lower Siwdliks, which attain a maximum thickness of 5,000 feet, have the
appearance of lying conformably on the nummulitic limestone. They

1W. T.Blanford, Memoirs, XX, 117, (1883).1 2 Infra, p. 416.
Chap. XI. 1 SIWALIKS OF BALOCHISTAN. 319

consist of moderately soft, fine grained, pepper and salt grey, sandstone,
interstratified with conglomeratic beds, composed of fragments of clay and
soft sandstone, apparently derived from contemporaneous deposits, im-
bedded in an argillaceous matrix and unaccompanied by any pebble of
harder rock. Clay beds also occur and are usually of a red colour.

In the lower beds of this group vertebrate remains occur in consider-
able abundance near Dera Bugti, among which are Mastodon, Rhinoceros,
Dinotherium, Anthracotherium, and Hyopotamus. The vertebrate fauna
has a distinctly miocene facies and is associated with seven species of
fluviatite mollusca, of which four belong to the genus Unio, two to Melania
and one to Paludina. All seven are extinct, and none are nearly related
to forms now living in western India, though two are allied to species
still existing in Burma. Three of the species of Unio are aberrant forms
with ribbed shells, exhibiting a superficial resemblance to the marine genus
Cardium

East of the Suldimdu range the upper Siwdlik conglomerate is on some
sections overlaid by a more recent conglomerate deposit, which has been
disturbed and dips towards the plain at moderate angles. The newer
conglomerate has the appearance of passing upwards into the recent
deposits, and though it has been regarded as uppermost Siwdlik *# should
probably, like the disturbed river gravels of the Bolan valley, be more
correctly classed as recent than as tertiary.

The various localities referred to in the previous passages form part of
one geological province throughout which the lower tertiaries maintain a
certain constancy of character, allowing the rock groups on one section
to be recognised on another, but when we pass northwards to the Punjab,
or north-westwards to Afghdnistdn, it is no longer possible to apply the
subdivisions adopted in Sind, and a fresh classification has to be adopted.
Before passing on to these areas it will be well to notice the exposures
to the east, in which the Sind subdivisions can be more or less recognised.

The tertiary rocks in the Cutch peninsula occupy a belt, varying in,
breadth from about four miles to twenty, between the alluvium near the
coast and the older rocks in the interior of the country. Tertiary
formations also fringe the Deccan traps and jurassic beds, on the borders
of the two openings by which the Rann of Cutch communicates with
the sea east and west of the province, and patches of the same tertiaries
are found here and there on the shores of the Rann, not only in the main
region of Cutch itself, but also around the detached hilly tracts or islands,
Patcham, Kharir, etc, and in Wdgad. The evidence of unconformity

*W. T. Blanford, Memoirs, XX, 162, 233, | ? W.T. Blanford, Memoirs, XX, 219, (1883).
(1883).
320 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [ Chap. XIII,

between the eocene rocks and the Deccan traps is very strong in Cutch,
The lava flows which appear to have covered the greater part, if not the
whole, of the jurassic region had been completely swept away from the
surface of the country, and the underlying jurassic rocks exposed and
largely eroded in places before the eocene marine beds were deposited.!
Despite this evidence of unconformity, there is every appearance, along
the southern border of the trap area, of the tertiary beds resting conform-
ably on the lava flows of the Deccan period.

The tertiaries of Cutch are far better known than those of Gujdrdt and
KAthidwdr, the materials for the first descriptions of marine fossils from
the later Indian deposits having been furnished by the rocks of the
first named province. Attention was first directed to the Cutch tertiaries
through the labours of Captain Grant, who carried with him to England
a considerable collection of tertiary organic remains, together with the
jurassic fossils mentioned in a former chapter. In accordance with the
ideas prevailing amongst geologists at the time, he separated nummulitic
rocks from the true tertiaries on his map,? and the same distinction was
preserved in the description of the fossils, but subsequently all the forms
described were classed as eocene by D’Archiac and Haime.? When the
rocks of Cutch were mapped in 1867-69 by Wynne and Fedden, and
described by the former, it was found that several distinct groups
could be recognised, and that the fossils of these groups differed, and it
was afterwards discovered that the groups corresponded very closely to
those determined in Sind. The succession of the rocks in Cutch, accord.
ing to Mr. Wynne, is the following, the probable Sind representative being
appended in each case. The supposed European equivalents differ some-
what from those originally suggested before the corresponding beds in
Sind had been examined *:—

Cutch, Sind, European equivalents,
Alluvium, blown sand, etc. . ‘ é - Alluvium, etc. Pezstocene ani recent
j Upper tertiary 200 to 500 ft. Manchhar . /liocene and upper
(unconformity). miucene,
Argillaceous group 800 to1,200,, GAj . « Miocene.
Arenaceous grou 1 ‘ari (? 2 mivce 9
ere group 30,, Nari (?) — . Lower miocene an! upper
7 eocene,
Nummulitic group , - 700,, Kirthar  , Eocene
G
Gypseous shales + 100 4, ps
; ik ‘ ;
Subnuniauhite : y too.) Ranikot Lower eocene
Stratified traps . . . FI ‘ - Trap , - Uppe: most c: etaceous.

1 This view is opposed to Mr. Wynne’s; ? Gel. Soc. Trans. 2nd series., V, 302, Pl.
opinion. He considered that the lowér eoceue | xx, (1840).
beds are conformable to the traps, and that | 3 Description des Animaux Fossiles du groupe
the traps never existed in northern Cutch. nummulitique de l’Inde, Paris, 1853.
— Memoirs, 1X, 72, (1872). 4 Memoirs, 1X, 48, (:872).
Chap. XIII. J LOWER EOCENE UF CUTCH. qat

The subnummulitic group consists chiefly of peculiar purple and red,
mottled with white, soft argillaceous beds, laterite of various kinds, and
coarse sandstones distinguished by brilliancy of colouring, white, red,
lavender, purple, and orange tints prevailing. There are also some shales
with impressions of leaves and carbonaceous layers, and occasionally with
gypsum.

Some of the peculiar argillaceous beds have a distinctly volcanic aspect,
but as they are much decomposed it is impossible to say that they are
really of eruptive origin. ‘lhe occurrence of these peculiar beds away from
the traps, in places where there is good reason to suppose that the traps
were removed by denudation in pretertiary times, and the fact that beds
reconsolidated from trap fragmeits must, when decomposed, frequently
be undistinguishable from a disintegrated eruptive rock, render it probable
that these soft mottled beds are of sedimentary origin and composed of
the detritus of volcanic rocks, Fossils are rare in the subnummulitic group
which extends along the southern edge of the traps in Cutch, overlapping
the volcanic rocks to the westward, and resting upon jurassic rocks
near Lakhpat. The same group is represented in several small patches,
deposited upon jurassic beds on the borders of the Rann, both on the
mainland of Cutch and on some of the detached hills or islands, especially
south of the hills in Patcham, Kharir, Bela, and Chorar, and intervening
in the hollow between two ranges on the first named. The group is
nowhere more than about 200 feet thick, and it frequently does not exceed
20 feet.

Above the subnummulitic beds there are in places from 50 to 100 feet
of fine laminated shales, bituminous and often pyritous, with fragments
of wood and leaf impressions. Allthe above rocks are classed by Mr,
Wynne apart from the true tertiaries, and with the bedded traps. It
appears, however, more probably correct, and more in accordance with the
sequence in Sind, to consider the main break in the series as taking place
between the traps and the next formation in ascending order.

The gypseous shales form a local and unimportant subdivision, not
more than from 50 to 150 feet in thickness, occurring in western Cutch,
round the Gaira hills and in a few other places. ‘They consist of shales,
with calcareous nodular bands and much gypsum, and with some beds
of laterite. Same of the marly beds abound in Nummulites and other
foraminifera oysters, etc,

The next group is of more importance, being the representative of
the massive nummulitic (Kirthar) limestone of Sind. In Cutch these beds
consist of pale yellow and white impure limestones, in bands of no great
thickness, interstratifed with marls and sandy beds. The upper portion
consists chiefly of marls, limestones being more abundant below ; Nummu-
dites, Alveoling, and echinoderms of several kinds abound and corals
322 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Qhap. XIII.

and mollusca are locally common. The nummulitics of Cutch are, however,
almost confined to the western part of the province, and occupy a band
extending from Lakhpat round the western termination of the Deccan
trap range in the Gaira hills.

Upon the nummulitic limestones and their associates there is usually
found a thin and unimportant band of, light coloured or white sand and
sandy shales, having at the base some finer dun, or blue coloured, silty
shales, These sandy beds are soft, friable, and obliquely laminated. In
the lower portion of the group the carapace of a small crab and casts of
bivalve shells have been found, in the upper part impressions of dicoty-
ledonous leaves occur. This group, originally described as the arenaceous
group, corresponds in mineral character and position to the upper Nati
of western Sind,

The Gaj group of Sind is represented in Cutch by what was originally
described as the argillaceous group, the best developed and most fossili-
ferous of the tertiary rock groups of that district, and it is this group
which yielded the bulk of the fossils described as tertiary in the ap-
pendix to Captain Grant’s paper, although it appears probable that there
were among these fossils some admixture of species from a lower horizon,
Until the whole of the Cutch and Sind fossils are thoroughly compared
and determined, some doubt must remain as to the original horizon of a
few Cutch species, but when the forms are common to the GAj beds of
Sind, and are not known to occur in older group of that area, it may fairly
be inferred that they are probably restricted to the same horizon in Cutch.

The GAj, or miocene, rocks of Cutch consist of sandstones at the base,
with a few nodular, marly and ferruginous bands often containing Zurri-
tella, Venus granosa, and Corbula, Above the sandy beds are marly
limestones and shales, next calcareous grits, and then a considerable
thickness of shales, clays, and marls. The most fossiliferous beds are the
marly limestones and shales. Only the echinoderms from these beds have
as yet been critically examined, and of 16 species no less than 8 are also
known from the Gaj group of Sind, one being Breynia carinata, one of the
most characteristic G4j forms.

The miocene beds are more extensively developed in Cutch than are
the nummulitics. They are found not only in the west of the province
around the extremity of the jurassic and trap area, but eastwards, resting
upon the subnummulitic group, as far as about half way across the province.
To the westward, however, the present group is overlapped by the next
in ascending order. .

The representative of the Siwdlik rocks in the sub-Himaldyan
tract, and of the Manchhar beds in Sind, appears to be widely developed
in Cutch, and covers a large area, but it is very ill seen, being greatly con-
cealed by alluvial deposits. The principal beds are more or less ferruginous
5

Chap. XII] KATHIAWAR, ‘iss
conglomerate, at the base, followed in ascending order by thick brown
sands and obliquely laminated, nodular, calcareous, and sandy clays. Marine
beds, with large oysters, are intercalated, as in southern Sind. It will pro-
bably be found on further examination that this uppermost tertiary group
in Cutch, as in Sind,.passes down into the underlying subdivision in
places, although to the eastward the latter appears to have been denuded
befure the deposition of the former. The upper tertiary group extends
throughout southern Cutch from east to west, resting on the older tertiaries
tothe westward, but gradually overlapping them and the traps to the east-
ward, and resting upon jurassic rocks in the extreme east of the province,

In Kathidwdr eocene beds have been recorded from Beyt island, off the
north-west extremity of the peninsula, by Dr. Carter,! and though the
Geological Survey of Kathidwdr failed to show the existence of any lower
tertiary beds on the mainland, it is impossible to accept Mr. Fedden’s
suggestion” that the record of the presence of nummulitic beds is due toa
confusion between Patellina and Nummulina, Dr. Carter’s statements
are too specific to allow of such an explanation, and unless he was mis-
informed as to the locality from which his specimens came, we May accept |
the presence of lower tertiary beds at the north-west extremity of the
Kathidwdr peninsula.

With this exception the only known tertiary deposits are miocene or
newer. They are found along the southern edge of the trap area, the
principal exposures being at the eastern and western extremities, with
some narrow strips in the intervening area. They lie almost horizontal, are
much obscured by recent deposits and cultivation, and in the absence of
any deep cut sections no good general succession has been made out. In
the south-eastern area&hey consist of shales and marls, which contain many
marine fossils, with interbedded bands of a rusty conglomerate of clay
pellets, and agates derived from the trap.

With the exception of the echinoderms, the marine fossils collected in K4-
thidwdr have not been examined in detail, but many of them are identical
with Gj species, and of the 13 species of Echinodermata, 6 are also found
in the Gdj groups of Sind, the whole distinctly indicating a miocene age.’

The uppermost beds of the Kathidwdr tertiaries appear to be sand-
stones and conglomerates, of which the best known exposure is that in the
small island of Perim or Piram—not to be confounded with the island
of similar name at the entrance to the Red Sea. Perim island is a small
reef of sandstones and conglomerates, only 1,800 yards long by 300 to
500 yards broad, which first achieved a geological celebrity through the

1 Geological Papers on Western India. Bom- | ® Duncan and Sladen, Pal. Indica, series

bay, 1857, p. 743 xiv, 1, pt. 4, p. 80, (1880),
9 Memoiis, XX1, 122, (1884).
324 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap. XII.

discovery of fossil mammalian bones by the Baron von Higel in 1836. .
According to the most recent examination these bones are found in the
conglomerate bands, but principally in a conglomerate bed which lies
considerably below high water level and is obscured by a thick covering of
mud for the greater part of the year. During the months of April, May
and June, however, the south-east end of the reef becomes scoured and
free from mud, and specimens of fossil bones can be found. As in all cther
Indian localities for fossil bones, the first collectors found a rich harvest
in the accumulation of ages, while their successors have to be content with
but occasional and fragmentary specimens.

The known mammals from Perim island comprise ten species, of which
four, Mastodon latidens, M. perimensts, Rhinoceros perimensis, and
Sus hysudrtcus, are common to the Siwdlik beds. All these forms are,
however, found in other fossil faunas; Mastodon perimensis and Sus
hysudricus being met with in the lower Manchhar beds of Sind, Riz-
noceros perimensis in the Irawadi deposits, and Mastodon latidens in
both, so that all the forms common to Perim island and the Siwdliks
are clearly species of wide range. The absence of Zvepzas and its sub-
genera, and of bovines, and the presence of Dinotiertum, tend strongly
to make the Perim island fauna appear of greater age than the Siwdlik
generally, but, on the otker hand, the presence of so highly specialised
a genus as Capra, if the generic determination be accepted,’ the occur-
rence of Camelopardalts and Antilope, and, above all, the absence, so far
as is known, of Anthracotherium, Hyopotamus and other older ungulate
types so abundant in the miocene beds of Sind and the Punjab, are op-
posed to the idea that the Perim island rocks can be of higher antiquity
than pliocene. They possibly occupy an intermediate position between
the Siwdliks proper and the Manchhars of Sind, but they are more nearly
allied to the former,

In the north-western portion of the Kdathidwdr peninsula the fossili-
terous Ga] beds are overlaid by what have been described as the Dwarka
beds, They consist of soft yellow earthy or marly clays, gypseous in part,
overlaid by more or less marly or arenaceous limestones, generally soft
and _ porous, largely composed of Foraminifera cemented by calcite, or of
comminuted shells and corals. No recognisable fossils have been found
in these beds, and their relation to the underlying fossiliferous miocene
beds was not determinable; it is probably one of conformity in spite of
the sudden change olf lithological character,

'F, Fedden, Memoirs, ¥XT, 39, (1884) ; see | with the horn-cases, and that nothing is known
also W. T. Blanford, Memoirs, V1, 374, (1869). | of the greater part of the cranium, the teeth

2 It should not be forgotten that Capra | or the limb bones. See Lydekker, Pal. Indica,
pevimensis is founded solely on a frontlet / series x, 1, 83, 170, Pl. xxviii, fig. 4, (1884).
Chap. XIIL] AFGHANISTAN. 325

Turning to Afghdn-Turkistén, there appears to be, as in Sind, a perfect
conformity between the cretaceous and tertiary beds. In the synclinal of
Mathar, too miles south-east of Balkh, the upper cretaceous limestones are
overlaid by sandstones and greenish shales in which no fossils were found,
except an Exogyra and Cerithium, and some fucoid- markings and badly
preserved remains of fishes and crustaceans. These beds; may be
regarded as eocene; they are overlaid by red sandstone, with a few
clay bands and, towards the top, conglomerates composed principally of
pebbles of upper cretaceous limestone, covered by a great thickness of sand-
stones and shales, containing a few fresh water shells and passing upwards
into soft gypseous clays.}

Besides the exposures in synclinal basins within the hills, a zone of
tertiary beds is found all along the edge of the alluvium of the Oxus val-
ley. Here, however, the eocene clays are wanting and the red beds with
conglomerates rest directly, in apparent conformity, on the upper creta-
ceous limestone. They pass upwards with a gradual transition into the
recent deposits, and in the upper portion of the section, there occur beds
which are undistinguishable from the recent wind blown loess of the Oxus

valley.?

The northern extension of the Suldimdn range has not been examined
geologically, but there is good reason to suppose that the fringe of Siwdlik
rocks is continuous with great area of tertiary deposits, extending from
the north-west corner of the Punjab along the outer edge of the Himdlayas
to the border of Nepal. The general description of the Himalayan terti-
aries, with which must be included those of the north-west corner of the
Punjab, will be deferred to the following chapter, but it will be well to
notice in this place a portion of the Kohat district west of the Indus,
where the lower tertiaries exhibit some peculiarities not noticed elsewhere.

The eocene rocks are well developed in this region, and the section
exposed, although only a few miles distant from parts of the Salt range,
differs in some important points. The following is abridged from Mr,

Wynne’s summary of the rocks exposed :—®
Thickness in feet,
Upper sandstones.—Soft, grey sandstones, clays and
conglomerates . . . . ° . e 500 to 1,500
ee Lower sandstones.—Harder grey and purple sand-
stones, bright red and purple clays, slightly cal-
careous and pseudo-conglomeratic bands. « 3,000 to 3,500

IC.L. Griesbach, Records XIX, 255, (1886). | (1886).
2C,L. Griesbach, Records, X1X, 257, 259; 3 Wynne, Menivirs, XI, 101, (1875).
Y¥2
326 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap, XITT.

Thickness mm feet,

Upper nummulitic.— Nummulitic limestone and some

shaly bands . . . . ‘ . 60 to 100
Red clay zone, or lower nummulitic.—Red clay,

lavender coloured near the top, occasionally with

Nummulites. The lower portion of the red clays

in places is partly or wholly replaced by fossilifer-

ous sandstones, thick greenish clays and bands of

limestones, all containing Nummultties ‘ é 150 to 400

Eocene é

Gypsum,—White, grey or black gypsum with bands
; ( of clay or shale ‘ : : . 5 50 to 300
Eocene? -4 Rock salt.—Thickbeds of salt, almost pure. The
( base not seen : . j . 300 to 700
(1,200),

The region examined is the hilly tract north of the Bannu plain and
of the Chichdli hills, and extending from the Indus, on the east, to the
British frontier. The ground is traversed by a series of east and west
ranges, chiefly formed of crushed and broken anticlinals of the nummulitic
limestone and the associated rocks.

The rock salt and gypsum at the base of the tertiary series in the
Kohdt region are very important and remarkable. ‘Ihe salt consists of
a more or less crystalline mass, usually grey in colour, with transparent
patches, and never reddish like the salt of the Salt range. A few earthy
bands occur, but the portion of the whole mass too impure to be worked

 

 

 

 

Fig. 18.— Hill of rock salt, 200 ft. high, at Bahddur Khel, after Wynne.

for commercial purposes is but small, although there is no attempt at
refining the salt, which is exported for sale in the form in which it is
mined, In some places the uppermost layer is dark coloured, almost black,
Chap. XIID] KOHAT SALT REGION. 329

and bituminous. The quantity of salt is something marvellous; in the
anticlinal near Bah&dur Khel alone, rock salt is seen for a distance of
about eight miles, and the thickness exposed exceeds 1,000 feet, the width
of the outcrop being sometimes more than a quarter of a mile. As a rule,

Eastern Salt hill

Fault 200 fot above Valley ***~

     

Fig. 19.—Section through the Bahddur Khel salt locality, after Wynne.

the salt contains sulphate of lime (gypsum), but none of the potassium
and magnesium salts of the Salt range beds,

Above the salt come gypsum and clays, as in the Salt range, but the
colours, white and grey, are very different, and the whole appearance, of both
salt and gypsum, is so distinct from the Salt range marl that, although
there is no indication of ‘salt beds at a higher level in the Salt range
itself, and although the outcrop of the salt marl close to Kdlabagh on the
Indus is only eighteen miles from one of the Kohdt rock salt regions, it
appears probable that the salt bearing rocks in the Kohat district may
belong to a very different horizon from that occupied by the same miner-
als in the Salt range series, a supposition supported by the fact that a great
series of mesozoic and palzozoic beds intervenes, throughout the Salt
range, between the nummulitic group and the salt marl, whilst in Kohat
the former rests with apparent conformity upon the gypsum and salt. It
is by no means certain that the Kuhdt salt and gypsum are eocene, but, in
the absence of any evidence to the contrary, it appears best to class
them with the nummulitic beds immediately overlying them.!

Overlying the gypsum, there is usually found a thick bed of deep
red clay, whose eocene age is proved by the occasional occurrence of
nummulites in the upper portion. Sometimes the red clay is wanting, and
apparently replaced by grey or olive coloured clays, marls, and limestones
containing nummulites, but the replacement is not clearly proved.
Above the red clay zone come earthy limestones, clays, and shales
with nummulites. The main band of limestone is very much thinner
than in the Salt range, but is, as usual, massive, pale coloured, and full
of Nummulites, Alveolina, etc. The overlying formation, consisting of
sandstones and clays, in which dark red and purple colours predominate,
exceeds all the eocene beds in thickness, and is probably, like the sand-

1These conclusions would have to be modi- | the salt and gypsum is adopted,—see supra,
fied if the hypothesis of hypogene origin of | p. ttt,
328 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. (Chap. XIIT

stones and clays overlying the nummulitic limestone of the Salt range,
really of much later age than the limestone on which it rests.

Pebbles of nummulitic limestone are said! to be found in the lowest
beds of the sandstone, and some reptilian bones (not determined), silici-
fied fossil wood, and a few ill preserved, ribbed, bivalve mollusca have
been found, but no characteristic organic remains. ‘Ihe heds resemble
those of the Murree group and pass upwards into undoubted newer tertiary
(Siwdlik or Manchhar) strata of the usual character, the red colours be-
coming rarer, and the usual drab grey sandstones and orange or drab clays
being the prevailing rocks.

When traced in a westerly direction towards Thal, the eocene beds
are said to change in character, the limestones being replaced by hard
sandstones and greenish grey or purple shales and, cn the right bank of
the Kurem river opposite Thal, volcanic beds occur among them. The
beds are said to be penetrated and indurated by intrusions of a dark
coarse crystalline trap, and besides the intrusions agglomerates: and fine
grained tufaceous trappean beds are found in alternating layers.2_ No
other occurrence of volcanic beds in the eocene of northern India is known,
and as the Thal locality was visited under circumstances extremely un-
favourable for geological investigation, too much importance must not be
attached to the observations.

No lower tertiary rocks are known south of the great snowy range of
the Himdlayas, between the small outliers folded among the pretertiary
slates in western Garhwdl and their re-appearance on the southern side
of the Assam range. Nearly the whole of this tract of the Himdlayas is
inaccessible to geological observation and it cannot consequently be
stated with confidence that the lower tertiaries are altogether wanting,
but the general run of their northern boundary, in the north-west portion
of the Himdlayas, appears to indicate that the eocene coast line trended
south of the present limit of the Himdlayas, and makes it improbable
that any lower tertiary marine beds would be found east of Garhwdl, even
if the country now inaccessible were open to observation.

At the western extremity of the Gdro hills the nummulitic deposits
contain only one thin band of limestone about 4o feet thick, resting conform-
ably upon the cretaceous sandstone. Even this limestone is ele earthy,
nodular. and ochreous, with shaly partings, the purer portion being generally
formed of a mass of Nummu/ites granulosa, in various stages of erowth. The
series gene-ally exhibits a shallow water type andan approach to the original
margin of the eocene sea. Throughout the formation there are no clear

1 A.B, Wynne, Memoirs, XI, 170, (1875). | 2 A.B. Wynne, Records, XII, 111, (1879).
Chap. XIILJ “GARO HILLS. 329

sandstones ; clays and soft earthy sandstones overlie the limestone and are
with difficulty distinguished from the succeeding upper tertiary deposits.
This character is most pronounced at the west endof the hills, where the
formation strongly resembles the most characteristic beds of the Subathu
group in the north-west Himalayas. No coaly layer has been found in the
series in the G4ro area. The lower tertiary beds are not confined to the
southern margin of the Garo hills, outliers being found in some of the basins
of the cretaceous rocks north of the Turd range. The country has not,
however, been sufficiently closely surveyed to show whether these indicate
the same thinning out and approach to the original shore line as is seen in
the Khdsi hills further east.

A number of fossils obtained from the lower tertiary beds of the Géro
hills appear to belong, so far as they can be identified, to eocene species,!
and the presence of bands of limestone full of nummulites is in itself sulfi-
cient to fix the age of these beds as eocene.

To the east, the nummulitics show a change to deposits of more pelagic
character, in which there is a great development of clear nummulitic lime-
stones. The transition between the prevailing types of deposits in both
the cretaceous and nummulitic strata is in the ground, described by
Colonel Godwin-Austen, on the confines of the Gdro and Khasi hills,
where the upper tertiaries have been almost denuded away from the base
of the range, and here a partial unconformity was noticed between these
systems.

The general character of the nummulitic series at the foot of the Khasi
hills is shown by the section iu the Tharia river, where the following beds
are exposed, all with a high southerly dip :—*

Feet.

7. Limestone, coarse, massive, blue . < . . * « 200
6. Sandstone, clear, yellowish, coarsish . Fs é . - 100
5. Limestone, fine, compact, blue or pink . ; . . - 200
4. Sandstone, earthy, greenish, ochrey . . . . * 50
3. Limestone . . . . . is . . . - 50
2, Sandstone, yellowish . : ‘ : : . . - 100
1. Limestone 7 . . ; . : - 3 » 200
Tova 2 900

On the highlands immediately to the south-west of the station of Cherra
Punji, there is a small plateau of nummulitic strata, very much reduced
in thickness. The bottom 8o feet are of limestone, covered by about an
equal thickness of sandstone, not markedly different from the underlying
cretaceous rock, At about ten feet above the limestone there is a thick
seam of bright coal, the well known Cherra coal. The limestone rests

1 Records, XX, 42, (1887}. 3 Memoirs, I, 140, 185, (1858); Records, XXU,
2 Memoirs, Vil, 164, (1869). 167, (1889).
330 GECLOGY OF INDIA—-EXTRA HIMALAYAN TERTIARY. [Chap. XIT,
directly upon the surface of the Cherra sandstone, without any sign
of denudation, and the bedding is parallel, having a southerly slope
of three degrees! The fossils from this limestone were determined
by Dr. Stoliczka; portions of the rcck consist almost entirely of small
specimens of Operculina canalifera, Nummulites lucasane, and
N. ramondi, both the last species very small, species of echinoderms,
fragments of oysters, Pecten, Cardium salteri and fragments of Natica
Cerithium, Turritella, etc?

Close to the north of the coal hill, the nummulitic beds occur again in
equal thickness, under the native town of Cherra Punji. The limestone
is not seen here, but this may be partly due to concealment. The num-
mulitic sandstone forms the highest ground of the plateau from Cherra
Punji_ to beyond Surarim. Carbonaceous markings are frequent in it,
and at Lairangao, four miles north of Cherra village, there is a workable
seam of coal, It is at about the same height over the cretaceous sandstone
as the seam at Cherra, but the underlying beds are all sandstone
and shale, except one bottom bed of limestone, resting on the Cherra
sandstone. In this limestone bed Operculina canalifera and Nummutlites
lamarcki ate very common ; with these occur a 7vochocyathus, Stylocenia
vicaryi, Echinolampas spherotdalis,a small Cardita, Pecten, Nattca roualtt,
Keilostoma marginatum, a Ziziphinus, the small Cerithium hookert, casts
of a large Natica, Cerithtum and Terebellum?

Under Surarim, only a mile from Lairangao, this bottom bed of lime-
stone is wanting, being apparently overlapped by the carbonaceous
sandstones, which themselves come to an end ina low bluff, two miles
further north, near where the road bifurcates.

In the corresponding section on the Bogapdni, below Chela, there is a
coaly layer in the sandstone just over the bottom limestone, which may
sufficiently establish the identity of these two bottom bands with those on
the plateau above, the marine bed having increased in thickness, and
this marine character is here strongly stamped upon the whole series.

In the Khasi hills, at least on the Cherra section, the northern thinning

As. Soc. Beng., XX XVII, pt. ii, 21, (1869).
2 Memoirs, VI, 167, (1869). ‘The fossils de-

1 This little plateau at Cherra offers a re-
markable instance of a form of denudation
that is not, perhaps, taken sufficient account

of in geological explanations. The scarp is
very regular and well defined at many points;

but the upper surface of the area, about a,

square mile in extent, is a chaos of tilted masses
of the upper sandstone. This is clearly due
to the more or less camplete removal by solu-
tion of the supporting limestone. Colonel
Godwin-Austen has described a much larger

instance of this form of denudation in the

Garo hills, where a considerable enclosed catch-
ment basin is drained underground.— Your.

 

scribed from eastern Bengal by MM. D’Archiac
and Haime in their “Groupe Nummulitique de
Inde” were all from the Khdsi hills; but their
Specimens were so mixed, probably even inclu-
ding fossils from the cretaceous beds, that the
value of their identitications is dou-tful. As
the authors themselves remark—“ On voit qu'il
y a un certain vague dans les rapports de
plusieurs des assises que mous venous
d'indiquer.” —J. ¢., p, 177.

3 These fossils were determined by Dr
Stoliczka,— Memoirs, VII, 167, (1869).
Chap. XiTl] KHASI HILLS. 331

out of the nummulitic series upon th2 gneissic plateau is not so clearly
seen as in the case of the cretaceous rocks, because only remnants of
the bottom bands are preserved there. It will be seen however, from what
has gone before, that there is a distinct indication of a northerly thinning
out of the lower tertiary beds and a disappearance of the marine type of
deposits, The nummulitics have not been observed anywhere to rest upon
the gneiss, so there is no proof of their having overlapped the cretaceous
deposits, but it is presumable that they did so, for the sequence is con-
formable, or at least parallel and undisturbed, and the nummulitics extend
close up to.the northern boundary, where the cretaceous beds are very thin.

The great thickening of the series in the sections at the southern foot
of the hills points to the same conclusion, and if the occurrence of coaly
beds in the Bogapdni below Chela at a horizon corresponding to bed No. 2
of the section may be taken to indicate an equivalence with the coal bear-
ing sandstone of the plateau, there is not only a great thickening of this and
the underlying group, but there are some 600 feet of overlying strata which
do not appear to be represented on the plateau to the north.

The nummulitic limestones form the most prominent features of
the low hills at the foot of the scarp along the Khdsi area, where they
are in much greater force than elsewhere; to the east or west. This may
be owing to a greater elevation in this position, whereby a deeper zone of
the basin of deposition has been exposed to view. Pure limestone is still the
chief rock of the group on the eastern confines of the Garo hills, west of
the Umblai, as described by Colonel Godwin-Austen, although the total
thickness must be much less than in the Tharia section. :

To the east the nummulitic limestone is known to be well developed
in the North Cachar hills, where it has a thickness of 300 feet, resting on
the cretaceous, and overlaid by the upper tertiary sandstones. In a north-

easterly directicn it has been traced as far as the hot springs of the Kapili
valley.!

No beds known to be of nummulitic age have yet been discovered in
upper Assam, but a series of coal and petroleum bearing rocks, conform-
ably overlaid by sandstones of the upper tertiary type, which appear to
represent them, is exposed near the edge of the hills north of the Brahma-
putra valley in the Sibsdgar and Dibrugarh districts.

For detailed information regarding the distribution of the rocks
Mr. Mallet’s Memoir must be referred to.? The measures are very
much alike in all the exposures, seams of less than a yard in thickness
are very numerous in some sections, and not unfrequently the coal beds
attain much greater dimensions. In the Nd4mddng, south of Rongreng
in the Mdkum field, there is a seam 100 feet thick, containing at least

1 T, D. LaTouche, Records, XVI, 201, (1883). | % Memoirs, XII, pt. 2, (1876).
332 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap. XII

75 feet of solid coal, and some very thick seams have been traced for more
than a mile without diminution. The sandstones and shales often contain
nodules and layers of clay ironstone. Earthy and ferruginous limestone
occurs sparingly in thin concretionary bands, also some layers of hard
tough magnesian limestone. The coal measure shales decompose into a
very tenacious blue clay, differing in this respect from the Disang shales,
which are more clunchy.

On the interesting question of the age of these very important and
extensive coal measures, there is little evidence for opinion and that little
suggests a middle tertiary horizon. The coal itself is a true coal of
superior quality, not lignite, as is attested by its composition—

Fixed Volatil
carbon, matter. Ash.
Average composition of 27 Assam coals . 60°0 362 3°8
Ditto of 17 Raniganj coals . 511 32°6 16°3

It is not unlike the nummulitic coal of the Khdsi hills, and quite unlike
the cretaceous coal, which maintains its peculiar characteristics into close
proximity to the Assam fields, The only fossils found in the seams
are bad impressions of dicotyledonous leaves and no trace of animal life
has been seen in any of the associated rocks. The strongest point in the
argument is the closely transitional relation between the measures and the
Tipam sandstone, which is a very typical representative of the Siwdlik
rock, and almost undoubtedly belongs to the upper tertiary period. In
the Khasi hills, as in the Punjab, the nummulitic coal occurs near the
very base of the formation. There are about 1,000 feet of marine nummu-
litics above the coal bed in the Tharia section, but the upper Assam
coal measures, if they are nummulitic at all, would seem to belong to the’
upper limits of the series,

The nummulitic beds in the Gdro hills pass into the upper tertiaries,
which may here be divided into a lower group of fine grained sand
stones with marine fossils, and an upper unfossiliferous group of rusty
sandstones, variable in grain, with grey shales. These uppermost tertiary
beds have not yielded any fossils to recent observers, but in 1821 Mr. Cole-
brooke read to the Geological Society of London ! some observations
on these rocks by Mr. David Scott, then Commissioner of Ktch Behar.
In those days the Brahmaputra flowed at the foot of the hills, and at one
spot on the left bank Mr, Scott found some fossils. The exact locality
is not given, but it was somewhere between the Kdlu-and Mahendraganj
(or Karibdri), probably nearer the latter place. The position in the series
is not known, but it seems certain that the bed belonged to the rocks of
the hills, not to the older alluvium, for it is described as at the foot of a

1 Geol, Trans., series ii, I, 132, (1824).
Chap, XII) UPPER TERIIARIES OF ASSAM, 333

small hill, rising about twenty feet over the general elevation of the plain.-
The fossils were a strange mixture of marine with land and fresh water forms,
and amongst them Mr. Pentland described! the teeth of Anthracotherium
stlistrense, a species that has since been found in the Manchhar (upper
miocene) beds of Sind.?

The change of character that is so marked in the cretaceous and
nummulitic deposits from west to east has not been observed in the upper
tertiaries ; there certainly can be no striking feature of this kind. A
massive soft greenish sandstone is the most prominent rock. It is some-
what like the common Siwdlik rock, but more earthy and of darker hue.
The associated bedsare mostly grey shales, unlike the brown and ochrey
clays of the sub-Himdlayan series, and the fossils show that even at the
western end of the range the deposits are, in part at least, marine. In Mr.
Scott’s notes of the section on the Brahmaputra, local contortions are
noticed, and this condition increases eastwards. On the Sumesari,
nearly sixty miles further east, where the tertiary zone is fourteen miles
wide, the state of disturbance is still only partial. At the southern edge
of the hills the dip is 40° to the south, in very new looking strata, there
is then a broad band in which the beds are flatly undulating, after which
they rise again with a steady and increasing southerly dip.

Nothing like a general unconformity in the tertiary series was noted in
the section of the Sumesari.

Further east lies the ground described by Colonel Godwin-Austen on the
confines of the Gdro and Khdsi areas, where the upper tertiary rocks have
been almost denuded away from the base of the range, the little that is left
of them being nearly vertical. Here, too, is the transition ground of the
prevailing marine type of deposits in both the cretaceous and nummulitic
strata. Here a partial unconformity was noticed between these formations
and this is the only position in the western part of the range where the
upper tertiaries have been found on the plateau inside the line of dis-
turbance. The summit of Nongkulang hill (2,070 feet) is formed of
rusty sandstones and shales, resting horizontally upon the undisturbed
nummulitic limestone. Ina collection of fossils from these beds, sent by
Colonel Godwin-Austen, Dr. Stoliczka found the genera Conus, Dolium,
Dentalium, Cardita, Cardium, Tellina, Nucula, Leda, Cucullza, and
several others, and he remarked that none of the species, so far as
recognisable, appeared to be identical with those known from the nume
mulitic beds of the same district. his fact suggests that these detached
beds on the heights may be an overlap of some beds of the series
higher than those that seem to be in transitional sequence with the

1 Geol. Trans., series ii, IT, 393, (1829). 3 Memoirs, VII, 193, (1869).
2 Records, X, 77, (1877).
334. GEOLOGY OF INDIA—ENTRA HIMALAYAN TERTIARY. [Chap, XII

‘nummulitics in the sections to the west. This view is strengthened
by the fact that Colonel Godwin-Austen observed some cases of local
denudation unconformity between these fossiliferous sandstones and the
nummulitic limestone, the strata being still quite parallel.

The exposure of upper tertiaries south of the Gdro hills narrows in an
easterly direction, and from about twenty miles east of the Sumesari river
the disturbed upper tertiaries have been removed by denudation, along the
foot of the Khd4si and Jaintia hills. They come in again where the strike
of the disturbance, which marks the southern margin of the plateau of the
Assam range, turns north-easterly, and occupy a large area in the hilly
country between Assam and Burma, and to the south of the Cachar valley,

The observations in this country are scanty and scattered. In the
Kasom range, on the eastern limit of Manipur territory, they consist of
sandstones with but few argillaceous bands, containing numerous car-
bonised fossil tree trunks, and nests of coaly matter, but no coal seams
were observed. Further south, however, well marked coal seams have
been found in what appear to be the same beds on the westside of the
Chindwin river. Fossil resin was found in the Manipur hills, and it is
probable that the amber mines of Upper Burma are excavated in beds of
this age.”

Further north marine fossils, among which is an undetermined
species of Venus, were found near Sdmaguting in beds which probably
represent the lower group of the Géro hills.

In upper Assam, the upper tertiary beds were described by Mr. Mallet as
_ the Tipam and Dihing groups. The greenish grey, pepper and salt,
sandstone of the Tipam range undoubtedly alternates with the top beds
of the coal measures, but the shaly beds rapidly cease and the sandstone
becomes very massive, Some coaly partings were also observed well up
in the Tipam group, and fossil wood, whether silicified or semicar-
bonised, sometimes in very large blocks, is common throughout this great
sandstone formation, as well as in the upper conglomeratic beds.

The upper (Dihing) group is less like the corresponding group of the
Siwdliks than is the Tipam sandstone, and two points are noteworthy in
it. Some of the bottom beds are coal conglomerates, made up almost
exclusively of rolled fragments of coal, presumably of the coal seams
underlying the Tipam group. This peculiar conglomerate has been
observed, at great distances apart, at the top of the Tipam sandstone
along the main fault. The ordinary conglomerates of the group are
composed of well rolled pebbles of fine hard sandstone, identical in
appearance with the Disang sandstone. The beds associated with these

) Four, As. Soc, Beng, XXXVIIL pt. ii, 14, 2 Memoirs, XIX, 226, (1883).
16, (1869).
Chap. XIII.) UPPER TERTIARIES OF ASSAM AND MANIPUR, 335

conglomerates are blue sandy clays, not like the brown and red clays of
the upper Siwdliks, at least of north-western India.

The distribution of these upper tertiary sandstones presetits some
peculiarities which require notice, Atthe easternend of the Assam range
they come in where the zone of disturbance which marks its southern
face diminishes in intensity and at the same time turns to the north-
east. From here they extend eastwards into the Patkoi range, but
do not extend across the hill country of Manipur. In the parallel of
Manipur city there is a stretch of about 100 miles broad of pretertiary
beds, before the upper tertiaries are again found capping the range which
looks down on the Chindwin valley. To the north of the valley of
Maripur the upper tertiaries are found capping the higher ranges, and
further north, in the Angd4mi N4gé hills, there is only some ten or twelve
miles between the boundary of the sandstones forming the high peak of
Japvo and the margin of the eastern area of upper tertiaries. No details
are known of the geology of the hills to the north-east of this, but the pre-
tertiary slates and sandstones of Manipur and the Angdmi Ndgd country,
appear to be represented by the Disang group in the coal fields of upper
Assam. Whether there is a continuous outcrop of pretertiary slates
in the intervening ground is not known, but in the Singpho country
further east, the pretertiary beds are completely cut out by the tertiaries,
which extend continuously from the plain to the crest of the range.

In a southerly direction from Manipur much the same appears to take
place, for east of Chittagong Mr. La Touche found nothing but tertiary
rocks on the route across the hills. It would seem, therefore, that the
Manipur hills are an area of special elevation from which the newer beds
have been removed by denudation.

In the last paragraph it has been assumed that these deposits originally
extended over the area where they are now wanting. There is no direct
proof that they ever did so, and the general resemblance in type to the
Siwdliks along the foot of the Himalayas might suggest that they were
deposited under similar circumstances, and that their original was not very
different from their present extension, but in spite of a certain lithological
similarity to the Himdlayan Siwdliks, there is a much greater uniformity
on different sections than they exhibit. The present boundaries, too, are in
all cases evidently due to denudation, and there can be little doubt that
they once extended right across the hills which separate Assam from
Burma, at any rate north of Manipur. The presence of marine fossils in
these sandstones shows that they were deposited under different condi-
tions to the Himdlayan Siwdliks, which we will see were formed
subaerially by streams, under circumstances closely resembling those
that now prevail along the foot of the Himdlayan range.
336 GEOLOGY OF INDIA~EXTRA HIMALAYAN TERTIARY. [Qhap, XIII,

Upper tertiary beds occupy a large area in Upper Burma, both in the
valleys of the Irawadi and Chindwin rivers, and in the hills to the west of
them. Little is known of this area, as no detailed geological survey
has yet been.practicable. Coal seams are found interbedded with the sand-
stones west of the Chindwin river, and of the Irawadi above Mandalay, as
also in the small outlying patches of upper tertiaries in the Shan hills,
Further south the petroleum of Burma is derived from upper tertiary beds,
though there are also some occurrences of very minor importance in the
older tertiary rocks of Pegu.

It is not till we reach the province of Pegu that anything approaching a
detailed account of the tertiary deposits is available, where they were
divided by Mr, Theobald into three main groups, supposed to correspond
more or less ta tlie eocene, miocene, and pliocene of the European classi-
fication, but it will be best to retain the local nomenclature, as the fossil
evidence is by no means sufficient to establish their complete equivalence.
The three main tertiary groups are as follows :1—

3- Fossil wood group : F P . Sand gravels, etc., with silicified wood and
mammalian bones.

2. Pegu group ; . . ; . Shales and sandstones, occasionally cal-
careous ; fossils numerous,

1. Nummulitic ‘ ‘ ; ‘ . Shales and sandstone, with some lime-

stone bands containing nummulites, etc.

There is but little to distinguish the nummulitic beds of Pegu from the
Ma-iiand Negrais rocks, beyond the much smaller amount of alteration
that they have undergone, the more frequent appearance of fossils,
and the occasional occurrence of limestone containing nummulites, especial-
ly in the higher part of the group. The ordinary beds are sandstones
and shales, unaltered but frequently hard and compact. The distinction
from the Negrais rocks is far from absolute. The tendency to a passage
between the two, at the foot of the hills, has already been noticed, and
there are in places, within the nummulitic area, hills formed of hardened
masses, perhaps older than the rocks around, but which have much the
appearance of being the same beds, slightly altered.

The main outcrop of the nummulitic rocks extends from north to south
throughout the province of Pegu, between the Arakan hills and the Irawadi
river. The beds have a general dip to the eastward, but to the southward
it is difficult, if not impossible, to define the base of the formation, on account
of the apparent passage from the nummulitic into the Negrais rocks. To
the northward, west of Thayetmyo, near the former boundary of British
territory, the section is better defined.2 In the Hlwa (Lhowa) stream,

i X, 227, (1873). — the nummulitics, The information necessi-
he beds formerly classed as triassic, on tating this change was not received in time to

the strength ofa mistaken identification of a | be incorporated in the text or in the geologi-
bussil (supra, p. 145) must be included with | cal map.
Chap. XII1.] NUMMULITICS OF PEGU. 337

sixteen miles west by south of Thayetmyo, upwards of 4,000 feet of hard
sandstones, mostly grey, and of blue, grey, or yellow shales, are exposed,
but throughout all this thickness of beds no fossil remains have been de-
tected, except a few carbonaceous markings. Apparently, at a somewhat
higher horizon on the Ma-tun stream, which joins the Hlwa from the
north, there is a great thickness of massive blue shales, of rather a dark
indigo blue in general, but sometimes of lighter colour, These shales
cannot be much less than 3,000 feet in thickness, but they are almost
as unfossiliferous as the sandstones and shales on the Hlwa, the only
organic remains found being some cycloid fish scales, Above these
there is again a great thickness of sandstones and shales, mostly un-
fossiliferous, but containing a few layers with nummulites, and at the top
of the whole group is a band cf nummulitic limestone, from 10 to 100 feet
thick, This limestone, however, is by no means continuous. Where it
occurs, it seems to be the uppermost band of the group, but it frequently
appears to thin out, and in fact to consist of irregular lenticular bands in
shale, rather than of an unbroken bed. Denudation may, perhaps, have
removed the limestone in places before the deposition of the next group.
Other bands of limestone occur at a lower horizon, but they are more
irregular than that at the top of the group.

The whole thickness of the formation must be considerable—probably
not less than 10,000 feet, but no estimate of any value can be made,
on account of the imperfect manner in which the rocks are seen. In
northern Pegu, west of Thayetmyo, the breadth of the eocene outcrop
from east to west is seventeen miles, but, afew miles to the south,
the width diminishes, till, west of Prome, it is not more than six, The
belt again expands in breadth near Akauktaung, on the Irawadi above
Myanaurg, but the beds are very poorly exposed in general, being covered
with gravel and other later deposits. Farther to the southward, west of
Myanaung and Henzada, the nummulitic rocks are much concealed by
posttertiary gravels and from Henzada to Bassein the only rocks seen
west of the Irawddi plain are the altered Negrais beds. The nummulitic
strata re-appear west of Bassein, and continue thence to Cape Negrais,
but still the rocks are much concealed by gravel. Throughout the area,
however, limestone with nummulites occasionally appears amongst the
higher beds of the group, and a peculiar, very fine, white or greenish,
argillaceous sandstone, with Foraminrfera, seen at Puriam point east of
the Bassein river, and in Long island of that river, is also probably one of
the uppermost eocene beds. This rock, known as Andagu-kyauk, or
image stone, is employed by the Burmese for carving into images of
Buddha, and is quarried to some extent for that purpose.

It is possible that nummulitic beds may crop out in places amongst the
miocene rocks of the Prome district, but the only known exposure of the
338 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap, XIII

former in Pegu, apart from the belt just noticed as extending along the
eastern side of the Arakan Yoma, is in a small ridge, known as Thon-
doung, or lime hill, about five miles south of Thayetmyo. This ridge con-
sists in great part of nummulitic limestone, resting upon shales and
sandstones, In 1855 a promising bed of coal, 4 feet thick, was discover-
ed in the latter, but it proved so irregular as to be of no value, the coal
thinning out, and passing into a clay with mere laminz and patches of
coaly matter, in the course of a few feet.!

To the west of the Arakan range, limestone with nummulites has been
noticed near Keantali, and there can be but little doubt that eocene beds
extend along the coast for a considerable distance. The islands of
Ramri and Cheduba consist of sandstones and shales closely resembling
those of Arakan, and doubtless belonging to the same series.? These
beds are also very similar to the nummulitic rocks of Pegu. A few
seams of coal have been found, resembling in character the nummulitic
coal of Assam, and petroleum is obtained in several places, The lime-
stone on the eastern side of Ramri island, as already mentioned, re-
sembles that of the Arakan coast near Ma-{ and Taung-gup, and may, there-
fore, be cretaceous, but there is no marked character by which the rocks
of the island can be divided into two series.

Above the nummulitic formation of Pegu there is an immense thick-
ness of soft shales and sandstones, often fossiliferous, but almost destitute
of any horizon distinguished either by mineralogical characters or by
organic remains. The base of this group is assumed to coincide with the
band of nummulitic limestone already mentioned, but there is no clear
evidence that this bed is the uppermost rock of the eocene group, and no
unconformity has been detected between the nummulitic rocks and the
next strata in ascending order. The upper limit of the middle tertiary
rocks of Pegu is equally ill defined, there being a gradual passage from
clays and sandstones with marine fossils into the gravels and sands with
silicified fossil wood and mammalian bones.

The fact is that without a thorough knowledge of the fossils the classi-
fication of rocks so obscure and so ill seen as those of Pegu is a simple
impossibility, and until the tertiary molluscs, echinoderms, and corals of
southern Asia are better known, it is hopeless to attempt more than a
general rough arrangement of the Burmese tertiaries. In the absence of
sufficient fossil data for the proper determination of different beds, all that
has been attempted at present is to class together all the marine beds ot
Pegu above the nummulitic limestone, and without nummulites, The
group thus constituted has been named the Pegu group from its form-
ing the greater part of the Pegu Yoma between the Irawadi and Sittaung.

'T. Oldham, Sel. Rec. Gout. India, X, 99, (1856). | 2 F, R. Mallet Records, XI, 191, (1878).
Chap. XIII.) PEGU GROUP. 339

There can be no doubt that a portion of this group is of miocene age, and
corresponds generally to the Gaj group of Sind Lut it is probable that
representatives of other groups are included.

The only approach to a subdivision of the Pegu group that has been
suggested is the separation of a considerable thickness of soft unfossili-
ferous blue shales, which rest upon the upper nummulitic strata near
Prome, and underlie the typical fossiliferous middle teitiary beds. These
shales have been called the Sitsyahn shales, from a village on the Trawadi,
eight and a half miles above Prome, whilst the overlying sandstones and
shales with fossils are distinguished as Prome beds, from their cccur-
rence in the neighbourhood of Prome. ‘The Sitsyahn shales consist of blue
clunchy clay with indistinct bedding, and greatly resemble some of the
nummulitic shales, except that they are somewhat paler in colour. ‘The
thickness of the subdivision is about 800 feet, and the beds have been
traced for a considerable distance along the upper limit of the nummulitic
rocks in the Prome district.

The Prome beds succeed the Sitsyahn shales conformably, and are
composed of grey sandstones, occasionally hard, but frequently argilla-
ceous or shaly, hard yellow sandstones, and shales or clays of various
colours. A section of about 2,500 feet of these beds is seen opposite
Prome on the right bank of the Irawadi, and probably a much greater
thickness exists east of tke river. One of the most fossiliferous beds is a
band of blue clay exposed at Ka-ma on the Irawadi, eighteen miles above
Prome. The position of this band is high, and, above it, a bed, abounding
in Zurritellz, and a hard sandstone containing corals belonging to the
genus Cladoecra, are the high est rocks of the group, and mark the passage
into the fossil wood beds.

It is almost useless to give any palzontological details. Suraminifera
and Echinodermata are rare, and the mollusca are not, as a rule, very
characteristic forms, A_ sessile cirriped, very common in some beds,
closely resembles Ba/anus sub/zvis.of the miocene in Sind. A few small
crabs occur, and small corals and sharks’ teeth are common.

In one locality, Minet-taung (Myay-net-toung), twenty-four miles east-
south-east of Thayetmyo, a bedded volcanic rock, consisting of greyish
trap, occurs interstratified with the rocks of the Pegu group and, to all
appearance, contemporaneous. Nothing has been ascertained as to the
source of this igneous formation.

The Pegu group forms nearly the whole of the great range of hills,
known as the Pegu Yoma, between the Irawadi and Sittaung, no older
rocks being known, with any certainty, to occur in the country between the
two rivers. The area occupied by the middle tertiary beds is very broad
to the northward, where it extends from considerably west of the Irawadi
to the base of the metamorphic hills east of the Sittaung, and contracts

zZ
340 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Chap. XIII,

gradually between the alluvial plains of the two rivers to the southward,
till it terminates in a long, narrow spur at Rangoon. West of the Irawadi,
the Pegu group extends to a little below Prome, and some hills on the
opposite side of the river below Prome are formed of the same beds. It
is, however, not quite certain that no older rocks appear between the
Irawadi and Sittaung for a species of Pseudodiadema, a genus of
echinoderms with cretaceous affinities, has been found in some beds in east-
ern Prome and a Zerebratula with a very cretaceous aspect was obtained
near the town of Pegu, In the former case the beds appear to be high
in the Pegu group, but owing to the great extent to which the surface of
the country is concealed, both by gravel and other alluvial deposits, and
by forest, it is most difficult to make out the geology satisfactorily, so that
lower beds may have been brought up to the surface by faults or otherwise,
In the case near Pegu the position of the beds is uncertain,

A small island, known as Kau-ran-gyi on the Arakan coast, is com-
posed of a very pale brown or cream coloured, calcareous sandstone
or earthy limestone, containing echinoderms, molluscs, sharks’ teeth,
and other fossils. The same rock occurs also at Nga-tha-mu on the
mainland opposite Kau-ran-gyi island, but has not been detected else-
where. The most abundant amongst the fossils are aspecies of Lobophora
(Echinodiscus) and an Echinolampas, apparently £. jacguemonttt, one
of the commonest fossils of the Gaj group inSind. The £chinodiscus also
closely resembles a Gaj species. The bed is somewhat similar to the
miliolite of Kathidwdr, and may represent a portion of the Pegu group,
but it is more probably of later date. One of the sharks’ teeth, however,
ciosely resembles one found in the Pegu group south of Thayetmyo.

The highest member of the tertiary series in Pegu is distinguished
by the abundance of silicified dicotyledonous wood, and is the source
whence all the fragments of that substance, so abundant in the older and
newer alluvial gravels of the Irawadi, are derived. The fossil-wood
group is much coarser than the underlying formations, and consists
of sands, gravels, and a few beds of clay or shale, all, as a rule, being soft
and incoherent, although occasionally hard sandstone or conglomerate
bands occur. The group is thus subdivided :—

a, Fossil-wood sands.—Sand, in part gravelly and conglomeratic, characterised by

a profusion of conncretions of iron peroxide.
b. Fine silty clay, with a few small pebbles,

c, Sands, shales, and a few conglomerate beds, with a litte concretionary iron
peroxide,

The lowest beds, which pass downwards into the marine bands of the
Pegu group, contain occasional silicified rolled ‘fragments of wood, and a
few mammalian bones. Some sharks’ teeth also occur. The thickness of
Chap, XITI.J FCSSIL WOOD GROUP. Bat

none of the subdivisions has been clearly ascertained, but the lower.
sands must comprise beds some hundreds of feet thick. The fine silty
clay dces not exceed about 4o feet in thickness. This bed is quite un-
fossiliferous, neither fossil wood nor bones having been found in it, and
pebbles are rare, though a few occur. It thus forms a marked band in the
group, and contrasts with the beds above and below it.

The upper fossil wood sands and gravels are by far the most import-
ant members of the formation, and it is from them that the greater por-
tion of the silicified wood is derived. This wood occurs ia the form of
large and small masses, some being trunks of trees 40 or 50 feet long ;
usually, however, such masses display marks of attrition, as if the tree
stems had been transported to a distance and rolled, before being silicified.
The wood is always, or nearly always, exogenous, a few rolled fragments
of endogenous woed, found in newer formations, being, nevertheless,
probably derived from the present group. The wood is not coniferous,
but owing to the very considerable amount of decomposition it had
undergone previous to silicification, its nature is difficult to determine.
Besides the fossil wood another characteristic of this portion of the group
is the abundance of concretionary nodules of hydrated iron peroxide,
which are in places so numerous as to have furnished a supply of iron ore
for the native furnaces. Mammalian bones are of only local occurrence.

The following is a list of the Vertebrata, exclusive of sharks’ teeth
hitherto obtained in the Irawadi valley from the beds of the fossil wood
group. Those marked with an asterisk being also found in the Siwaliks
of the sub-Himdlayas :—

MaMMattia.
Ursus, sp. : Tapyrus, sp.
* Elephas (Stegodon) cliftii. Equus, sp.
* Mastodon latidens. Hippopotamus (Hexaprotodox) travadicus.
= ” stvalensis. *Merycopotamus dissimilts.
Rhinoceras travadicus. Cervus, sp.
* » — pewimensis, Vishnutherium iravad cem.
R., sp. Bos, sp.
Reprizia.!
Cracodtlus, sp. Emys, sp.
Ghartaits, sp. Trionyx, sp.
Testudo, sp. Einyda, sp.
* Colossochelys atlas.

 

The proportion of species identified with Siw4lik forms is rather less
than in the case of Perim Island, only five out of fourteen species of mammals
being regarded as identical, yet the general facies of the two faunas is

1These genera are recorded amongst the | Asiatic Society of Bengal, Calcutta, 1859,
Ava specimens in the collection of the Asiatic | p. 30); but as all the specimens were unla-
Society (Falconer, Catalogue of the Fossil Re- | belled, there is some doubt about the locality.
mains of Vertebrata, in the Museum of the

ea
342 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY, {Chap. XIII.

very similar. Both contain a considerable proportion of living genera un-
known in the middle tertiaries of Europe, together with some older forms,
and the Irawadi fauna may be regarded as approximately of the same age
as the upper Siwdliks, or pliocene ifthe views as to the relations of the
Siw4lik fauna advocated in the present work be accepted. Silicified wood
abounds in places in some of the Siwdlik beds of the Punjab and in the
Manchhar beds of Sind, and is occasionally found in the sub-Himdlayan
Siwéliks though the remains of trees are for the most part carbonised.

Independently of the fact that the rocks supplying the materials
from which the beds have been derived east and west of the Bay of Bengal
are probably very distinct, there is some doubt as to the conditions under
which the Burmese beds were deposited, owing to the frequent occurrence
of sharks’ teeth, and it has been suggested that the fossil wood group may
have been, in part at least marine or estuarine. The silicified wood itself
is never bored by xylophagous mollusca (Tevedinidz or Pholadidz) and
as not only all wood floating on the sea, butall found anywhere in tidal
creeks in India at the present day, and even the dead trunks and branches
of trees in places flooded by the tide, are riddled by boring molluscs, it is
extremely improbable that the wocd found in the Burmese pliocene beds
can have been immersed for any length of time in salt water, whilst the
tree stems can scarcely have been silicified before being imbedded, as they
would have been in that case too heavy to be transported. It is true that
the beds containing sharks’ teeth are not those in which the fossil wood
is most abundant, but still some fragments of wood occur with the teeth,
and mammalian bones are common. The beds generally are much too
coarse for estuarine deposits, and if they are marine it is difficult to under-
stand why no molluscs, echinoderms, or corals occur. ‘On the whole, it is
most probable that the fossil wood beds, like the Siwdliks and Manchhars,
are fluviatile or subaerial, deposited by streams and rainwash, and that
the sharks inhabited rivers, as some species do at the present day.

The fossil wood beds in Pegu are evidently the mere remnants of a
formation which once occupied a far more extensive area, the former exist-
ence of the beds being shown by the occurrence in abundance of fragments
of silicified wood far beyond the present limits of the group. Judging
from the occurrence of the larger blocks of fossil wood alone, and neglecting
the small fragments in the alluvial gravels, the beds of the present group
formerly extended far to the southward of their present limits, probably
along the whole eastern side of the Arakan Yoma, and almost certainly
us far as Rangoon along the Pegu range. A considerable area in the
Sittaung valley, north of Taung-ngu (Tonghoo), is also occupied by the
fossil wood beds, but no traces of the former existence of this group is
found south of the Kabaung stream, which joins the Sittaung from the
westward a little below Taung-ngu. There is rather more clay associated
Chap. XIIt.) TENASSEKIM. 343

with the pliocene beds in the Sittaung than in the Irawadi valley whilst
in two small outlying patches, east and north of Taung-ngu, the group
is represented by a form of laterite containing numerous pebbles.

In Upper Burma the same beds are very extensively developed. They
occupy large areas both east and west of the Irawadi, and between the
Jrawadi and the Chindwin above their confluence, as well as west of the
latter river. The details of their stratigraphy and distribution have
not yet been worked out, and there is no published information suffi-
ciently detailed or connected to be incorporated in this work. The
petroleum of Upper Burma is derived from beds which appear to lie at
the base of the fossil wood group, or the upper limit of the Pegu group.

North of where the Irawadi crosses from the eastern to the western
side of the depression between the hills of western China and the Shan
states on the one hand, and of Manipur and the Arakan Yoina on the
other, there are a number of intrusions and volcanic outbursts in the ter-
tiary beds where exact age has not been determined, but appears to range
from upper tertiary to pleistocene,

Although nothing definite is known as to the age of ‘the beds
associated with coal in Tenasserim, except that they are in all probability
tertiary, there is more likelihood that they belong to the newer tertiaries
than to the older. They consist of conglomerates, sandstones, soft shales,
and beds of coal. The conglomerates are never coarse, the pebbles seldom
exceeding a few inches in diameter; the sandstones are fine, gritty, and.
pebbly, clean white quartzose sands, or earthy and of a yellowish tint; and
the shale beds are of a bluish green or blackish tint, very regularly disposed
in thin lamina. The coal is also in thin laminz, with earthy bands.

These coal bearing deposits, whose total thickness nowhere exceeds
goo to 1,00¢e feet, are never traceable continuously over any extended
area, They are found occupying isolated and detached basins in the
great north and south valley of the Tenasserim river, between the main
dividing range separating British Burma from Siam to the eastward,
and the outer ridges near the sea coast to the westward. The small
tracts of tertiary rocks are in all probability of fresh water origin, and
have much the appearance of having been deposited in the small basins
they now occupy. The only organic remains found are dicotyledonous
leaves and scales and bones of fish.!

In the Andaman and Nicobar islands the upper tertiaries are repre-
sented by soft limestones formed of coral and shell sand, soft calcareous

'T, Oldham, Sel. Ree, Govt. India, X, 34, (1856).
344 GEOLOGY OF INDIA—EXTRA HIMALAYAN TERTIARY. [Ohap. XII}

sandstones and white clays with some bands containing pebbles of coral,’
They form the whole of the archipelago east of the great Andaman, whencc
they have been calledthe Archipelago series ; they are also developed in the
great Andaman, more especially in the northern islands and on the west
coast, as well as in the Nicobar islands, The only fossils that have been
found are Polycystinz, which were described by Ehrenberg from the clays
of Nancowry harbour?

No contact section has yet been observed, showing their relations
to the rocks, sandstones and shales, which are regarded as the equiva-
lents of the Negrais rocks of the Arakan Yoma, but the generally newer
appearance of the recks, their tess disturbed condition, and the fact that
pebbles of serpentine have been found in beds which there seems good
reason for referring to this series, leaves little room for doubt that they
are newer, and not, as has been suggested, merely a lithological variety
indicating different conditions of deposition.

* For details, see H. Rink, Die Nikebaris- | Ball, four. As. Soe. Beng., XXXIX, pt. 2,
chen Inseln., Copenhagen, 1847, and Sel. Ree. | 25, 231, (1870); R. D. Oldham, Records,
Govt. India, LX XVII, 109—154, (1870);F. von | XVIII, 133, (1885).

Hochstetter, Reise der Novara, II, 83—112, | 2 Abhandl. K, Akad. Wiss, Berliz, 1875, p.
1864), and Records, Il, 59—73, (1859); V. | 116.
CHAPTER XIV.

TERTIARIES OF THE HIMALAYAS.

(/ncluding the N.-W, Punjab.)

Tertiaries of the central Himélayas—Tertiaries of the outer Himdlayas—Sirmur series — SiwAlik
series—Homotaxzis of the SiwAliks— Relations of Siwdlik and recent faunas,

The description of the tertiary deposits of the Himdtayan range was
excluded from the last chapter, as their bearing on the question of the age
of the Himdlayas as a mountain chain, and certain peculiarities which they
exhibit, resulting from their mode of origin, render it more convenient
to consider them separately. They are exposed in two distinct areas, and
may be distinguished, according to their geographical position, as the ter-
tiaries of the central, and of the outer Himdlayas. The latter of these
are much the most important, whether on account of the superficial area
they cover, or the interest of the sections they exhibit, but it will be most
convenient to take up the description of the central Himdlayan tertiaries
first and then pass on to the tertiaries of the outer or southern edge of the
range.

The existence of tertiary rocks in the central Himdlayas has long been
known, but even now the only information available is derived from ob-
servations made on rapid journeys through an elevated and inhospitable
region, where the rarefaction of the atmosphere offers a serious impedi-
ment to physical exertion.

The best known area is that of the upper Indus valley, where the
tertiaries extend, for a distance of two hundred miles, in a south-easterly
direction from Kargil in Kashmir territory. Along the whole of the north-
eastern boundary, from Kargil to beyond Leh, if not to the extreme limit
of Kashmir territory, they rest in unconformable contact with the metamor-:
phic rocks. Between Khalsi and Leh the lowest beds consist of coarse
grained, sharp, felspathic sandstones, containing a large proportion of
grains of undecomposed felspar, and including numerous large boulders of
syenite and angular blocks of an intensely hard hornstone porphyry, whose
original source is unknown, These beds have been looked upon, with
346 GEOLOGY OF INDIA— HIMALAYAN TERTIARY. [Chap. XIV.

some degree of reason, as probably of glacial origin.’ They occur at the
base of the nummulitics and no break or unconformity has been detected ;
they are however overlaid by black carbonaceous shaly beds very like
those of the carboniferous of Kashmir, and as no similar rock has been de-
tected in the western portion of the outcrop, where the original contact
with the pretertiary rocks is exhibited, it is possible that the glacial beds
belong to the carboniferous period, and that there is an undetected break
between them and the overlying unmistakeable nummulitics.

The conglomeratic beds are succeeded by orange and brown sand-
stones, often calcareous, which form the lowest member of the series
further to the north-west. 4/e/anza, and a bivalve shell, which is probably
a Unio, though it has been referred to Pholadomya or Panopea, have been
found in the neighbourhood of Kargil, marking the beds as fresh water
or estuarine in origin.

The sandstones are succeeded by green and purple or dark red shales,
and these are overlaid, between Khalsi and Nurla, by a thick band of
coarse, blue, shelly limestone containing numerous discs, which are probably
ill preserved nummulites. Above this comes a coarse limestone conglome-
tate containing pebbles of the same limestone succeeded by shales and
slates, generally of a grey colour.

In the sections eastwards of Leh conglomerates are said to occur near
the upper limit of the series, and these conglomerates contain pebbles of
the volcanic beds, which will presently be described, and of nummulitic
limestone. The occurrence of these last shows that the beds had locally
been elevatcd and exposed to denudation, while elsewhere the process of
deposition had gone on continuously.

In the central portion of the exposure the sedimentary beds are in
direct contact with the older rocks along their south-western margin, but
at either extremity they are separated by a great series of volcanic rocks
of a very basic type. There can be no doubt that these rocks, which form
the upper limit of the tertiary system of this region, are in the main con-
temporaneous eruptive products, as they include beds of volcanic ash and
aggiomerate,*® but there are also numerous intrusive masses associated with
the bedded traps. Basic trappean intrusions are also found in the preter-
tiary rocks south-west of the boundary, which are evidently connected with
these same eruptive rocks. These intrusions are interesting as, at Pugha
and in the Markha valley south of Leh, they are composed of peridotite,
until lately the only recorded instances of ultrabasic rocks having been
found in India.@

On the north-eastern and north-western boundaries the tertiaries rest

1 R. Lydekker, Memoirs, XXU, 104, (1883) ; | (1885); R. D. Oldham, Records, XXI, 154,
R. D. Oldham, Records, XXI, 158, (1889). (1888),

*C. A. McMahon, Records, XIX, 118,| *C.A. McMahon, Records, XIX, 115, (1886).
Chap. XIV.] » CENTRAL HIMALAYAS. 347

on an eroded surface of metamorphics, showing that the present boundary
marks aun original limit of deposition in these directions, The south-western
boundary, on the cther hand, is marked by great disturbance and the ter-
tiaries certainly extended some distance beyond their present limit.?
Clear evidence of this original extension is to be found not merely in the
nature of the boundary but in the existence of an outlier of nummulitic
limestone, originally discovered by Dr. Thomson in 1852, on the Singhe
La. The correctness of the observation was at one time questioned, but
was completely confirmed in 1888 by Mr. La Touche,? who described the
rock as a black foetid limestone, full of nummulites, resting directly
on the palzozoic quartzites, without any intervening deposits of littoral
type. The occurrence of an open sea formation in this outlier points
to a southerly extension of the eocene sea, and it would be interesting
to know whether there was direct communication with that in which
the nummulitics of the Punjab were deposited, No outliers of tertiary
beds are known in the country intervening between the two principal
exposures, but this country has not been examined in any detail, and
has undergone such extensive denudation that, even if the nummulitics
once extended over it, they may well have been completely removed
or only represented by small patches, folded up with the older rocks,
which would escape notice in any but the most detailed examination,
The considerable lithological resemblance between some of the beds of the
tertiaries of the upper Indus valley and those of the Subathu and Dagshdi
groups has been noticed by more than one observer,® but this would not in
itself be proof of former continuity, while the general shallow water type of
the Subdthu group, and its complete overlap by the Dagshdi group on the
inlier of pretertiary limestone which lies north of that at Ridsi, indi-
cate a northern limit of deposition of the tertiaries on the southern face of
the Himalayas.

The fossils do not help us in any way, for, with the exception of those
already mentioned, the only recorded fossils are Nummulites ramondi
and N. exponens, fromthe Markha valley ;* the former species has also been
found in the outlier of the Singhe L4,5 and is known from the Kirthar and
Ranjkot groups of Sind. Besides these, some obscure remains were obtained
near Khalsi, which have been supposed to be Aippurites and a cepha-
lopod allied to Hamites, but the fossils are altogether too ill preserved
for determination.

Besides the outlier already mentioned, there is one composed of basic
traps forming the peaks known as D 24 and D 25 in Zanskar. It is not

'R. D, Oldham, Records, XXI, 156, (1888). | 4 F. Stoliczka, Memoirs, V, 344, (1865).

® Records, XXI, 160, (1888). 5 D‘Archiac et Haime: Groupe Nummu-

8 F. Stoliczka, Memoirs, V, 343, (1865); R. | itique de ’'Inde, Paris, 1853, p. 176.
Lydekker, Memoirs, XXII , 118, (1883).
348 GEOLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV,

known whether these are bedded or intrusive. If the former, they area true
outlier of the Indus valley eocene volcanics; if the latter they are doubt-
less an old volcanic core, and indicate a former extension of the volcanic
beds beyond their present limits. Some small patches of sandstone and
conglomerate, which have been regarded as tertiary,’ are also found in the
Chang-cheng-mo valley and near Drds, associated in the latter locality
with basic traps. The information regarding them is scanty, and they do
not need detailed notice here.

Further to the east, in Hundes, the tertiaries are described by Mr.
Griesbach® as composed of highly altered rocks, schists, phyllites and
crystalline limestones, in which some distorted sections of Nummulites
were observed associated with intrusive diorite. Above these rocks come
pepper and salt grey sandstones, very like some of the lower Siwélik
sandstones, whose relation to the nummuliferous beds was not observed,
but is said to be probably one of unconformity. Both these are highly
disturbed, dipping to the north-east, and are unconformably covered by
the horizontal deposits of the Hundes plain, once regarded as upper
tertiary, but now shown to be of pleistocene age.

The only igneous rocks in this section appear to be intrusive diorites,
but a large development of basic traps was observed by Gen. R. Strachey
further to the north, in the neighbourhood of the great lakes of Tibet,
which probably represents the volcanics of the Indus valley tertiaries.

Nothing is known of the geology of the whole northern face of the Himé-
layas east of the Hundes plain except for one spot north of Sikkim. In
the neighbourhood of the Cholamo lakes Sir J. Hooker observed con-
glomerates, slates and earthly red clays, and a compact blue limestone “ full
of encrinitic fossils and probably nummulites’’ which may be a con-
tinuation of the central Himalayan tertiaries.*

The tertiaries of the outer Himalayas are found in a narrow zone of
upper tertiary rocks, extending the whole length of the Himalayas, -and
continuous at the surface, so far as is known, except for a stretch of about
fifty miles at the foot of the Bhutdn hills, where they are covered by a great
accumulation of recent deposits. West of the Ganges the tertiary area
begins to widen out, and lower beds come in, till in the extreme north-west.
beyond the Jehlam, there is a very complete representation of the tertiary
sequence. It is only that portion of the tertiary area which lies beyond

'F. Stoliczka, Records, VII,15, (1874); R.Ly-| ® Himalayan Journals, London, 1855, Il, 156,
dekker, Memoirs, XXII, 113, 115, (1883). 177.
2 Memoirs, XXIII, 83, (1891). :
Chap. XIV. SIRMUR SERIES. 349

the western frontier of Nepal that has been examined in any detail, and
here the best known areas are the hills below Kumdun and Garhwél, the
section south of Simla, and the area beyond the Jehlam river. The classi-
fications of the rocks that were adopted by the surveyors of the north-
western area and of that south of Simla differ from each other, but the
intermediate country has since been examined sufficiently to show the
probable equivalence of the rock groups as indicated in the following
table :—

Nortu-WeEst AREA. Sma AREa,
Upper Siwdlik é Upper Tertiary . Upper Siwalik.
Lower (red and grey) Siwalik Siwdlik series Middle Siwalik.
Lower (Nahan) Siwalik.

Murree beds . S . ‘ Lower Tertiary . pe ae
Upper Nummulitic . Siemur‘seties Subdathu group.
Lower Nummulitic. i

According to the order of description adopted in this work the lower
nummulitic strata of the hills beyond the Jehlam, which do not, strictly
speaking, form part of the Himdlayas, would stand first for description,
but as the reasons for considering them older than the Subdthu group can
only be understood after the description of the Sirmur series, the strict
chronological order will not be adhered to, and the description of these
beds will be left till after that of the series which overlies them. It will
be best to begin with the area which, from priority of description, must
rémain the standard to which the sections of other districts must be
referred.

In the hills near Simla, the marine nummulitics, with the upper groups
of the lower tertiaries, appear from below more recent deposits a short way
west of the Sutlej river, and rising into the high ground of what are known
as the lower Himalayas, extend for some 80 miles till they disappear 12
miles east of Nahan, re-appearing as outliers east of the Ganges.
Throughout part of this region the Sirmur series, as it has been named,? is
separated from the upper portion of the tertiary system by a narrow
strip of pretertiary rocks, and by the great dislocation, or main boundary,
which forms one of the leading features in Himalayan geology. The
rocks ate everywhere highly disturbed, and structurally belong rather to
the Himdlayan area proper than to the tertiary area of the sub-Himdlayas,
but at the north-western extremity of the exposure they run down into the
sub-Himdlayas, and the marine group becomes covered by more recent
deposits in a manner that has not been worked out in detail.

1 Manual, 1st ed., p. §24. The series was | description in the text is partly based on
first described as the Subdthu group (Memoirs, | Mr. Medlicott’s descriptions (toc. cit.), and
HiI,pt. i, p. 74),a name which was subse- | partly on subsequent unpublished observations
quently restricted to its lowest member. The ! by the present writer.
350 GEOLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV,

The lowest of the three groups, into which the Sirmur series has
been divided, is named after the military station of Subdthu, near which
it is well exposed. It consists principally of greenish grey and red gypseous
shales, with some subordinate lenticular bands of impure limestone and
sandstone, the latter principally found near the top of the group. The
beds are everywhere highly disturbed, and the boundary with the pretertiary
slates and limestones is almost always faulted, but wherever an original con-
tact section is found, in the Jammu hills, at Subdthu, at the termination of the
main Sirmur area east of Ndhan, and again in the outliers east of the Ganges,
there is always a parallelism of stratification between the beds below and
above the junction, and the bottom bed of the Subdthu group is a peculiar
ferruginous rock, containing pisolitic grains of iron oxide and closely re-
sembling the laterite of the Peninsula, whose occurrence at this horizon
in the extra-peninsular area is interesting in connection with the occur-
rence of laterite in the nummulites of Sind, Jaisalmer, Cutch and Surat,
This rock is very well seen at Subdthu itself, where it was first observed
and described, and in the shaly beds immediately overlying it there isa
seam of impure coal. The coal is too impure and.too crushed to be of any
economic value, but its occurrence is of interest, as will appear in the
sequel.

The Subdthu group.is overlaid, with perfect conformity, by a great
thickness of hard grey sandstones, interbedded with bright red nodular
clays, known as the Dagshdi group. The transition from the Subéthu to
the Dagshdi group, though perfectly conformable, is somewhat abrupt, and
marked by the presence of a group of passage beds, comprising a pecu-
liar pisolitic marl with small calcareous concretions scattered through
a matrix of red clay, a white sandstone full of irregular shaped highly
ferruginous concretions of some inches in diameter, and pure white sand-
stones associated with dark purple or liver coloured shales, differing
markedly in appearance from the general run of those above or below
them.

The beds of the Dagshdi group proper consist almost ‘exclusively of
two distinct types of rock. One is a bright red or purple, homogeneous
clay, weathering into small rounded nodular lumps; the other'a fine grained
hard sandstone of grey or purplish colour. The clays prevail in the lower
part of the group and the sandstones, in beds of to to 50 feet thick, form
but a small proportion of the total thickness, but in the upper portion of
the group they increase, at the expense of the clay beds, till at the top
there is about 200 or 300 feet of sandstones, with a few thin bands of red
clay, which it is impossible to class definitely either with this group or the
succeeding one.

As will have appeared from the preceding paragraph, the passage from
the Dagshdi to the Kasauli group is perfectly transitional, indeed the
Chag. XIV.) SUBATHU GROUP. 451

distinction of the two merely depends on the absence of the bright red no-
dular clays of the Dagshdi group. The Kasauli group is essentially a
sandstone formation in which the argillaceous beds are quite subordinate
in amount. The sandstones are mostly of grey or greenish colour, and
though some of the beds are as hard as anything in the Dagshdi group,
they are, as arule, softer, coarser, more micaceous, and at times distinctly
felspathic. ‘Ihe clay bands are gritty, micaceous, and but seldom shaly ;
in the lewer part of the group they often have a remarkably trappoid ap-
pearance, owing to their dull green colour and mode of weathering, first
into rounded masses and afterwards into small angular fragments.

At the upper limit of the Kasauli group some reddish clay bands are
seen on the cart road to Simla. These clay bands are softer and paler than
those cf the Dagshdi group, and resemble the clays of the lower portion
of the upper tertiaries near Kdlka. This, the only trace of a connection
_ between the Sirmur series and the upper tertiaries of this region, will be
referred to further on.

The Subdthu group is most palpably of marine origin and of nummu-
litic age, as is shown by the numerous fossils it contains. The Dagshdai
group has yielded no fossils, except some fucoid markings and annelid
tracks, which are of no use for determining either the age or mode of origin
of the beds, but the great contrast of lithological character suggests a cor-
responding change of conditions of formation, and it is probable that they
were deposited either in lagoons or salt water lakes cut off from the sea, or
were of subaerial origin. The Kasauli group has so far yielded no fossils
but plant remains, and this, taken in conjunction with its general similarity
to the upper tertiary deposits, renders it probable that it is composed of
fresh water, if not subaerial deposits.

A short distance west of the Sutlej river the Subdthu group become
covered up by the newer beds, and is not again seen till the Jammu
hills are reached. Here there are some inliers of marine nummulitics, but
the most interesting exposures are those in which they rest on the pre-
tertiary limestones of Ridsi and the Punch valley.! In these the Subdthu
beds rest, with perfect parallelism of stratification and every appearance of
conformity, on the older rocks, and at their base is found the same peculiar
pisolitic ferruginous bed as is seen at Subdthu. Separated from this by
about 70 feet of shales, there is a coal seam of 2 to 5 feet thick, over-
laid by some 350 feet of shales, with a couple of thin bands of nummulitic
limestone,? the group being conformably overlaid by the sandstones and
red clays similar to those of the Dagshdi group.

On the Ridsi inlier the pisolitic bottom bed of the nummulitics is

. H. B. Medlicott, Records, 1X, 53, (1876);| 2? T.D.La Touche, Records, XXI, 62, (1888).
R. Lydekker, Memoirs, XXII, go, (1889).
sea GEOLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV.
underlaid everywhere by a silicious breccia of variable thickness, composed
of perfectly angular silicious fragments, cemented in places by cellular
limonite. The true age or nature of this rock is uncertain. Mr. Medli-
cott thought it was a shattered condition of a sandstone band which often
occurs at the top of the pretertiary limestone series, and in any case the
angular nature of the fragments forbids the supposition that they have

been transported for any great distance.

So far there has been no difficulty in recognising the equivalent of the
Subdthu group, but in the sections west of the Jehlam there is a very great
thickness of marine nummulitics, which doubtless in part represent the
Subdthu group, but probably cover a larger period of time.

In the Salt range the principal member of the marine nummulitics
is a band of fine, compact, grey or white, limestone, frequently cherty, of
some 400 or 500 feet in thickness, which is unconformably overlaid by
upper tertiary beds. Below the limestone there is from 50 to 100 feet of
soft variegated shales or clays, with one or more coal seams. ‘The clays
are pyritous and decompose readily on exposure, the decomposed shales
being burnt and employed in the manufacture of alum. These shaly beds
contain a number of fossils, but the collections have not yet been examined
in detail. Cardita beaumonti is, however, known to occur, and marks the
age of the beds as lower eocene at latest,

North of the Salt range, in the hills of the Hazara district,! the oldest
member of the tertiary system is a great thickness of dark bluish grey,
or blackish, limestone, with brownish olive shales. The rock is generally
foetid and massive, sometimes distinctly, and sometimes obscurely, strati-
fied. The distribution of this hill type of nummulitic limestone, as it has
been called, is peculiar. It forms a broad belt throughout the Hazdra
and Murree hills, from the neighbourhood of Abbottdébdd past Murree, and
along the spurs traversed by the Grand Trunk Road north-west of Rawal-
pindi. The same rock forms the greater part of the Chittapahér range,
and is continued west of the Indus in the Afridi hills, which are principally
composed of this formation.

The tract so defined lies immediately to the north of a great line of
disturbance, accompanied by much faulting and displacement of the strata,
which runs westwards from the neighbourhood of Murree. Though less
sharply defined, this zone of disturbance corresponds to the great faulted
boundary which separates the pretertiary rocks of the Himdlayas from the
tertiary formations of the sub-Himdlayas, and is in the main a line of sepa-
ration between the pretertiary deposits and the hill type of nummulitic

1The account of the tertiary beds of the | in Records, X, 107, (1879); XII, 114, 208,

North-West Punjab is derived from Mr. A. B. | (1881).
Wynne’s papers, of which the principal are
Chap. XIV.] NUMMULITICS OF THE PUNJAB. 353

limestone on the one hand, and the newer tertiaries on the other. ‘The de-
marcation is not absolute, for outliers of the newer beds are found to the
_ north, and inliers of the hill nummulitic limestone to the south of the line.

The most important of these inliers is that of the Khaire Miérut ridge,
south-west of Rawalpindi, Here, as in the hills north and west of Murree,
the hill nummulitic limestone is overlaid by an upper nummulitic group,
composed of grey, red, and deep purple clays or shales, associated with
masses of gypsum, and alternating with thin bands of limestone. The com-
position of the group varies ; sometimes the limestones and at others shales
are most developed in the lower part, and at times there are but few cal-
careous beds. Strong zones of yellowish grey sandstones are found in the
western localities, while eastwards hard grey sandstones and purple clays,
resembling those of the overlying group, are said to be included. Except
in the sandstones, which are unfossiliferous but for a few fucoid mark-
ings, marine fossils are abundant, among which Nummulites, Oporculina,
etc., are common, None of the fossils have been critically examined, but
the age of the beds is evidently eocene, and probably upper eocene.

There is still some doubt as to the equivalence of the marine nummu-
litics west of the Jehlam with those to the east, that is to say, it is uncertain
whether the thicker series to the west merely represents the same period
of time as the thinner series to the east, or amore extended one. Although
the fossils of the Jammu inliers have not been examined, the presence of the
ferruginous bottom bed and associated coal seam, together with the general
similarity of the type of deposit and relations to the Dagshdi beds above,
leave little room for doubt that the Jammu nummulitics are the equivalent
of the typical Subdthu group. -

It would be natural also to regard the shaly marine beds, immediately
underlying the sandstones and red clays of Dagshai type west of the Jeh-
lam, as of contemporaneous origin with the typical Subdthus, but as the
distinction between the two groups is merely due to the cessation of marine
conditions of deposit, there is an uncertainty as to whether the change took
simultaneously on all the sections, and it is possible that the upper portion
of the nummulitics west of the Jehlam were formed contemporaneously with
part of the typical Dagshdi group, while the hill type of nummulitic lime-

stone may be merely a deep water formation of the same age as the typical
Subdthus.

In this connection the Salt range section is of some importance. The
presence of coal seams in the shaly beds at the base of the tertiary rocks
naturally suggests their equivalence to the Subéthu group, and it is even
stated that the pisolitic lateritic bottom bed of the Subdthus is found in the
Salt range. There is, however, some doubt as to this last. It is not pos-
sible to recognise this bed with certainty in Mr. Wynne’s description, and
354 GEOLCGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV.

it is not evident whether the only specific statement of identity is based on
actual observation.!

The mere presence of coal seams is no proof of contemporaneity, and
what paleontological evidence is available points to an older age for the
Salt range nummulitics than that of the Subéthu group. The presence of
Cardita beaumontt, a species characteristic of the passage beds between
the cretaceous and nummulitic in Sind, in the shales below the main lime-
stone of the Salt rauge, shows that they cannot well be later than oldest
eocene in age. Of the 46 species described by Messrs. D’Archiac and
Haime from the nummulitics of the Salt range, (3 are found in the
RAnfkot, group of Sind, 14in the Kirthar, of which 6 are also found in the
Rdnikot, while 3 are found in Gdj and Nari beds. From the Subdthu bed
49 species were described, of which not one is also found in the Salt range,
a difference of fauna which must be almost entirely due to the Subdthu bed
having been deposited in shallow muddy water, while the Salt range
species inhabited a deeper and clearer sea. The same cause that led to
the distinction between the fauna of the Subdthu and Salt range nummu-
litics restricted the number of Sind species found in the former, and we
find that there are only 10 species in all, of which 2 are Rdnfkot, 7 Kirthar
and t Gdj. The general facies of the fauna is consequently, so far as the
more limited evidence allows us to judge, newer than that of the Salt range
nummulitics.

The upper portion of the nummulitic series is wanting in the Salt range,
where the clear nummulitic limestones are unconformably overlaid by the
upper tertiaries. It is not clear how far the hill type of nummulitic lime-
stone to the north may represent the nummulitic limestone of the Salt
range. Palzontological evidence is wanting, and though there is a consi-
derable lithological diversity, there is none that could not be accounted for
by local variations in the conditions of deposition and by the greater dis-
turbance that the northern beds have undergone.

Whether there isan exact equivalence in time of the two types of num-
mulitic deposits or no, the existence of marine conditions in the Salt range,
at a period anterior to the formation of the typical Supéthu group, deprives
us of any compulsion regard to the nummulitic series west of the Jehlam as
coeval with that further east, and it is on the whole more natural to adopt
Mr, Wynne’s original correlation of the upper shaly beds in the westeru
area with the typical Subéthus,? and to look on the underlying limestones
and associated beds as older, and unrepresented on the Subdthu section.

In the north-west Punjab and in the Jammu hills the eocene beds are
conformably overlaid, and pass by interstratification into a great thickness
of red and purple clays, with interbedded grey or purplish sandstones,

» Records, 1X, 54, (1876). I 2 Records, X, 109, (1877).
Chap XIV.J AGE OF THE MURREE BEDS, 355

precisely similar in lithological compesition as in stratigraphical position to
the Dagshdi and Kasauli groups of the Sirmur area. The ridge on which
the hill station of Murree stands, as well as the hills to the south-east of it,
are composed of these beds, which have consequently been known as the
Murree beds.!. They have, however, been traced to the Ravi, where
they are directly continuous with a band of rocks, originally regarded as
representing the Ndhan or lower portion of the upper tertiaries, but
now recognised as belonging to the Sirmur series. There is thus a
direct continuity of outcrop, as well as a similarity of stratigraphical po-
sition with regard to the marine nummulitics, which leaves little recom for
doubting that the Murree beds represent the Dagshdi and Kasauli groups
of the Simla region. What little possibility of doubt there might remain is
removed by the discovery of the palm Saéa/? major at Kasauli, and
in the Murree beds in the Jehlam valley,? and by the recognition of the
Kasauli plant bed in the Ravi valley?

The plant in question ranges from lowest to middle miocene in
Europe, and as it is found in india near the top ofa series of beds inter-
mediate between the eocene and thetower Siwdlik beds, which are regard-
ed as upper miocene on independent grounds, it will be seen that the
stratigraphical position of the species is similar in the two areas, and that
the Dagshdi and Kasauli groups may be regarded as covering the oligocene
and lower miocene periods of European geology.

Apart from these fossil plants no organic remains are known with
certainty to occur in the Murree beds, any more than in the Dagshdi group
in its typical area. Bone fragments, crocodilian scutes and exogenous
fossil timber are found in certain beds, resting on the nummulitic lime-
stone of the Salt range, which Mr. Wynne regarded as forming part of
his Murree group.* It is, however, very doubtful whether these beds can
be regarded as the equivalents of any portion of the Dagshdi or Kasauli
groups, or of the Murree group as originally defined, in spite of the very
indefinite nature of its upper limit. There is a well marked uncons
formity between the nummulitie limestone and the beds immediately over-
lying it, whose importance Mr. Wynne was inclined to minimise, but
it certainly represents a considerable lapse of time, as is shown by the
occurrence of conglomerates composed of pebbles of the underlying num-
mulitic limestone at the base of the upper tertiaries, by the complete
cutting out of the eocene beds at either end of the Salt range, and by
the occurrence, as determined by Mr. Theobald, of Mastodon latidens and
Rhinoceros palztndicus in a fossiliferous zone, about 100 feet above the
nummulitic limestone. These two animals mark the age of the beds in
which they are found as upper miocene at oldest, and show that a great

1A. B, Wynne, Records, VII, 66, (1274). 3H. B. Medlicott, Records, 1X, 52, (187A).

20. Feistmantel, Records, XV, 51, (1882).| 4 Records,'X, 119, (1877).

2A
256 GECLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV.

interval of time must have elapsed after the deposition of the middle eoceze
nummulitic limestone, an interval which would te filled by the oligocene
and lower miocene Dagshdi and Kasauli groups.

How far the lithological distinction between the Dagshdi and Kasauli
groups is maintained in the Jammu and Hazara hills isnot clearly determin-
able from the published descriptions, but as the Murree beds are said to
pass with perfect transition into that great series of upper tertiary deposits
known as the Siwdlik series, it is probable that there is a distinction
between the lower and upper portion, analogous to that between the two
groups in the Simla region,

The name Siwdlik, originally applied to the range of hills separating
-\the Dehra Dun from the plains, has been extended by geographers to the
fringing h’lls of the southern foot of the Himalayan range, and applied by
geologists to that great system of subaerial river deposits which contains
remains of the “ Fauna antigua stvalensis.” In spite of local variations
of texture, inevitable from their mode of formation, these upper tertiary
beds of the Siw4lik series maintain a great uniformity of type along the
whole length of the Himalayan range.

Lithologically the lower portion of the system is characterised by a
great thickness of fine grained grey, micaceous, pepper and salt sandstone,
interbedded with clay bands near its lower portion, while the upper part
of the system is composed of soft earthy clays, undistinguishable from the
alluvium of the plains except by the disturbance they have undergone, and.
coarse conglomerates, of well rounded pebbles and boulders, of . crystalline
and metamorphic rocks derived from the Himdlayan ranges.

In the neighbourhood of Ndhan this system was originally divided
into two. members,! a lower, to which the name of Ndhan was ap-
plied, and an upper, to which the name Siwdlik was restricted. In
this area the boundary between the two yroups is a great fauit, but
there must be a real, if local, unconformity, for the upper Siwalik con-
glomerates contain numerous. pebbles? of the Ndhan sandstones they:
are faulted into contact with. The distinction between the Ndham and
Siwdlik zones appears to be well maintained in a south-easterly direc-
tion as far as the borders of Nepal, but to the north-west it Misa ppeats,
and there appears to have been a continuous series of deposits, ranging
from the bottom to the top of the upper tertiary formations. No fossils
have yet been found in the typical Ndhans, though it would appear that
they do occur,’ but to the north-west representatives of the Siwdlik fauna

1H. B. Medlicott, Memoirs, UI, pt. i, pp. 17, ® See H. B. Medlicott, Memoirs, YH, pt. its

101, (1864). Pp. 16, (1864) ; Records, XIV, 71, footnote
* HB Medlicott, Records, X1V, 172, (1882). | (:864),
Chap. XIV.) NAHAN GROUP, - 357

occur low down im the series, iw beds which very possibly represent the
Ndhan group as originally defined. Under these circumstances it has
been found inadvisable to retain the separation between Ndhan and Siwélik,
and the former are now classed as lower Siwélik, though the term
may be retained as a useful local designation for a particular type of
formation,

The Ndhan group is composed of alternating beds of a fine grained,
usually grey, firm sandstone, and of clays, usually bright red in colour
and almost always some shade of red or purple, which weather in a
nodular manner. The clays usually prevail im the lower part of the group
and the sandstones in the upper.

The lithology of this group resembles very closely that of the Dag-
shdi group, and one might be tempted to regard them as equivalent to
each other. The equivalence cannot be absolutely disproved till the
area west of the termination of the typical lower Himalayas, inthe Kdngra
valley and the Jammu hills, has been examined in greater detail than has
yet been done, but in the meanwhile there are good feasons for sup-
posing that the lithological similarity between the two groups is due to
a similarity in their condition of deposition and does not mean contem-
porancity of origin, In the first place the two groups are found in distinct
areas, separated by a marked structural feature, exhibiting itself atthe pre-
sent day as a fault of many thousand feet throw. As will be shown in a
subsequent chapter, this fault—commonly known as the main boundary—is
connected in a peculiar manner with the elevation of the Himdlayas, and it
is highly improbable that the beds exposed south of it are of the same age
as those found to the north. Another argument depends on the fact that
no exposure of the Subdthu group has been found even in the deepest
cut sections of the typical Nahan group, and a third may be derived from
the smaller degree of induration, indicating, though not proving, a younger
age. In the country north of Ndhan town, where the Néhan and Dagshai
groups are brought into contact with each other, on opposite sides of the
main boundary fault, the sandstones of the former always weather into
soft rounded lumps, while the Dagshdi sandstones weather into angular
fragments, which have lost the sharpness of their angles, but exhibit a
much less degree of weathering than that to which the N4han beds’ have
undergone, Finally the red clay beds which have already béen mentioned
as occurring at the top of the Kasauli group, though they differ some-
what from the typical Nd4han clays, resemble them sufficiently to point to a
return of the conditions of deposition which prevailed in the Dagshdi and
Ndhan periods, and suggest that on an unbroken section the Ndhan would
-be found to overlie the Kasauli group.

No fossils have been described from the typical Ndhans. It is possible
that some of thé lower Siwdlik fossils found in the north-west Punjab may

x 2A 2
358 GEOLOGY OF INDIA—HIMALAYAN TERTIARY, [Chap. XIV.

have been derived from beds of the same age, but the supposition lacks
proof. There seems however to be little room for doubt that Sir Proby
Cautley did find fossils on the northern side of the hill on which the town
ot Ndhan stands, and consequently in the beds of the Ndhan group, but the
specimens were lost before they had been examined by a palzontologist.

The Na&han group is succeeded, on those sections where the sequence
is complete, by an immense thickness of soft sandstones, generally coarser
in grain and more micaceous, mostly of a pepper and salt grey colour,
with some interbedded bands of earthy clay, occasionally slightly tinged
with red, but never assuming the bright red colours of the Ndhaa clays.
The argillaceous beds are, for the most part, confined to the lower part of
the group, the middle part being usually composed of some thousands of
feet of sandstones, without any intercalation of shale on the one hand, or
any included pebble on the other. In the upper part of the group strings
of pebbles occur among the sandstones, which become more numerous till
bands of conglomerate appear and increase in abundance and coarseness,

The uppermost group of all varies very much in character. Near the
large rivers draining from the central Himdlayas, it consists principally
of coarse conglomerate, composed of rounded boulders of the harder rocks
of the Himalayan chain. In the intermediate stretches of ground it is
composed largely of soft earthy beds precisely similar to those of the
modern alluvium of the plains.

The details of this variation in lithology of the upper Siwdlik beds, as
well as the structural features of the Siwdlik zone, are of great interest, but
their principal interest lies in their bearing on the age and elevation
of the Himdlayan chain, and they will consequently be considered in the
chapter devoted to that question. At present it will be sufficient to point
out the conditions under which the Siwdlik series was deposited. The
earlier observers regarded this great series of beds as having been deposited
in a sea, a supposition which is sufficiently disproved by the complete
absence of any marine organisms, and by the occurrence of the remains of
fresh water molluscs, fishes and tortoises. It is hardly possible that they
could have been deposited in a fresh water lake, for it is not conceivabl2
that a fresh water lake extending the whole length of the Himalayas could
have existed. Moreover, the fresh water organisms whose remains have
been found are all such as inhabit streams, and not lakes. But the most
conclusive proof of all lies in the evident unity of the whole SiwAiik series,
pointing to the whole of it, with the possible exception of the Ndhan group,
having been formed under very similar conditions, while the very close
resemblance between the upper Siwdlik beds and the recent deposits of
the Gangetic plain leaves little room for doubt that the Siwdlik beds were
deposited subaerially by streams and rivers,

The thickness attained by the Siwdlik series is immenge. Mr. Wynne
Char. XIV. SIWALIK FAUNA. 359

estimated it at 14,000 feet in the north-west Punjab. In the Siwdlik
hills there are at least 15,coo feet of beds, and the series is by no
means complete, and' similar vast thicknesses may be measured on any

section.

The few mollusca which have been found in the upper Siwdliks belong
solely to fresh water or terrestrial forms, and the first comparison that was
made! was carried out under circumstances so unfavourable, with so poor a
collection of recent species from India, and at a time when the latter were so
imperfectly known, that but little weight can be attached to the conclusions
formed. The majority of the specimens obtained are in poor preservation,
but allthe forms collected from upper or middle Siwdlik beds, since the
recent fresh water shells have been better known, have proved to be
either identical with living species, or closely allied tothem. Amongst
those hitherto identified, the only land shell is Bulzmus ¢nusudaris,? a
species which ranges at the present day from Africa to Burma, whilst
amongst fresh water molluscs; the two common Indian river snails Pa/u-
dina bengalensis and P. dissimilis. have been recognised, and forms of
Melania, Ampuilaria and Unio also occur3

So fat as the evidence extends, therefore, the few mollusca of the
Siwdliks tend to show that the beds must be of upper and middle tertiary
date. But the evidence afforded by the mollusca is imperfect, and both
closer comparison anda larger series of fossil specimens are desirable
before any very positive assertions can be made as to the antiquity of
the Siwdlik series, on the dataafforded by the invertebrata. In investi-
gating the question of age, we are consequently forced to depend, first
upon the vertebrata, and especially the mammalia, and secondly upon
such geologicat evidence of connection with other formations of known

_ age as the rocks afford.

The first question, then, is the homotaxial relation of the mammalian

1 E. Forbes, in Falconer's Palzontological | examined by him were too imperfect to enab’e

Memoirs, London, 1886, I, p. 380.

2 Theobald, MS.; Geoffrey Nevill, Records,
XV, 106, (1882). As in thisand the followng
paragraph Mr. Theobald’s view of the affinities
of the Siwdlik mollusca has been accepted in

preference to the high authority of Prof. E..

Forbes, it is only just to say that Mr. Theobald
has a far more extensive knowledge of living
Indian fresh water shells than it was possible
for any naturalist in Zurope to acquire at the
period wlen Prof. E. Forbes’ note was written.
Indeed, it is evident from Prof, Fo. bes’ remarks
that the collections of recent Indian shells

 

him to form a competent opinion. Mr. Benson,
a better authority on this particular subject
than Forbes, considered the most, if not
the whole, of the Siwdlik mollusca identi-
cal with existing species (Falconer: Palz-
ontolozica’ Memoirs. I, pp. 26, 181). Of three
Species in the Survey collections, two were
identified by Mr. Geoffrey Nevill with living
forms.

3 Some extinct fuviatile mollusca have been
found in the lower Siwalik strata of the Bug-
ti hills associated with a vertebrate fauna of
miocene facies. See supra, p. 319.
360 GEOLOGY OF INDIA—HIMALAYAN TERTIARY. [Chap. XIV,

fauna which has been obtained from the two upper groups of the Siw4liks
series, ‘The true age of this fauna, whether miocene or pliocene, was at one
time disputed and, though there is happily no further controversy regard-
ing this point, the question presents points of sufficient interest to claim
somewhat extended notice,

The following list of genera, with the number of species of each that
are known, comprises all that have been found in the typical Siwdlik
area, excluding the fauna of the Manchhar beds in Sind and of Perim
island which appear to belong to an older period than the fossiliferous
beds of the sub-Himdlayas; genera still living are distinguished by an
asterisk :—

PRIMATES—
* Troglodytes, 1; * Simia, 1; *Semnopithecus, 1 5 *Macacus, 1; *Cynocephalus, 2.

Carnivora
* Mustela, 13 *Mellivora, 2; Mellivorodon, 1; *Lutra, 3; Hyenodon, 13; Ursus, 1 ;
Hyenarctus, 33 Amphicyon,1; *Canis, 23 *Viverra, 2; *Hyena, 43 Lepthy-
@na, 1; Hyenictis, 1; Eluropsis, 1; Elurogale, 13 *Felis, 5; Machaeratus, 2.

PRoBuscIDEA—
*Elephas, 6 ; *(Euelephas, 1, *Lorodon, 1, Stegodon, 4); Mastodon, 5.

UneuLata— i :
Chalicotherium, 1 5 *Rhinoceros, 3; *Equus,1; Hipparion, 2; *Hippopotamus, 1;
Tetraconodon, 13; *Sus, 5; Hippohyus, 1; Scnitherium, 1 3 Merycopotamus, 33
*Cervus, 3; Dorcatherium, 2; *Tragulus, 1; *Moschus, 1; Pr pal@omerys, 13
*Camelopardalis, 13 Helludotherium, 1; Hydaspitherium, 2; Sivatherium, 33;
* Alcelaphus, 1 3 *Gasella, 1;*Cobus, 2; *Antilope,1 ; Hippotragus, 1 ;* Oreas(?) 1;
4Stvepsiceros, (2) 13 Boselaphus, 1; Palaeoryx, (?) 13 Hemibos, 1; Leptobos, 1;

*Bubalus, 2;*Bison, 1; Bos, 3; Bucapra, 1; *Capra, 2; *Qvis, 13 *Camelus, 2,
—_—_— .

Ropenr1—
* Mus (Nesokia), 1; *Rhyzomys, 13 *Hystrix, 15 *Lepusy 1.

AvEs—
* Phalacrocorax, 1; *Pelecanus, 2; *Leptoptilus, 1; *Mergus, 1; *Struthio, 1;
* Dromaeas, 1.

REPTiILiA—
Crocodilia—* Crccodilus, 1; *Gharialis, 33 Rhamprosuchus, 1.
Lacertilia—* Varanus, 1,
Chelonia—Colossochelys, 1; *Damonia, 13 *Bellia, 2; * Kachuga, 3; *Hardella,
1; *Emyda, 4; *Trionyx, 13; * Chitra, 1.
PIsces—

* Carcharias, 133 Ophiocerhalus, 13 *Clarias, 1; *Heterobrauchus, 13 *Chrysichthys,
13 * Macrones, 1; *Rita, 13 *Arius, 1; *Pagarius, 1.

Only very imperfect information exists as to the exact horizon in the
Siw4lik series at which the bones of a large proportion of the species
have been found, but the great majority are from the upper and middle
Ohap. XIV.] AGE OF THE SIWALIK FAUNA, 36t

Siwdliks, none in the typical area being known to occur in the lower or
Ndhan subdivision. It is, however, by no means certain that some of the
specimens from the north-western Punjab are not derived from beds of
the same age asthe Ndhan group, and it is highly probable that some
other forms with middle tertiary affinities would be found to be confined
to lower Siwdlik beds, if the precise horizon of all the bones collected were
known!

On the other hand, one pleistocene form, Bos (Bubalus) pa/lzxindicus,
has been found in the highest Siwdlik strata, asscciated with Camelus
sivalensis, Colossochelys, etc.; and two species of elephant, belonging to
the subgenus Stegodon, viz. £. insignis and E. ganesa, range throughout
the upper Siwdliks, and recur in the pleistocene deposits. The species
of proboscidians generally appear to have had a more extensive range,
both in space and time than most of the forms belonging to other mam-
malian orders, but Bos palzindicus is an animal of exceptionally recent
aspect, even in the pleistocene mammalian fauna, since it is only distin-
guishable from the living Bos dudalus (Bubalus buffelus v. B. arni, auct.)
by comparatively trifling and unimportant osteological details. It must
evidently have been a very near ally, and in all probability the not very
distant progenitor, of the buffaloes which now inhabit the Ganges valley,
Assam, and parts of the Central Provinces of India.

The Siwdlik forms, however, which might be excluded on account of
belonging to an older or a newer fauna, and of being supposed, on more or
less strong evidence, to be confined to either the lowest or the uppermost
portions of the series, are too few in number to affect the general facies,
and there are unquestionably several miocene types and some pleistocene
species found in the highly fossiliferous upper Siwdlik beds. It is best
therefore, for the present, to include all the forms enumerated.

Proceeding then to classify the genera of mammalia above given,
it will be found that thirty-nine, comprising seventy-one species, still exist
(the livirg species being, however, different in nearly every case), whilst
twenty five with thirty-seven species, are extinct.

Of the extinct genera, excluding those that are purely Indian, @/urogale
and Hyznodon are found in oligocene beds, the latter ranging into the mio-
cene ; of the miocene genera, Dorcathertum and Amphicyon are not known
from newer beds, while Wyznarctus, Chalicothertum, and Hipparion range
into the pliocene, Wacherodus and Mastodon into the pleistocene ; He//a-
dotherium and Palgoryx are purely pliocene forms, while the distinction
of Hemibos from the living Budalus is very doubtful.

Of the other extinct forms, not known out of India, two (A¢ppohyus and

1A very large proportion of the Siwdlik | tors, and of course the precise locality of the
remains have been obtained by native collec- | bones is in most of these cases doubtful.
362 GEOLOGY OF INDIA—IIIMALAYAN TERTIARY. [Chap. XIV.

Merycopotamus) belong to the less specialised types characteristic in
general of the older and middle tertiaries. Several others, such as 7e¢tr¢-
conodon, with its enormously developed premolar teeth, and the huge
four-horned Stvatherium, differ widely from anything now existing, but,
being highly specialised forms there is nothing in their organisation to
indicate that they are of earlier age than newer tertiary.

Amongst the recent genera represented in the Siw4liks, ten, viz. A7ustela
Felis, Canis, Viverra, Lutra, Rhinoceros, Sus, Cervus, and Hystrix
are known to range as far back as upper miocene, and in one or two cases
even further ; twelve, viz. Macacus, Semnopithecus, Ursus, Hyzno, Elephas,
Equus, Hippopotamus, Camelopardalrs, Gazella, Bos, Capra, and Mus,
are known from the European plioceue’ beds, but not earlier; whilst
Troglodytes, Stmia, Cynocephalus, Mellivora, Cobus, Antilope, Oreas,
Strepriceros, Capra, Ovis, Came/us, and Rhizomys, have hitherto only
been found in recent or pleistocene deposits, outside of India.

This examination of the relations between the Siwdlik genera and the
distribution of similar forms in European tertiaries leads, as might be
anticipated, to a somewhat uncertain result. The proportion of living to
extinct genera is greater than is found in most miocene deposits, but not
more than appears to exist in the characteristically middle tertiary ossi-
ferous beds of Sansanin France. The presence of four extinct genera
not known to range above the miocene period elsewhere, is contrasted
with the occurrence of twenty-five genera not found elsewhere at a lower
horizon than pliocene or pleistocene. There is perhaps rather more pro-
bability that early forms, like Dorcatherium and Hyznodon, should have
survived longer in India than they did in Europe, just as rhinoceroses,
tapirs, and elephants still exist in the tropics, associated with a fauna
amongst which they appear antiquated and out of place, than that
such eminently specialised types as Wacacus, Bos, Capra, or Equus,
should have lived in miocene times, but the argument is of small value,
for the miocene Cervus and Antilope were in all probability as highly
specialised, or nearly so, as the Siwdlik genera. The fact, however, that
the recent genera contain more species than the extinct forms is of some
importance, since it is probable that types which were dying out would be
represented by fewer species than those which were sup planting them,
and which might fairly be credited with the vitally important power of.
producing distinct specific stocks by variation. A stronger argument for,
the newer age of the Siwdlik beds is to be found in the close approxima-
tion between some of the mammals and the living species of the same
genera, the most remarkable of all being the connection already noticed

1 Gervais, Zoologie et Paiéontologie Frangaises, 2nd ed., Paris, 1859, p. 338.
 

Capra sivaleuisis, Lyd, x,

  

Rhinoceros sivalensis, f. & C, Camelopardalis sivalenisis, F’. & C.
right upper molar X 4. left upper molar

 

Cynocephalus subhi-
Mmalayanus. ‘Dheger, right
upper roolar,

 

 

Hyocenarctos paleindicus, Lyd. right upper molars and premolars Hyoena colvini lryd right upper

premolars.
SIWALIK FOSSILS.

Calcutta Phototype Co,
 
  
   

  
 
 
 

 

fo FS)

‘ vs { Lhe

aaa « We
RAY Ke

mY

 
    

‘SE

       

Oe
Pe ~
Ce

 
 

Ag

Stegodon ganesa. F, & C. X YK

 

Jfastodon falconeri, 2nd right lower molar X 4. Mastodon latidens, 2nd right upper molar X 4.

  

Dinotherium indicum, F. left lower molar X% 4, Sus hyssdricus F. & C right upper moars.

SIWALIK FOSSILS.

Calcutta Phototype Co,
Chap. XIV.]  STRATIGRAPHICAL RELATIONS OF THE SIWALIKS.
between the fossil buffalo of the uppermost Siwdlik strata, that of the
pleistocene Jumna and Narbadd beds, and the common Indian species now
existing.

363

The anomalies exhibited by the mammalian fauna taken as a whole would
doubtless disappear to a considerable extent if we knew the exact horizons
from_which the fossils were obtained, those forms with miocene affinities
being more characteristic of the lower beds and those closely allied to
recent forms of the upper. Yet, after allowing for this, when we consider
that the great bulk of the fauna was obtained from what is practically a
single group of beds, and that the life of a species is probably in every case,
and certainly in most, more extensive than the period represented. by the
particular beds in which its remains are found, there remainsa strong prob-
ability that there was an admixture of types not found in beds of later than
miocene age in northern Europe with the more typically pliocene forms
which prevail in the Siwdlik fauna,

The evidence afforded by the reptiles, birds, and fishes is even more
decidedly in favour of attributing a later data than miocene to the
Siwdlik beds. Among the reptiles only two out of twelve genera, and
only eleven out of twenty species are extinct, the remainder, AHarde/la
thurgi, Damonia hamittont, Kachuga lineata, K. tectum, K. dhongoka,
Chitra indica, Emyda vittata, Gha:talis palustris, are all found living
at the present day,’ while Crocodz/us sivalensis seems undistinguishable
from the recent -C. palustres.

The six genera of birds are all represented in the living fauna of the
world, and the same holds good of the fishes, no extinct genus of either of
these orders having been determined in the Siwdlik fauna. Moreover, two
fishes (M/achrones aor and Bagarius yarrel(t) are still living.

Putting together the data derived from the fauna asa whole, it is im-
possible to deny that the balance of evidence is in favour of a pliocene
age* and this view is supported by some stratigraphical data. As
the approximate age of the Siwdlik rocks is a necessary element in
any argument founded upon their faura, it will be best to show how the

*R. Lydekker, Records, XXII. 58, (1889). one Of the tertiary periods of Europe” (Palce-

3 This appears at first sight to be in direct
opposition to Dr. Falconer’s conclusions, but a
study of his writings leaves it doubtful
whether he ever expressed any decided con.
viction on the subject. He repeatedly noticed
the close connection between some Siwdlik
forms and those now found in India, and ap-
peared for along time rather disposed to consi-

der that the tertiary mammalia of India “ lasted
through a period corresponding to more than

ontological Memoirs, I, p. 28) tnan to class
the Siwdlik fauna with the miocene of Europe.
In his later writings he certainly spcke of the
Siwdlik fauna as miocene, but on'y incident-
ally, whilst in some of his latest papers he ar-
gued in favour of man having been a probable
sentemporary of Colvssochelys and the Si-
wdlik mammaljia—an idea which it is diffi-
cult to reconcile with the miocene age of
the fauna,
364 GEOLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV.

stratigraphical evidence supports the view that these beds are of pliocene,
and not of miocene age, before proceeding to notice some other interest-
ing points of connection between the Siwdlik and other faunas, recent

and extinct.

The stratigraphical, as distinguished from the purely homotaxial
relations just discussed, depend entirely upon the connection between the
typical Siwdliks and the Manchhar beds of Sind. The position of the
latter has been already described, and it was shown that’ the
Manchhar beds, comprising where thickest but little less than 10,000
feet of strata, rest upon the G4j group, whose age is determined
by the far more satisfactory data afforded by marine organisms, and is
shown to be more probably upper than lower miocene. The lower
Manchhar beds pass down into the Gj rocks, so that it is reasonable
to believe that no important difference of age exists between the
two, It has also been shown that the mammalian fauna of the lower
Manchhars, although containing several species in common with the
Siwdliks, is altogether older in aspect, and that the majority of the forms
hitherto recognised belong to the peculiar types of even toed ungulates
allied to Merycopotamus and Anthracotherium, intermediate in character
between pigs and ruminants, and characteristic of the miocene
epoch. In these lower Manchhar beds there is also found a form of
Dinotherium, miocene type unknown in the Siwdliks proper. though
found (the species being distinct) in the beds of Perim island. Now,
there can be no reasonable doubt that the Manchhar beds of Sind,
as a whole, correspond with the Siwé4lik formation of northern India,
for the two are portions of one continuous band of upper tertiary rocks, and
viewed in this light, the relations of the faunas are very striking, the
fossiliferous lower beds of the Manchhar group corresponding to the un-
fossiliferous Néhans, and the almost unfossiliferous upper Manchhar beds to
the ossiferous strata of the Siwdliks. If, therefore, the lower Manchhars of
Sind are upper miocene, so is the N&éhan group of the Punjab, and it is
impossible, either on stratigraphical or paleontological grounds, to class
the fossiliferous middle Siwdliks lower than pliocene, the upper Siwdliks,
which contain Bos (Bubalus) pala@indicus, being probably upper plio-
cene. Briefly stated, the evidence is that the Siw4lik fauna is newer
than the Manchhar fauna, and found in higher beds, and that the Manchhar
fauna is not older than upper miocene.

It is impossible to enter at length into the detailed relations of the
Siwdlik fauna with the forms found in various newer tertiary strata. There
is however one Eurcpean fossil fauna which is of singular interest from
its resemblance to that of the Siwdlik beds, In this collection of extinct
Chap. XIV.] SIWALIK AND PIKERMI FAUNAS. 365

mammalia, which was first discovered at Pikermi in Attica,’ and has since
been found at Samos and Mytilene on the coast of Asia Minor, at Baltavar in
Hungary and at Maragha east of lake Urmia in northern Servia, not only is
there aremarkable admixture of typically miocene forms with other species
which have a later aspect, but there is the same remarkable abundance of
true ruminants, amongst them several species of Girafide and of Anti-
lope, as in the Siwdliks. In miocene strata, although ruminants occur, they
are in general but little, if at all, superior in number to the other
artiodactyle ungulates, but in the Pikermi beds (including only the forms
found at Pikermi) there are fifteen ruminants to one’pig and one Chalico-
thertum ; in the Siw4lik fauna thirty-seven ruminants and Lut twelve other
artiodactyle ungulates. Another point of similarity in the two faunas
is the absence of small mammals.

The following is a list of the genera found in the beds of Greece, with
the number of species belonging to each genus :—

PRIMATES—
Mesopithecus, t.

CaRNIVORA~

Simocyon, 1; Mustela, 1; Promephitis, 15 Ictithertum, 33 Hyena, 1; Hyenarctus, 1;
Hyeanictis, 1; Macherodus, 1; Felis, 4.

PROBOSCIDEA==
Mastodon, 23; Dinotherium, t.

Uncutata PEerissopacTy La— .
Rhinoceros, 3; Aceratherium, 13 Lepto.lon, 1; Hipparion, r.

UneGuaTtsa ARTIODACTY LA—
Sus, 1; Chalicotherium (Ancylotherium), t; Dremotherium 2; Antilope, 33 Palaotra-
gus, 1; Pal@oryx, 2; Tragoceros, 2; Pal@oreas, 13 Antidorcas, (?) 1; Gazella, 1;
Camelopardalis, 1; Helladotierium, tr.

RopentTia—
Mus (Acomys), 1; Hystrix, tr.

Of birds, a Phastanus, a Gallus, and a Grus have been identified; of
reptiles, bones of Zestudo and Varanus.

Of the above twenty-nine genera of mammals eighteen, including
Helladotherium,* are found in the Siwdliks of India, besides which the fauna
bears in many respects the same similarity to that of Africa at the present
day as the Siwdlik mammals bear to their living Indian representatives.
Now, this Pikermi fauna has been freqently quoted as upper miocene, and its
connecticn with the miocene beds in other parts of Europe is unmistakable,

1 Gaudry: Animanx fossiles et Géologie de 2 See Fcrsyth, Major, Proc. Zool. Soc., 1891,
l'Attique, 4° Paris, 1862. P- 323.
366 GROLOGY OF INDIA-—HIMALAYAN TERTIARY.

[Chap. XIV.

no less than fifteen species being undistinguishable from those found in
various miocene deposits. Several of these species are doubtfully identified,
but amongst the number are such characteristic forms as Macherodus cul-
tridens, Mastodon turicensis, and Hipparton gractle.- But, as M. Gaudry
points out in the clearest manner, the ossiferous beds of Pikermi contain
at their base, and below the horizon whence the bones have been obtained
a layer with pliocene marine fossils, and all the beds containing the bones,
together with the pliocene marine beds, rest unconformably on lacustrine
miocene rocks. There can be, therefore, no reasonable doubt that the
Pikermi fossils, like the middle Siw4liks, are of pliocene age, and that the
quotation of them as miocene is an error.t At the same time the absence
of some characteristic living genera, such as E/ephas and Gos, gives a
somewhat older facies to the Pikermi than to the Siwdlik fauna.

The points of similarity between the European miocene faunas and
the animals now inhabiting either tropical Asia or Africa south of the
Sahara may be due either to migration, and survival in a more favour-
able climate, or to the fauna having been formerly more uniform over
large areas, and to the modified descendants continuing to live in one
region, whereas they have died out and been replaced by distinct types
in other parts of their old province.? On the latter hypothesis we may
suppose that the fauna of central Europe and Malayasia was more or
less uniform in the lower miocene period, and that Greece and Africa
formed a single zoological province in pliocene days, but that the gibbon-
like apes, Zupatade and other Malay types, died out in centra! Europe,
and the giraffes, antelopes, etc., in Greece, whilst the descendants of their
relatives survived in the Malay countries and Africa respective'y. The
theory of migration presents, on the whole, fewer difficulties, and is im ac-
cordance with the little we know of the Indian miocene (Manchhar) fauna,
in which living tropical forms appear to be less represented than in the
deposits of that age in Europe, It is not unreasonable to suppose that
some of the forms named, and especially the rumiaants, migrated into
southern Asia at the close of the miocene period.

' For the theory adopted by M. Gaudry to
account for the survival of these miocene
animals in pliocene times, see “Animaux
fossiles et Geologie de l’Attique,” p. 343. It
appears simpler to believe that the miccene
fauna of Europe migrated to the southward,
and that many species survived in Greece after
they had died out north of the Alps, Hence
the admixture of pliocene and miocene types,
A further contribution to the question, by
M. Gaudry, was pubhizhed in 1886, Bull. Svc,

Geol. Franc,, 3rd series, XIV, 288.

7 It is assumed in the present and in other
arguments employed in this work that similarity
of organisation implies relationship by descent,
i.e. that animais having similar structure are
descended more or less remotely from the same
ancestors. The theories of evolution, and
of origin of species py descent with modifica»
tion, are now so widely accepted amongst
naturalists that it is unnecessary to explain or
defend them.
Chap. X1V.] SIWALIK AND RECENT FAUNAS. 367

‘It is true that amongst the marine invertebrates there is a well marked
resemblance between the miocene genera of Europe and living tropical
forms. The Indian and African land faunas of the early and middle ter-
tiaries are as yet too imperfectly known for any comparison to be made
between them and those of the same epoch in extratropical regions, It is
not improbable that there may prove to have been a greater similarity
than exists amongst the terrestrial forms living at present, and it is also
probable that, if such similarity existed, it will he found to have consisted
mainly in the greater richness of the extratropical fauna in middle tertiary
times, and in a number of types now extinct or confined to the tropics
having been represented in both tropical and extratropical zones of cli-
mate. This last probability is founded on the fact that the temperature
of Europe in the miocene epoch was in all probability nearer to that of
the present tropics than to the temperate climate of recent times, and that,
consequently, whole families of animals, and of plants intolerant of cold,
then ranged to much higher latitudes than they now do.

It is by no means an improbable inference that the representation of
European miocene genera in the Indian Siwdliks is due to changes of cli-
mate in later tertiary times and to a migration of the fauna towards the trop-
ics, There is good reason for believing that Europe and south-eastern Asia
were connected by land after the eocene period, and as it is certain that,
a great portion of the disturbances affecting the Himdlayan strata are of
pliocene or postpliocene date, it is reasonable to conclude that, at the close
of the miocene epoch, the mountain barrier which now separates the Indian

-peninsula from Central Asia did not exist, or was so much lower that it
afforded little or no obstacle to migration,

But the immigration of the European miocene forms may not be the only
way in which only the Siwdlik fauna was affected by the secular refrige-
ration of the earth’s surface, culminating in the glacial epoch, It is true
that there is a considerable amount of similarity between the Siwdlik fauna
and that of India at the present day, but, nevertheless, there is a very
striking distinction—a distinction due less tc change and replacement
than to disappearance. Even after making allowance for the fact that the |
whole assemblage did not exist contemporaneously, there is nothing so
striking in the fauna of the Siwdlik epoch as the wonderful wealth and
variety of forms. It must be recollected that we know little or nothing of
the smaller mammals, and that animals of size inferior toa pig or a sheep
are scarcely represented. It would be premature to infer that, as at the pre-
sent day, the more minute forms exceeded the larger types in abundance,
for the conditions of intermediate ages may have affected the more bulky

' This will be found to agree with the con- | of the Himdlayas derived from other conside.
clusion regarding the probable date of origin | rations. Sce Chap. XVIL
368 GEOLOGY OF INDIA—HIMALAYAN TERTIARY. (Chap. XIV.

animals far more than the minute Rodentia, [nsectivora, Chtroptera, etc.
Still it is only reasonable to suppose that the ancestors of the present M/iero-
mammalia lived in the same profusion as they do now, and it is incre-
dible that the living rodents and insectivores can play the parts on the
modern stages, and fulfil the functions, of the great ungulates and carnivores
of past times. Comparing like with like, and especially passing in review
the Carnivora, Proboscidéa, and Ungulata, all represented, and all,
except the Proboscidea, well represented in the living fauna of India,
indeed better than in most other parts of the world at the present day, it
is impossible not to be struck with the comparative poverty in variety of
the existing mammalian types. We have of course but an imperfect
knowledge even of the larger Siwdlik animals, and remains of Carnivora
are rare, 80 much so that probably many species remain undiscovered,
but even at present the known Siwdlik carnivores are more numerous
than the living forms of similar size in the same area, and the ungulates
exceeded their living representatives in number in the proportion of more
than three to one, there being fity-six known Siwdlik species and only
eighteen recent, The superior wealth of the older fauna is both generic
and specific; not only are the types more varied, but there is a greater
variety of forms in many of the genera, and no less than eleven extinct
elephants and mastodons are represented by a solitary living form. Even
such modern types as Bos have dwindled in numbers from six to two.
This great impoverishment of the recent mammalian fauna is not
peculiar to [ndia. It is found in other parts of the Old World and in
America, wherever remains of animals have been preserved in sufficient
quantities amongst the deposits of the later tertiary epochs for a good
idea of the fauna to be presented. In the words of Mr, Wallace, ‘We
live in a zoologically impoverished world, from which all the largest and
fiercest and strongest forms have recently disappeared ;’’ and he makes
the happy suggestion, that this enormous reduction in the numbers of
the greater mammals is due to the glacial epoch. Thus, we have an addi-
tion to the arguments urged in the first chapter,® ‘in favour of India
having been affected by a cold period in the geologically recent past.

1 Geographical Distribution of Animals, I, ® Supra, p. 14.
p- 150.
CHAPTER XV.

LATERITE,

General characters and composition:—Distinction of high level and low level laterite —
Distribution and mode of occurfencé —Theories of the origin Resumé.

All who have paid any attention to the geology of India must be
familiar with the term ‘laterite,’ and no one cati have travelled far in India
without meeting with the substance itself, which is still one of the stum-
bling blocks of Indian geology. Although it is difficult to conceive
that a rock, so widely spread in India and Ceylon, and said to be extensively
developed in Malacca and Sumatra, while some occurs in Burma, can be
peculiar to these regions,’ it is uncertain if anything precisely similar has
hitherto been detected elsewhere, It is almost invariably a surface form.
ation, and according to some observers, nothing but,a form of soil; yet it be-
comes. an important formation from the very large area in India, which it
superficially covers, and a treatise om Indian geology would be imperfect
without a full deseription of the rock.

The order in which its description comes is governed by the fact that
while there can be but htthe doubt that some forms of laterite are
in process of formation at the present day, others date from tertiary, and
perhaps from eocene, times, and as the rock is usually unfossiliferous it
appears best to describe all the varieties together before proceeding to
the description of the posttertiary rocks.

The description of laterite, given in many geological works, is far from.
accurate, althouyh the rock has been well described by several Indiaw
geologists.? In its normal form it is a porous argillaceous rock much

k Voysey states that it is found at the Caps
of Good Hope. It ig a noteworthy fact. that
no laterite has been detected in Abyssinia,
where the rocks throughout a large area ci
country are precisely similar to those of the
Bombay Deccan. In map No. 4 of Berghaus’
Physical Atlas, laterite is rerresented as cover-.
ing nearly a quarter of the dry land of the earth.
The term is, howevec, used in a diffcrent sense
to that here applied.

2 It would be difficult to give a description
of any rock more clear and accurate than

Newbold’s account of the laterite of Bidar, Four,
i As, Soc. Beng, XIN, 989; (7844); Four. As. Soc.
Beng. XIV, 299, (1845) and Four. Roy. As.
|, Soe, WALL, 227 (1846). Tne descriptions of
laterite scattered through the writings of various
Indian geologists are too numerous.to quote.
Amongst the more important are the following—
’ Buchanan, Jousney from Madras through My-
gore, Canara, and Malabar, London, 1807, II,
rp. 440. Stirling, As. Res, XV, 177, (1825);
Christie, Edin. New Phil. Four., VI, 117,(1829);

   
a7°

GEOLOGY OF INDIA—LATERITE,

[Chap. XV.

impregnated with iron peroxide irregularly distributed throughout the
mass, whose composition may be gathered from the following analysis of a
very richly ferruginous variety from Rangoon,'—

Sotusee 1n AciIDs.

Peroxide of iron. F . .
Alumina . * . - ‘
Lime 3 é ‘ is ‘
Magnesia . . . 7 .
Silica : ‘ ‘ ‘ ;

46°279.
5'783
"742
‘ogo
*120

InsoLuBLe iN ACIDS.

Silica (dissolved by potash)
Silica (by fusion) .

Lime, iron, and alumina 3
Combined water alkalies and lose

.

The iron exists e:ther entirely

6°728
30°728
2'728
6802

. . e e .

 

100°000

in the state of hydrated peroxide

(limonite) or else partly as hydrated and partly as anhydrous peroxide,
The surface of laterite after exposure is usually covered with a brown or
blackish brown crust of limonite, but when freshly broken, the rock is

and Mad. Four. Lit., Sci, 1V, 468, (1836);
Calder, As. Res., XVII, 4, (1833) ; Cole, Mad.
Your. Lit, Sci., 1V, 10, (1836).; Voysey, Four.
As. Soc. Beng., XIX, 273; (1850); Kelaart,
Edin. New Phil. Four., LIV, 28, (1853);
Carter, Four. Bo. Br. Roy. As. Soc., 1V, 199,
(1852); V, 264, (1857); Aytoun, Edin. New
Phil. Four., and series, UV, 67, (1856); Buist.,
Trans, Bo. Geog. Soc. XV, p. xxii, (1859).

The subject has also been frequently treated
in the publications of the Geological Survey,
especially Memoirs 1, 69, (1856), 265-280,(1859);.
li, 78, (1860) ; IV, 260, (1864) ; X, 27, (1873);
XII, 200, 224, (1876) ; XIII, 222, (1877); XVIII,

-

 

122, (1881); XXIV, 217 and 239, (1890); where
an account of all previous notices is: given.
See also Records, XV, 93, (1882) ; XXII, 220,
(1889).

1 Your, As. Soc. Beng., XX, 198, (1853).
The result given is the mean of three an-
alyses made in the laboratory vf the School of
Mines, London. The following are assays of
the quantity of iron contained in the portion of
laterite soluble in acids. The analysis of first
fiveand.No. 8 were made by Mr. Mallet for
the first edition of the present work; the other
three are from the paper on the laterite of
Orissa (Memoirs,1I, p. 288) :—

Percentage of

Percentage of
metallic iton,

iren peroxide,

. High-level latesite overlying Deccan trap, Amarkantak  35°6 50°8
2. Ditto from Main Pat, Sargaja - 5 é 16°6 237
3. Ditto from Baplaimali plateau, K4ldhandi,

; south of Samba!pur . . 15 21°4
4. Ditto from top of Moira hill in the Kharakpur

range, south of Monghyr ‘ ° 28° 0"

5, Ditto from Mahudgarhf hill, RAjmabl hills ie se
6. Laterite (high-level) from Kathidw4r, Western India ‘ 22°8 325
7- Low-level laterite, from Daltola, Cuttack, Orissa : ‘ 24'5 349
8, Ditto from near Cuttack ‘ ‘ : . 25°6 36'5
9. Ditto from Tanjore . é . . . 23'4 33°4
Chan. XV.] LATERITE AND LITHOMARGE. 37t

mottled with various tints of brown, red, and yellow, and a considerable
proportion sometimes consists of white clay. The difference of tint is
evidently due to the segregation of the iron in the harder portions, the
pale yellow and white portions of the rock, which contain little or no iron,
being very much softer, and liable to be washed away on exposure. Oc-
casionally the white portions have a brecciated appearance, and consist
of angular fragments ina ferruginous matrix. In this case the rock has
not unfrequently a compact texture resembling jasper, but it is never sc.
hard as a purely silicious mineral.

The iron peroxide not unfrequently occurs in the form of small pisolitic
nodules, which are sometimes employed as iron ore. Veins and nests of
black manganese have been observed by Newbold in some laterites of the
Deccan.!

In many forms of laterite the rock is traversed by small irregular
tortuous tubes, from a quarter of an inch to upwards of an inch in dia-
meter. The tubes are most commonly vertical, or nearly vertical, but
their direction is quite irregular, and sometimes they are horizontal. They
are usually lined throughout with a crust of limonite, and except near the sur-
face are often filled with clay. Besides these, there are sometimes horizontal
cracks, occasionally expanding into small cavities, and giving an appearance
of irregular stratification to the formation. In the more massive forms of
laterite some horizontal banding is usually present, the cavities beneath the
surface being mostly filled by more or less sandy clay. When first quar-
ried, the rock is so soft that it can easily be cut out with a pick, and
sometimes with a spade, but it hardens greatly on exposure, ;

The exposed surface, whether vertical or horizontal, is characteristic
and peculiar. It is extremely irregular, being pitted over with small
hollows, caused by the washing away of the softer portions, and gener-
ally, though not always, traversed by the tubes and cavities just described.
At times it is so much broken up by small holes as to appear vesicular,
whilst the crust of limonite forms a brown glaze, often mammillated or
botryoidal, so that the rock has a remarkably scoriaceous appearance
and bears a very curious resemblance to an igneous product. It is not
surprising that many observers should have looked upon laterite as vol-
canic, for not only does it often present this remarkable superficial resem-
blance to a scoriaceous lava flow, but it is found, in several parts of India,
associated with basalt and other igneous rocks. Laterite, however, as will
be shown presently, is never an original form of igneous rock, It is in all
cases either produced by the alteration of other rocks, sometimes igneous
sometimes sedimentary or metamorphic, or else it is of detrital origin.

The laterite frequently appears to pass into the underlying rock,

! Jour. As. Soc. Beng., XII, go2, (1844).
372 GEOLOGY OF INDIA—LATERITE. [Chap. XV.

whether this be igneous, metamorphic, or sedimentary. In the case of basalt
or gneiss underlying laterite the upper part is decomposed, forming a
clay, which becomes a kind of lithomarge passing by insensible grada-
tions into laterite itself, through its impregnation with iron by the water
trickling througii the laterite above. In fresh sections, where a detrital form
of laterite is the overlying rock, the limit of the two can usually be traced
without difficulty, but surfaces which have been exposed for a length of
time are generally covered with more or less of the limonite glaze and the
lithomarge can no longer be distinguished from laterite.!| This lithomarge
is always more ferruginous above than below; it varies in colour from red
through yellow to white, being usually mottled, not frequently coloured
purple or lilac in patches, and a few pipes often occur, apparently produced
by the percolation of water.

Another form of lithomarge, found beneath the laterite in many places,
and especially to the northward, consists of hardened clay, sometimes
sandy and generally highly ferruginous, which shows no tendency to pass
into the underlying rock, although it usually exhibits unmistakable transition
into the laterite above. In these cases, the laterite and lithomarge together
form a group of beds superposed, as a rule unconformably, upon
older rocks of various kinds. In some instances, as in Bundelkhand, this
infra-lateritic formation contains pebbles,? and there is every reason for
believing that it is a rock of sedimentary origin. In some cases the
present form of lithomarge contains hematite or limonite in quantities
sufficient to enable the inineval to be collected for iron ore, as in Bundel-
khand, near Jabalpur, and on the eastern flanks of the Rajmahdl hills. 3

One peculiarity possessed to an eminent degree by all forms of laterite
is the property of broken or detrital fragments being recemented into a mass,
closely resembling the original rock. Laterite itself has great powers of
resisting atmospheric disintegration, being produced by long action of the
atmosphere upon various ferruginous clays, but the underlying formation
decomposes, is slowly washed away, and the originally horizontal cap of
laterite, falling down, b:comes reconsolidated on the irregular surface,
which it still covers. Another form of reconsolidated laterite is composed
of broken fragments, washed down by rain and streams to a lower level,
at which they become recemented.

The surface of the country composed of the more solid forms of la-
terite is usually very barren, the trees and shrubs growing upon it being
thinly scattered and of small size. This infertility is due, in great part,
to the rock being so porous that all water sinks into it, and sufficient
moisture is not retained to support vegetation. The result is that laterite
plateaux are usually bare of soil, and frequently almost bare of vegetation.

1 Memoirs, 1, 283, (1859). | 8 Memoirs, Il, 81, (1860); XIII, 241, (1877).
2 Memoirs, Il, 84, (1890). :
Chap. XV.J | HIGH AND LOW LEVEL LATERITE, 273

Of course, this barrenness is not universal, soil sometimes accumulates
on laterite caps, and some of the more gravelly or more argillaceous
varieties support a moderate amount of vegetation. Still the general
effect of the rock is to produce barrenness.

Several writers have divided the laterite into two forms, high level and
low level laterite, the former of. which was supposed not to be, the other
to be, of detrital origin, and by some it has-been urged that the term should
only be applied to the latter of these,

The high level form, which is found cappivg the summits of hills and
plateaux on the highlands of central and western India, is a rock of
fine grain, and apart from the irregular distribution of the iron it contains,
fairly homogeneous in structure; it is not sandy, and only exceptionally
shows any indication of a detrital origin. This type Mr. Foote! has pro-
posed to distinguish as iron clay, a term used by Voysey, one of the
earliest observers, but also used by him in describing the low level detrital
laterite of Nellore.-

The low level laterite, which covers large tracts in the neighbourhood |.

of both coasts, on the other hand, frequently contains grains of sand and
pebbles, imbedded in the ferruginous matrix. It is, as a rule, less homo-
geneous than the high level form, and passes by insensible gradations into
sandy clay or gravel with a very small proportion of iron, especially in the
exposures that have been classed as laterite on the east coast, many of
which have little or no claim to the name if it is to be used in any litho-
logical sense, On the west coast the exposures are more truly lateritic in
their nature, and there are large areas of reck which do not appear to be
detrital in their origin and are undistinguishable, except by position, from
the high level laterite of the Deccan.

This fact shows that no hard and fast distinction can be drawn
between the high and low level laterite, put there is undoubtedly on
the whole a difference in age and origin between the two types, and
those geologists who consider that the name of a rock should distinguish
not only its composition and structure, but also its mode of origin, are
justified in refusing to use the same word to designate both. At the same
time there is nothing in the description of the original propounder of the
name to indicate that he restricted the word to one form rather than the
other. There is good reason to suppose that some of the laterite in the
district he examined is, and some is not, of detrital origin ; moreover, the
word has by convention come to be used so generally as an ill defined but
convenient term, descriptive of the constitution and nature of the rock,
irrespective of its mode of origin, that it would be inconvenient, if not
impossible, to attempt a restriction of its meaning. But, though it is im-
possible to distinguish between the high level and low level laterites, if by

1 Memoirs, XU, 201, (187¢).
2B2
374 GEOLOGY OF INDIA—LATERITE. LChap. XY,

those terms there is implied any theory involving the supposition of an
essential difference in age or mode of origin, yet the terms are convenient
when they merely refer to the position occupied by the rock, whether on
the highlands of the interior, or on the lowlands near the coast.

The fact is, that it is frequently impossible to determine whether the
laterite is detrital or not, a difficulty well shown by the beds which occur
interstratified with lower tertiaries near. Surat. These beds are sedi-
mentary, for marine fossils were found in one of them, yet they are not in
the least sandy, and they resemble the laterite or iron clay of the Bombay
Deccan so closely that, when these laterite beds were first examined by
the Geological Survey, they were supposed to be volcanic rocks altered by
surface action. The mistake, it should be added, was partly due to an
apparent intercalation of basalt and nummulitic limestone, subsequently
found to be caused by faulting.

The high level laterite is chiefly developed on the Deccan plateau,
especially on the highest portions of the Sahyddri range, and of the
spurs running from the Ghats. It forms a cap on the uppermost traps
exposed on the plateau, being also found at lower elevations, but the
lower beds are, as a rule, of small extent or thickness. The summit bed,
as it is termed by Mr. Foote,’ is not more than 50 to go feet in thickness
in the southern Maratha country. It is about the same at Mahdbaleshwar,
but from too to 200 feet at Bidar north-west of Haidardbéd. It occurs
at varying heights above the sea, 4,700 feet at Mahabaleshwar being probably
the highest point, whilst at several places in the southern Mardthd country
it is found capping ridges and isolated hills from about 2,000 to nearly 3,500
feet above the sea, always overlying the highest lava flowin the country.
At Matheran, near Bombay, and on some neighbouring plateaux, there are
caps of laterite at about 2,000 feet above the sea, but these do not
belong to the summit bed, as the traps on the tops of these hills are
not high in the ‘series. Some- of the laterite caps are very extensive.
The bed at Bidar is said to be twenty-eight miles long from west-north-
west to east south-east and twenty-two miles broad, and the area of late-
rite at Kaliani, forty miles west of Bidar, is of even greater extent.

The greater portion of the trap area in the Deccan has not been closely
examined, but so far as is known, laterite is of rare occurrence, except
near the Western Ghats and in the southern Maratha country. A few
very small caps are found south-west of Ndgpur, in south-east Berar, and
probably similar small outliers occur here and there along the south-eastern
margin of the volcanic region. Farther to the north-east, laterite occurs
at Amarkantak and on the eastern outliers of the Deccan traps, at Mdin
Pat and Jamfra Pat in Sdrgdja, being from 100 to 200 feet thick on the

» Memoirs, XII, 203, (:876). 2 Tene Four. Roy. As, Soc. Vill, 228,
46).
Chap. XV.1 DISTRIBUTION OF HIGH LEVEL LATERITE, 375

former. North of the Narbadé also, in Rewd, Bundelkhand, and in other
states as far west as Gujerdt, laterite is found, sometimes as much as
200 feet in thickness, capping outliers of the trap series.

In all the localities hitherto mentioned laterite occurs resting upon the
_Deccan traps, but the high level laterite overlaps the traps, rests upon older
rocks, and is found in places some hundreds of miles beyond any existing
outlier of the volcanic series. Instances of this kind have been noticed
by: various observers in the southern’ Maratha country,’ the same laterite
bed being apparently sometimes ‘continued from the trap surface on
to the transition or metamorphic rocks, whilst numerous outliers on
the older formations are known to exist. Caps are said also to occur
at high elevations on the Dambal or Kappatgod hills, east of Dhér-
war, and on hills in the neighbourhood of Bellary and Cuddapah.? More to
the north-east, in the high grounds of Patnd, Kdldhandi, Bastdr, Jaipur,
etc., between the Mahdnadi and Godavari, caps of laterite, 50 to 100 feet
thick, occur on many of the higher hills® at elevations of between 2,000
and 4,000 feet above the sea. The most eastern exposure known to occur
in this neighbourhood is on the Kopilas hill, about 2,050 feet above the
sea, and 12 miles nearly due north of Cuttack. To the northward a
great expanse of laterite is found on the Chutid Nagpur plateau at eleva-
tions varying from 2,000 to 3,000 feet above the sea in several places, and
especially to the north-west of Jashpur ;° it caps ridges and peaks in the
usual manner, but differs from the usual high level laterite in covering hills
and valleys alike, and is probably in part a reconsolidated formation.
Leaving, for the moment, the Rdjmahdl hills, which require separate
notice, a thick mass of laterite occurs at an elevation of 1,500 feet on
Moira hill, the highest peak of the Kharakpur range. Turning thence
westward, caps of the same rock are found, outside of the trap area, at
several places in Bundelkhand,® and at two near Gwalior,’ alt on the highest
ground of the country.

Besides the above mentioned localities laterite has been reported to
occur on some of the hills of Southern India, but ferruginous clays have
possibly been described under the name of laterite, which have little of
its true character. Such is the case with the Nilgiris, one of the localities
mentioned by several geologists, No well authenticated occurrence of
laterite is known at an elevation exceeding 5,000 feet above the sea.

There is, however, a very important bed of this rock on the R4j-.

3 Newbold, Four. As. Soc, Beng., X'I, 906,{ 4* The information of the occurrence of laterite
(1844); Four. Roy. As. Soc, VIII, 228, (1846); | on Kopilas hill was obtained by Mr. Ball from

Foote, Memoirs, XII, 205, 217, (1876). Dr. Stewart, of Cuttack,
2Newbold, Four. Roy. As.. Soc,, VIII, 228, 5 Ball, Records, X, 170, footnote, (1877).
(1846). ®t, B. Medlicott, Memoirs, II, 82, (186c).

3 Ball, Records, X, 169, (1877). * Hacket, Records, III, 41, (1870).
376 (Chap. XV.

mahal hills in Bengal,! These hills, tike the highlands of the Bombay
Deccan, are composed of bedded basaltic traps, and, as in the Deccan, the
very highest bed consists of laterite, Mahudgarhi, the highest plateau in
the range, 1,655 feet above the sea, being capped by this formation. The
laterite in the R4jmahal hills is, in places, as much as 200 feet thick and
it slopes gradually from the western scarp of the hills, where it attains its
highest elevation, to the Gangetic plain on the east. Here, too, there is
sometimes, as in the Deccan, an apparent passage from basalt into laterite,
but the latter rock to the eastward is distinctly identical with the low
level laterite of Bengal, and is clearly of detrital origin, whilst, even at
considerable elevations in the hills, fragments, derived from the shales
interstratified with the basaltic flows, are found imbedded in the laterite,
so that, no distinct line having ever been drawn between the beds at differ-
ent elevations, we appear in this case to have a passage from the high
level into the low level laterite, and reasons for supposing that both were
originally of sedimentary origin. The case, it should be remembered, is
not clearly proved, the laterite of the Rajmahal country not having been
specially examined with a view to test the connection between the beds to
the eastward and those to the westward, but the two appear to be parts
of the same formation, and it is certain that both are in this instance
detrital?

The evidence hitherto collected is insufficient to justify the conclusion
that the high level laterite once formed a continuous bed, occupying the
whole surface of the Indian peninsula from the Ganges valley to the neigh-
bourhood of Madras, but the manner in which caps now occur upon isolated
peaks and ridges clearly shows that they were once much more extensive,
and that only the remnants have been left undenuded. It is difficult, in pre-
sence of the great amount of denudation which has taken place since the
laterite caps were part of a more extensive formation to escape the conviction
that the high level laterite must be of considerable geological antiquity.

GEOLOGY OF INDIA—LATERITE

Before proceeding to discuss the very difficult subject of the origin of

laterite, it willbe best to point out

1 The laterite has been but briefly noticed;
see T. Oldham, Four. As. Soe. Beng., XXIII,
273, (1854) ; Ball, Memoirs, XIII, 222, (1877).

2 There is a possibility that the connection
between the high level laterite amd the low
level laterite of the eastern coast is not con-
fined to the solitary instance of the Rajma-
hal hills, although no other equally well mark-
ed case of passage can be traced, and in
some cases, as at Kopilas near Cuttack, the
difference in level is very great; but the low
level laterite of the eastern coast rises gra-

.

the general distribution of the low

dually from the neighbourhood of the sea, at
a slope which, if continued inland, would con-
nect the bed with the high level formation.
The latter is of greater antiquity than the low
level bed, but the process of formation may
have been continuous, the rock now found at
a higher level being first formed, that at a
lower elevation being gradually consolidated
as the lower portion of the country was raised
above the sea. It should be remembered that
the higher part of the country was, in all pro-
bability, never depressed below the sea levels
Chap. XV.] DISTRIBUTION OF LOW LEVEL LATERITE. 377

level laterite, especially in the neighbourhood of the coast. On the west
coast of the Peninsula laterite has not been observed in the Konkan,
or lowlands, north of Bombay,! it appears, a little farther to the south-
ward, between Bombay and Ratnagiri, and extends thence throughout
large tracts of the low country, intervening between the Sahyddri range
and the sea, as far as Cape Comorin. It does not, of course, cover the
whole surface. In many places it has been cut away by streams, so that
the lower formations are exposed, and in parts of the country it appears
tobe wanting. The greater part of the region, however, has never been
geologically mapped, and very few details of the distribution of laterite are
available.

In the country between Ratnagiri and Goa the rock appears to
form a plateau, having a general elevation of 200 to 300 feet above the
sea. Qn the coast it terminates in cliffs, the trap being exposed
beneath it. The plateau extends for from 15 to 20 miles inland, and is
cut through by numerous rivers and streams, in all of which the trap is
exposed, the lignite and clays, which were mentioned in the last chapter,
-being found between the laterite and the traps at Ratnagiri. Farther in-
land the laterite is found at a higher elevation than near the coast, so that
the rock appears to have a low slope towards the sea. The laterite is
distinctly of detrital origin, and even conglomeratic in places, the thick-
ness is considerable, but no exact measurements have been recorded, except
at Ratnagiri, where it amounts to 35 feet, probably less than the average.
It is evident that the plateau formerly extended much farther to the east-
ward, and it probably covered the whole of the country as far as the base
of tne Sahyddri range, for caps of laterite are found in places on the trap
hills, and masses, reconsolidated from the detritus of the principal beds,
are found at lower levels.

South of Malwdn the underlying rock is no longer trap but gneiss, or
some other metamorphic formation. ‘The laterite, which is extensively
developed, appears to be similar to that of the Bombay Konkan. In
Travancore it overlies the fossiliferous tertiary beds.

On the east coast of India laterite occurs almost everywhere, rising
from beneath the alluvium which fringes the coast, and sloping gradually
upwards towards the interior, but this laterite is, as a rule, a much less
massive formation than the rock of the western coast. It is seldom more
than 20 feet in thickness, and is often represented by a mere sandy or
gravelly deposit, not more than four or five feet thick. Where it is thicker
the lower portion usually consists of lithomarge, produced by the alteration
of the underlying rock. The laterite is frequently conglomeratic, and

This idea of the whole laterite being one{ !' Except near Surat, where the outcrops are
continuous formation appears to have occurred | of nummulitic age. The rock differs from
to Newbold.—Four. Roy, As, Soc. VAL, 240, | all superficial laterite, in being distinctly in-
(1846). tercalated between other beds.
378 GEOLOGY OF INDIA—LATERITE. (Chap. XV,

includes large rounded, or subangular, fragments of gneiss and other rocks,
good instances being found at Trichinopoli, at many places near Madras,
amongst which are the Red Hills, seven miles to the north-west of the city,
and around the detached hills north-west of Cuttack, in Orissa. In the
Madras area quartzite implements of human construction have been found
in the laterite in considerable numbers.?

The fringe of laterite is of very unequal width. In places it forms a
broad, low slope, stretching for many miles from the edge of the alluvium ;
in ‘ethers it only remains as caps upon the older rocks. Jn one form
or another it appears to be traced, at short intervals, from Cape Comorin
to Orissa, and thence northward through Midnapur, Rardwén, aad
Birbhim, to the flanks of the RAjmahdal hills, where it is well developed and,
as already noticed, it appears to pass into the high level laterite.

The low level laterite is not confined to the neighbourhood of the coast,
It is frequently found in patches over many parts of the country, but these
patches are rarely of large size and they often appear to be due to local
Conditions, such as abundance of iron in the rocks, or reconsolidation of
fragments derived from a bed of high level laterite. Many such lateritic
deposits are rather of the nature of ferruginous gravel than of true lateritic.
‘The small pisolitic nodules, so characteristic of some forms of laterite, are
found abundantly in the older alluvium of the Ganges valley, and in many
other superficial deposits in the plains of India, and whenever they are suf-
ficiently abundant, appear to become cemented, with the accompanying
sand and clay, into a rock closely resembling laterite in many of its pecu-
liarities.

{n Burma, laterite of the detrital low level type is found in places
on the edge of the alluvial tracts of the Irawadi and Sittaung rivers in
Pegu and Martaban, forming as usual a cap to other rocks, and having a
very low dip towards the river from the sides of the valleys. The laterite
appears to form the basement bed of the post-tertiary gravels and sands,
and laterite gravels, apparently derived from the denudation of the mas-
sive laterite are largely dispersed through the older alluvial deposits.

West of the Irawadi only a few patches of laterite occur in the Myan-
aung district, but the rock is more common along the western foot of the
Pegu Yoma. To the east of that range laterite is generally wanting, but
there is a well marked belt of this formation along the base of the meta-
morphic hills east of the Sittaung river, forming a plateau which rises 50
or 60 feet above the alluvium of the Sittaung valley. Some laterite is
also found in Tenasserim, whence it extends into the Malay peninsula.

Having thus stated, as briefly as is consistent with the object of afford-

'R.B. Foote, Memoirs, X, 27-58, (1879).
Chap. XV.) MODE OF ORIGIN. 379

ing a tolerably complete account of the rock, the distribution and mode of
occurrence of the different varieties of laterite, the question of the manner
in which this rock has been formed, must next be considered. The
subject has already been noticed as difficult, the difficulty arising from
the fact that the rock has evidently undergone a considerable amount of
change, both chemical and structural. The difference between laterite,
when first cut from the quarry and the same rock after exposure, is
well marked. The rock becomes harder, and the hardening appears
not merely due to the desiccation of the argillaceous constituents, but
also to a change taking place in the distribution of the peroxide of
iron, the change being shown by the colour becoming darker, and by the
surface being covered with a glaze of limonite. Whether the anhydrous
iron peroxide, which occurs in some forms of high level laterite, be-
comes converted by exposure into hydrated peroxide, has not been
ascertained, but it is quite clear that the process of segregation of the
iron has tended greatly to obscure any structure which may have exist-
ed originally in the rock, and that this segregative action is constantly
in progress, It has already been stated that iron has been dissolved out
of the laterite and redeposited in the underlying lithomarge, where the
latter is merely an altered form of the rock beneath, and it is a common
circumstance to find pisolitic nodules of hydrated iron peroxide, evidently
due to segregation, in some forms of laterite. These facts, and the pro-
cess by which the surfaces of the rock, and of the tubes by which it is
traversed, become coated with a glaze of limonite, render it evident that a
transfer of iron oxide from one part of the rock to another is continually
going on.

One view, which has been held by several good observers and has
been strongly supported by Mr. Foote’s examination of the laterite or
iron clay in the southern Maratha country, is that the high level laterite
is simply the result of the alteration zm sz/u of various forms of rock, and
especially of basalt, by the action of atmospheric changes. Many of the
dolerites of the Deccan contain iron in the form of magnetite, and large
quantities of magnetic iron sand are found in the beds of streams which
flow over the traps, whilst bands, both of magnetite and hematite, are lo-
cally common in the metamorphic rocks. The gradual change from dole-
ritic trap into laterite has been noticed by several observers,! and so far as
the Deccan alone is concerned, the evidence in favour of laterite being
merely the result of atmospheric change acting upon very ferruginous vol-
canic rocks, appears so strong that, if there were no conflicting phenome-
na, it might be accepted as a satisfactory explanation. At the same time
there are some difficulties, to which attention was first called by Captain

1 Voysey, Four. As. Soc, Beng., XIX, 274, (1850); Foote, Memoirs, XU, 202, (1876).
380 , GEOLOGY OF IND!A—LATERITE. (Chap. XV,

Newbold,! and although Mr, Foote* has shown that they are not insuper-
able, they must not be overlooked, because the apparently sedimentary
origin of the rock, in Bundelkhand and elsewhere, tends to invalidate
the conclusion that the high level laterite is merely the result of surface
change.

The main argument in favour of supposing the high level laterite of
the Deccan to be merely altered basalt, is that the two rocks are
seen to pass into each other. This fact, which is unquestionably
established, may be considered proof that laterite’ may result from
the alteration of basalt or a similar rock, but it is, of course, insufficient
evidence to show that such is the origin in all cases. It is always
possible that the upper portion of the laterite is, in each case, of ex-
traneous origin, and that the surface of the basalt beneath has been
affected by the infiltration of iron, in the manner already described when
explaining the origin of lithomarge. Numerous instances are found, on
the other hand, in which the laterite rests upon the surface of basalt,
which is either hard and unaltered, or soft and decomposed, without any
appearance of a passage from one rock to the other. But this, again, is
no proof that the laterite above the unaltered trap is not itself the result
of alteration of a different lava flow, the rock beneath not being sus-
ceptible of the same change. It is clear that the evidence afforded by
the circumstance that basaltic trap sometimes passes into laterite, and
sometimes does not, is insufficient to decide the question as to whether
the latter is derived from the former by a process of chemical alteration.

It has been stated that magnetite occurs in many of the Deccan
basalts, but until far more analyses have been made, it is impossible to
say whether any of these rocks contain as large a proportion of iron as
the laterite. It is probable that some may, but it is certain that so large
a proportion of iron as 15 or 20 per cent.® in any basalt is exceptional,
yet this is not above the average amount in the Deccan laterite. At the
same time the larger percentage may perhaps be explained by a process
of concentration, some of the other constituents of the rock having been
removed, in the manner explained further on, but not the iron.

One difficulty, to which especial attention was drawn by Captain New-
bold, is the complete absence in the laterite of those nodules, large or
small, of various forms of silica, such as agate, jasper, and crystalline,
quartz, so frequently found in the different forms of trap. It is difficult
to understand, if laterite simply results from the alteration 7m situ of the
Deccan basalts, why no amygdaloidal structure, especially where the
amygdules contain so indestructible a mineral as agate, should be detected

1 Four. As. Svc Beng., XIII, 098, (1844); Four. 8 That is of metallic iron; 15 per cent. of
Roy As. Soe., VIII, 238, (1846). iron corresponds to 19°3 per cent. of protexile,
Memoirs, XII, 203, (1876), and 21-4 per cent. of sesquioxide.
Chap. XV.] HYPOTHESIS OF ORIGIN IN SITU. get

in the altered rock. Mr. Foote suggests! that, in the case of the summit
bed, which appears to rest upon the highest traps, the absence of
amygdaloidal structure may be due, in the first place to the lava flow,
having been of a peculiarly dense nature,” and secondly to the fact that,
being the uppermost flow, the water which percolated it did not contain
silica in sufficient quantity .to form silicious nodules in the vesicular
hollows. He also points out that the underlying bed into which the
summit bed laterite is seen to graduate in several sections, is a very
argillaceous rock without vesicular cavities or enclosed minerals,

One conclusion is clear, if the high level laterite of the Deccan has
been produced by the alteration 7” sétu of volcanic. rocks, only particular
varieties of such rocks are capable of undergoing the alteration. If all
were similarly liable to be converted into laterite at the surface, the
occurrence of that rock would be more general, and less restricted to
particular elevations. The great difficulty in the way of explaining the
origin of the high level laterite, so widely spread in MAlw4 and the Deccan,
by asimple process of atmospheric alteration is, in brief, that the hypo-
thesis demands the occurrence, over an enormous area of country, of a
volcanic rock, whether a tuff or a true lava flow is immaterial, of peculiar
and unusual composition, containing a much larger proportion of iron than
usual, and wanting the amygdaloidal structure, so common in the Deccan
traps. This difficulty, it must be remembered, is, so far, only a reason for
caution in coming to a conclusion, and does not show that the hypothesis of
alteration 7” széu is impossible.

The great extension of the laterite beyond the trap area might
be explained by supposing that the highest volcanic stratum covered a
wider surface than any of the inferior lava flows, but this theory is unten-
able in some cases, for instance in that of the Gauli plateau, south of
Belgaum,’ where a bed of laterite at a lower level than the summit bed
was traced by Mr. Foote on both sides of the Mahddéyi ravine, passing
into the underlying trap to the northward and into metamorphic rocks to
the south, as is represented in the sketch section fig. 20, given on the fol-
lowing page. In this case, the southern portion must have been formed from
gneissic rocks if the laterite be the result of alteration alone, and it
is difficult to understand how two rocks, so totally dissimilar in constitution
as basalt and gneiss, can have produced precisely the same rock, by a
simple process of disintegration 7 sztu.

On the other hand, the difficulties in the way of supposing the high
level laterite to be sedimentary are considerable. The idea of its being
a marine deposit can scarcely be entertained, as there is not a shadow of

3 Memoirs, XIil, 203, (1876). rare, although they do not, as a rule, preserve

2 Such dense beds do certainly occur in the | their non-vesicular character over large areas.
Deccan traps, indeed they cannot be said to be Memoirs, XII, 217, (1876).

,
&
20
1)

Phonda Ghat

Hills south of
Musda

Malwan

   
      
     

LE

SZ

IAS

mth
= SSS

Mmm TT

    
  

mn

met
WOLGibaa

7
N

   
 

ro

    

Ee

WK

gneiss ; Q =Kalddgi quartzites.

Deccan trap; M

laterite ; D=

Fig. 20,— Section across the Gauli plateau, after Foote, L

GEOLOGY OF INDIA—LATERITE. (Chap. XV.

evidence in any part of India to render it probable that
the whole of the great trap plateau has been beneath
the sea in tertiary times. It is inconceivable that
fluviatile deposits should be so enormously extended,
yet so thin. One objection, which at the first glance,
appears important, is apparent rather thaa real. I[t
is that a sedimentary deposit could not be formed
on the highest portions of the country, because there
could be no higher Iand in the neighbourhood from
which the sediment might be derived, whilst the
singularly small amount of disturbance which the
Deccan rocks have undergone renders it improbable
that any great relative change of elevation has taken
place. But it must be remembered that laterite is
a rock which resists atmospheric action far more than
most forms of doleritic trap, as is shown by the
manner in which hard unaltered caps of laterite
rest upon softened and decomposing basaltic rocks.
Consequently those portions of the plateau which
were originally highest may, if not capped by
laterite, have disintegrated more rapidly than those
protected by the lateritic formation, until the latter
remained, forming the highest ridges, long after
the unprotected portions had been swept away.

The evidence afforded by the laterite outliers in
Bundelkhand? is distinctly opposed to the theory of
alteration zz sztu. The whole group, laterite above
underlaid by ferruginous clay, frequently containing
sand and pebbles, is found indifferently capping the
trap and Vindhyan sandstones. Now, whatever may
be the case with dolerite and gneissic rocks, no
conceivable process of alteration could convert a
purely quartzose rock, containing a mere trace of
iron, like the Vindhyan sandstone, into an argillaceous
one with 20 per cent. of iron entering into its com-
position, and the circumstance that the lower portion
of the lateritic group is clearly detrital, proves that the
laterite is not an altered outlier of the Deccan trap.

It appears almost impossible to separate the Bundelkhand laterite from
the high level laterite of the Deccan. Lithologically and stratigraphically

' Memoirs, 1, 79—86, (1860),
Chap. XV.] ORIGIN OF THE LATERITE IN MALABAR, 383

the two rocks are identical. There can be no reasonable doubt that the
trap once occupied the surface of the ground now cut out inte valleys by
the feeders and main streams of the Son, Narbadd, and Mahdnadf, and
that Bundelkhand and Malwa were united with Mandl4 and Sargdja into
one plateau of horizontal trap rocks. If this be conceded, and it
appears impossible to doubt it, the caps of laterite near Sdgar (Saugor),
-occupying precisely the same relative position as those at Amarkantak
and the Mdin Pat, may fairly be considered part of the same bed, and the
Rewa outliers, which are probably either beyond the original range of
the trap, or else on ground which was above the general trap level, must
be referred to the same origin. Now the Amarkantak and Sargija
laterites are not merely similar in every respect to the other Deccan high
level outcrops of the rock, but they appear to be connected, by a series of
small caps at intervals, with the typical formation of the southern Mardth4
country. There may be a break in the chain, as the distances are too great
for any safe conclusions to be formed, and all that can be done is to indi-
cate the probabilities,but it appears to be a fair inference that, if the
Bundelkhand laterite is of detrital origin, the rock of Amarkantak and
the Deccan is the same.

The laterite of the Rajmahdl hills is separated by so great a break
from that of Sargija, and the Rdjmahdl traps are in all probability so much
older than those of the Deccan, that it is impossible to say whether the
R4jmahél laterite is of the same age as that of central and western India.
Lithologically it is identical, and like the Deccan laterite it occurs, in
part at least, at a considerable elevation, whilst its sedimentary origin has
already been mentioned.

On the west coast of the Peninsula we havea careful study of the low-
level laterite of south Malabar by Mr. P. Lake! He describes two distinct
varieties, the first, distinguished as vesicular laterite, is characterised by nu~
merous vermicular branching and anastomising tubes which, in the portion
of the rock not exposed to the air, are filled with a white or yellow clay,
containing a much smaller proportion of iron and a larger percentage of
potash than the walls of the tubes. Away from the surface, as the rock
becomes less affected by the weather, the distinction between the tube
walls and their contents gradually disappears, till what was laterite above
passes into clay below. The laterite is thus seen to be the result of a
sort of superficial concretionary action, the iron tending to segregate in
the form of tubes from which the clayey non-ferruginous parts are washed
out, leaving the resulting laterite with a higher proportion of iron than the
clay from which it was formed,

! Memoirs, XXIV, 217, (1890).
384 GEOLOGY OF INDIA —LATERITE. (Chap. XV,

The other variety of laterite, distinguished as pellety, is derived from
the vesicular laterite. After the tube contents have been washed out the
tube walls slowly break up into little subangular pieces, which are washed
away and deposited at a lower level, where they are recemented into a rock,
usually much more solid than the first variety. It consists of small irre-
gular ferruginous pellets cemented by a similar material, is devoid: of
vermicular tubes, and has a higher percentage of iron than the unexposed
parts of the vesicular variety, the percentage being probably much the
same as in the tube walls.

The origin of the pellety form is sufficiently obvious: it is a detrital
rock fo-med of the debris of the vesicular form, which is regarded by Mr.
Lake as a product of decomposition ¢# situ of the gneiss. The gneiss of
Malabar, consisting principally of quartz, felspar, hornblende or mica, and
garnets, weathers readily into a ferruginous clay. If exposed to the
weather, the iron segregates and hardens the clay, where it collects, while
the rest, deprived of ferruginous cement, is loose and easily washed away.
As this process goes on the gneiss becomes covered with a cap of laterite,
gradually increasing in thickness till it acts as a protection to the under-
dying gneiss from further decomposition.

There can be no doubt that the vesicular laterite is, in part at least,
formed by a laterisation of the decomposed gneiss, for the lamination of
the gneiss can often be traced into the laterite, and the more quartzose
beds stand up as ribs several feet into the laterite which has replaced
the more felspathic portions on either side.

One more hypothesis of the origin of laterite requires notice. Mr. F.R.
Mallet, in noticing the resemblance of certain ferruginous beds with under-
lying bole, interstratified with the volcanic rocks of Ulster,! to the laterite
and lithomarge of India, suggests that the laterite is of lacustrine origin,
During the decay of vegetable matter in the presence of the higher oxides
of iron oxygen is absorbed, red ucing them to protoxide, which unites with
carbonic acid, a product of the decomposition of vegetation, to form ferrous
carbonate, soluble in water containing an excess of carbonic acid. When
the water carrying this ferrous carbonate in solution is exposed to the
air in streams or lakes it absorbs oxygen, the ferrous carbonate is decom-
posed with the escape of carbonic acid and a re-oxidation of the ferrous
oxide into insoluble ferric oxide, which is precipitated wherever the
water comes to rest in a lake or marsh. :

On this hypothesis the high level laterite would have been formed in
the shallow depressions left between the lava flows and ash heaps of the
surface of the land at the close of the Deccan trap period, while the low

1 Records, XIV, 139, (1881).
Chap. XV.] SUMMARY OF THEORIES OF ORIGIN. 385

level laterite, excluding the ferruginous gravels, would have been formed
in the depressions on the plain of marine denudation whose surface it caps.
As already explained, the resistance which laterite offers to denudation
would be sufficient to account for its being now found capping the hills,
and if it were originally formed in the depressions of the surface, for the
complete reversal of contour is an indication of the time that has elapsed.

From what has gone before it will be seen that the subject of the origin
of laterite is still wrapt in obscurity. None of the various hypotheses that
have been propounded is completely satisfactory, nor is it possible to come
to any final conclusion till an agreement is come to as to the meaning of
the word laterite. ‘It must be used either as purely a lithological, or as
a chronological, term, not indifferently as either, but as the word was
originally intended to describe a peculiar variety of rock, irrespective of
its age or origin, for which too a fresh name would have to be adopted if
the familiar one were abandoned, it is the lithological sense of the word
which it is most desirable to retain.

Using the term as defined by its proposer for a vesicular, highly ferru-
ginous, clay, soft in the mass but readily hardening on exposure to the
weather, it seems that there is nothing essentially volcanic in the rock.
The high level laterites are doubtless derived, directly or indirectly, from
the debris of volcanic rocks, but it is to the gneiss and granite that we must
look for the ultimate source of the laterite of the south Konkan, Travancore,
and Ceylon.

According to some geologists this laterite is in reality asoil and formed
by the direct decomposition 7” situ of the underlying rock, but) some doubt
attaches to the observations on syhich this conclusion is based, as the ap-
parent transition may be due to an infiltration of iron from the overlying
laterite and a conversion of the decomposed portion of the underlying rock,
where it is suitably argillaceous, into laterite. It is certain that, though
laerite appears to a certain extent to be still in course of formation, the
bulk of the reek is of ancient date, for it is now found on the summits of
steeply scarped trap hills, or a deeply eroded plain of marine denudation,
the exgeptions being for the most part those cases where it is palpably or
prebably of detrital origin, and derived from pre-existing laterite.

There are two difficulties in the way of accepting this explanation
without modification. The first is the large proportion of iron present in
laterite, a proportion which is only exceptionally found in any volcanic
rock, and much exceeds that present in any of the gneissose rocks, apart
from certain highly ferruginous bands. Such local and exceptional accu-
mulations of iron are not sufficient to account for the amount present in the
laterite caps far removed from them, nor can the concentration of iron
386 GEOLOGY OF INDIA—LATERITE. (Chap. XV.

caused by the washing away of the less ferruginous and consequently less
coherent portions, account for its presence, any more than for the argil-
laceous nature of the rock, where it rests on sandstones almost devoid of
any trace of clay or iron. The supposition that laterite is derived from the
decomposition of a specially ferriferous eruption, marking the close of the
Deccan trap period, is insufficient to account for the facts even within the
Deccan trap area, and quite fails when applied fo the remote exposures,
far beyond the limits of the area within which the influence of these erup-
tions could have been felt. ; 7

One objection which might be urged to the hypothesis of the origin of
laterite by the decomposition 7” sétu of volcanic rock, that it should in that
case be found interbedded with the trap, may be dismissed. However
laterite originated, time was certainly a factor in its “formation, and the
occurrence of beds of bole suggests that the same causes which subse-
quently led to its formation, were at work during the Deccan trap period,
but that the rapid succession of lava flows did not leave them time to pro-
duce the full effect which resulted when the eruptions had ceased.

The second difficulty is the great thickness of some of the patches of
high level laterite, which seems to preclude their having been formed as a
-soil, by the segregation of the oxide of iron and removal of the less
coherent non-ferruginous portions, and the more probable explanation is
that they were formed in lakes or marshy hollows by the deposition of oxide
of iron from the stagnant waters, mixed more or less with fine grained ash
and decomposed volcanic debris. The laterite of the lowlands may have
been similarly formed in marshy hollows, left on the surface of a plain of
marine denudation after its elevation above sea level. It is described
as a thin, fairly uniform layer covering the undulations of this plain, but in
Mr. Lake’s description of south Malabar there are said to be gneiss hills,
or islands, rising from this plain, on which no laterite is found, and if the
laterite were a direct product of decomposition of gneiss, it is difficult to
see why it should be restricted to the lowland near the coast, which was
evidently once covered by the sea.

As to ‘the conditions necessary for the formation of laterite little can be
said. Those countries where it has been supposed to be still in process of
formation are characterised for the most part by a warm, moist, climate, and
an abundant vegetation. .But there is one characteristic of all the laterite
regions that appears to be important ; the laterite is without exception only
found on level or gently undulating surfaces, if we ignore the irregulari-
ties produced by subsequent denudation. It is found on the terrace border-
ing the sea coast, and on the plateaux capping the hills further inland, but
whenever a rock which could pass for laterite is found on the intermediate
slopes, it is clearly of derivative origin. The rounded surfaces of the
gneiss hills of Ceylon and Southern India are often covered to a great
Chap, XV, ] GEOLOGICAL AGE. 387

depth with a more or less ferruginous subsoil, which never passes into
laterite, except in such localities as the summit plateau of the Shevaroy hills
-or the plain, now intersected with valleys cf denudation, which borders the
sea coast. It is also said to be found in the bottoms of the valleys of south
Malabar, but wherever laterite is found on the sloping ground, it is clearly
derived from the disintegration of some bed at a higher level,

The geological age of the high level laterite must, of course, remain
undetermined, until the mode of formation has been more definitely ascer-
tained. If the rock be merely the result of surface alteration, it may
be of any date subsequent to the termination of the volcanic out-
bursts. Indeed, it must still be in process of formation, as has been
justly pointed out by several observers. But, as its occurrence in the
form of a few isolated caps shows that it was once a much more extensive
formation, it must have existed before the denudation of the area
had much advanced, and must, therefore, have been formed, in part at
least, soon after the termination of the volcanic eruptions. The great
similarity between the high level laterite and the beds of the same rock
interstratified with the nummulitic limestones and gravels of Gujardt and
Cutch, tends to suggest the possibility that the two are contemporaneous,
and also that they may have been produced in the same manner, with
this important distinction, however, that the Gujarat beds are marine,
whilst there does not appear to be any evidence in favour of supposing
that the highlands of the Deccan were submerged during any portion of
’ the tertiary period. Had they been submerged, the amount of denudation
which the traps must have undergone would, in all probability, have
caused the high level laterite to be more distinctly unconformable, At
the same time, it is far from clear that the laterite is truly conformable to
the highest trap flows. It has been hitherto assumed, rather than proved,
that all the beds of laterite, at lower elevations than the summit bed, are
of later age. The occurrence of laterite at various elevations presents
no difficulty on the theory of the laterite being an altered form of
the traps, but if this rock be of any definite date, it is clear that ex-
tensive denudation must have reduced the level of such hills as Mé-~
therdn, the uppermost beds of which are at least 2,000 or 3,000 feet
velow the highest volcanic flows, before the laterite was deposited.
Nevertheless, the laterite of Matherdn, although apparently non-detrital,
may be a secondary product. This question of the conformity of the high-
level laterite to the highest traps requires, in fact, further investigation.

Whether the true laterite of the low grounds near the sea is to be re-
garded as newer than that of the high level plateaux depends on the
hypothesis of origin adopted. If they are both products of decomposition

2c
388 GEOLOGY OF INDIA—S.ATERITE. CChap. XV,

in situ of the underlying rock, they may both be of the same age, but if,
as seems probable, the formation of the laterite was anterior tothe excava-
tion of the valleys which now limit its extent, then the low level laterite is
shown, by the less degree of denudation that has taken place since its for-
mation, to be newer than the high level. On any hypothesis, except the
impossible one of direct volcanic origin, there is no reason why the pro-
duction of laterite should be restricted to any particular geological age,
and Mr. Lake’s observation of the apparent passage of gneiss into laterite
in the bottoms of some of the valleys of scuth Malabar,!' would bring the
date of origin of some part of the laterite down to a, geologically, very
recent period.

‘The foregoing remarks regarding the origin of laterite refer only to
what may be called the non-detrital form, in the sense that it is not pal-
pably formed of the debris of pre-existing rocks of a similar nature ; but the
bulk of what has generally been described as low level laterite, principally
along the east coast, is evidently of clastic origin, and is often merely a sub-
recent gravel with a ferruginous matrix. Although, as has heen explained,
it seems advisable to use the term laterite in a purely lithological sense
and not to apply it to such rocks, even though there is every degree of
transition between the two, it is impossible to ignore these so called late-
rites, seeing that the term has been so generally used in the past.

Like the true laterite of south Malabar, they are found resting on
what appears to be a plain of marine denudation, and the enquiry naturally
arises as to whether the low level laterite isa marine formation. A priori
it would appear improbable that a marine formation should be deposited
during the process by which a plain of marine denudation is elevated
above the sea. On the other hand, the frequent occurrence of pebbles,
often of large size, in the laterites of the east coast appears due to the
action of the waves, especially where, as around the isolated hills in Orissa,
which may originally have been islands, a mass of well rounded shingle,
in every way resembling a beach, is found cemented together by laterite.
‘The absence of marine fossils may be due to their having been obliterated
by. the forces which produced the peculiar concretionary structure of the
rock.

There are, however, two circumstances which appear to militate
strongly against considering the laterite a marine formation. One of these
is its remarkable thinness, which, so far as is known, rarely exceeds 20
feet along the east coast, and the other is the very frequent occurrence,
in the Madras country, of paleolithic implements imbedded in the rock.
Some of these might have been dropped into the sea from canoes, but

" Memoirs, XXIV, 226, (18go).
Chap. XV. ] AGE OF LATERITE OF MADRAS, 389

it is incredible that the men who used the stones should have lost them
in the sea in such numbers as would account for their present abundance.

On the whole, it appears most probable that the low level laterite is a
subaerial deposit, due, however, in many cases, to the rearrangement of
marine gravels and sands by rain and streams. All rain and stream
action would tend to carry away the lighter sand and clay, and to leave
behind the heavy iron sand, and to this may be due the concentration of
the ferruginous element.

The presence of palzolithic human implements in the Madras laterite
proves that the rock is of post-tertiary origin. The implements! found
are chiefly of quartzite, and have evidently been fashioned from peb-
bles, derived originally from the rocks of the Cuddapah system.

Despite the geologically recent origin of the low level laterite, the con-
siderable amount of denudation which it has undergone shows that it is,
in part at least, a formation of ancient date, counting by years, It has
already been mentioned that, on the west coast the plateau near the sea
has been cut through by streams to a great depth, and the underlying trap
exposed, and that farther inland, at a higher level, only a few caps of the
low level laterite remain. On the eastern coast, which, owing to the large
amount of deposits brought down by rivers, is protected from the action of
the sea, the laterite has undergone less denudation, in consequence of its
being frequently covered by later alluvial deposits, but still it has been
removed by atmospheric action over large areas away from the coast. It is
probable that the land rose very slowly from the sea, the laterite forming
on the raised slope pavz passu with the elevation, and that, consequently,
the farther inland the rock the older its date, and the longer the period
during which it has undergone denudation from atmospheric agencies.
But the deep ravines cut by the streams close to the western coast, near
Ratnagiri, mark the lapse of a considerable period of time since the low
level laterite was first consolidated, and a curious piece of evidence of the
same kind has been recorded by Mr. Foote? in the neighbourhood of
Madras.

Between two villages called Amerumbode and Maderapaucum, east of
Sattavedu, and about 30 miles north-west by north of Madras, are some
stone circles, made of blocks of the laterite, in which palzolithic imple-
ments are found in abundance in the immediate neighbourhood. The
stone circles are known in the country as Karambar rings, and precisely
similar rings of stone are found in many parts of India, associated with
various other rude stone buildings such, as kistvaens and cromiechs.
That these stone circles are of much later date than the palzolithic

1R,B. Foote, Mad. Four. Lit. Sct., Oct. 1866,]} Soc., XXIV, 484, (1868).
grd series, Pt. 2, p.1; also Quart, four. Geol. | * Memoirs, X, 47, (1873).
2C2
390 GEOLOGY OF INDIA—LATERITE. [ Chap. XV.

implements is evident, first because in the particular case near Madras the
circles are construct ed of rock in which the implements are imbedded, and
secondly, because iron implements, which mark a far more advanced stage
of human progress, have been repeatedly found within the enclosures,
Nevertheless, the stone circles themselves must be the work of a very
ancient period, for all record of their construction, or even of the people
who built them, has passed away.
CHAPTER XVI.

PLEISTOCENE AND RECENT DEPOSITS,

(Exclusive of the Indo-Gangetic alluvium.)

PenINsuLAR AREA—Extent, and distinction from tertiary beds—Various forms of posttertiary
deposits—Cuddalore sandstone—Warkalli beds—Cave deposits—Older alluvium of the
Narbad4, Tdpti, Godavari and Kistna— Newer alluvium of the east coast—Smooth water
anchorages and recent deposits of the west coast—Lake deposits—Reguz, or cotton soil—
Blonw sand —ExrTra-PENINSULAR AREA—Hills west of the Indus—North-West Punjab—
the Himdiayas —Eastern frontier—Alluvium of the Irawadi.

The posttertiary (postpliocene, pleistocene or quaternary) and recent
formations of India occupy an immense area, They form the wide plains of
the Indus, Ganges, and Brahmaputra, and cover large tracts of country
south of the Gangetic and east of the Indus plain. No older formation is
exposed throughout the greater portion.of the belt of alluvial lowland
fringing the eastern coast of the Peninsula, and subrecent accumulations
occupy a large area in Gujarat and in some other districts near the western
coast. Large deposits in the valleys of the peninsular rivers and upon the
fertile plains of the interior are also of recent or subrecent origin, The
most important and extensive of these forms the great Indo-Gangetic plain,
and, as the extent and variety of the recent and subrecent deposits render
it impossible to treat of them all in a single chapter, its description, with
all the important and interesting questions it raises, will be deferred for
the present.

It is very difficult to draw a clear and distinct line between tertiary
and posttertiary formations in India. The limit of the two in Europe
coincides with the glacial epoch, but as no physical trace of this cold
period has been detected in peninsular India, the distinction between the
pliocene tertiary formations and the postpliocene beds is there less easily
defined, Practically, no difficulty has hitherto arisen, because the
tertiary beds of the Peninsula are comparatively unimportant, and those
which occur belong to the older or middle tertiaries, and not to the
newer beds, so that there is a marked break -between the tertiary and
posttertiary deposits; but in the extra-peninsular area, where the upper-
most tertiary deposits are largely developed, it is often extremely
difficult to say where the line should be drawn,
392 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [ Chap, XVI.

In dealing with the recent and subrecent deposits of India it is impos-
sible to observe a strictly chronological order, and it is necessary to classify
them more or less according to their nature. In the case of those in
peninsular India the following classification will be followed, the oder
being roughly, though not strictly, that of their date of origii. :—

1. Subrecent or doubtfully tertiary deposits of the coastal region.
2. Cave deposits.

3. Older alluvial deposits.

4. Newer alluvial deposits of the river valleys and deltas.

5. Raised littoral accumulations of sand, shells, etc.

6. Soils.

7. Blown sand.

Along the eastern coast cf the Peninsula, from the neighbourhood of
Raj4mahendri to the Tinevelli district, a peculiar formation, consisting
chiefly of sandstones and grits, is found underlying the laterite and asso-
ciated gravels which ferm a low slope on the edge of the east coast
alluvium. This sandstone formation has received several local names, but
is now generally known as the Cuddalore sandstones,’ from being well
developed in the neighbourhood of the civil station of Cuddalore, about
100 miles south of Madras.

The greater portion of the Cuddalore group, throughout the area in
which it is found, consists of gritty and sandy beds, sometimes highly
ferruginous, and coloured of various tints of yellow, brown, red, and
purple, sometimes white or pale coloured, and not infrequently mottled-
In some cases the rock is argillaceous, and occasionally thin bands of
clays or shales are interstratified, The beds are soft, loose textured and, as
a rule, ill consolidated, being rarely safficiently compact to be used as
building stone. Bands of conglomerate have been found.

As already stated, these beds have been traced throughout a large
portion of the east coast. Their most northerly extension known is
between Vizagapatam and Réjdmahendri. The coast nerth of Vizagapa-
tam, as far as the Chilk4 lake, has not been examined geologically, and
throughout Orissa no outcrops of the Cuddalore beds have been detected,
but there is a possibility that they may be represented by some clays and
sandy beds associated with the laterite of Midnapur.? There is rather
more probability that certain sandstones, grits and clays, which occur east
of Raniganj, and extend northwards as far as Surf in Birbhtim, belong to
the same group as the Cuddalore sandstones of Madras.

From the neighbourhood of Rajd4mahendri the Cuddalcre beds have
been mapped at intervals along the coast for fully 600 miles to the

1 For further information see more particu- | Foote, Memoirs, X, 59, (1873) and XX, 35,
larly, H. F. Blanford, Memoirs, 1V, 165, (1863); | (1883).
King and Foote, Memoirs, 1V, 256, (1864); 2 Memvirs, |, 268, (1859).
Chap. XVII CUDDALORE SANDSTONES, 393

vv

southward. They usually form a low slope, dipping at a very slight angle to
the eastward, in the direction of the sea, and are, as a rule, much covered
and concealed by the deposits associated with the low level laterite of the
east coast. To the westward they rest indifferently, but always unconform-
ably, upon rocks of various ages,—-metamorphic, jurassic’ or cretaceous,—
and they often terminate in this direction in a low scarp. Tothe eastward
they disappear in places, with their capping of laterite, beneath the alluvium
of the coast, but they quite as often, especially to the southward, terminate
in a small cliff. Their outcrop is repeatedly interrupted by the broad alluvial
valleys of rivers, and in some places, as for nearly 100 miles south of
Madras, they appear to be wanting altogether, whilst in other parts of the
country they form a broad tract, usually sandy and infertile, raised above
the general level, occasionally no less than 25 miles wide from east.to west,
as near Cuddalore, but generally much less.

From the paucity of sections and the extent to which the Cuddalore
sandstones are concealed by laterite and sandy soil, their absolute thick-
ness can nowhere be estimated with accuracy. ‘The scarp in which they
terminate to the westward is sometimes as much as roo feet high and
they must. be somewhat thicker than this, but it is doubtful if they attain
any considerable thickness. ‘They are perfectly undisturbed, and have all
the appearance of being a comparatively late formation,

The only fossiis found in the Cuddalore beds consist of exogenous sili-
cified fossil wood, some of which is coniferous and has been described
under the name cf Peuce schmidiana,' The genus Pexce is not acknow.
ledged by all palzobotanists, and it appears too ill defined to justily any
conclusions as to the age of the rocks being founded upon its occurrence.

This silicified wood is especially abundant at Tiruvakarai(Trivicary’, about
fourteen miles west north-west of Pondicherri, whence the name of Trivicary
grits has been applied by some writers to the local development of the Cudda-
lore sandstones. The trunks of trees occurring at this place are of large
size, one having been found as much as roo feet in length, while stems 15
to 20 feet long and 5 to 6 feet in girth are not uncommon, ‘They occur
prostrate, imbedded in ferruginous grit.

‘The age and mode of origin of the Cuddalore sandstones are obscure,
as but little importance can be attached to the identification of one generic
form of coniferous wood, They are quite unconformable to the cretaceous
beds, which they overlap in a most irregular manner, as near Pondicherri,
where they rest on beds of the Ariyalir group, forming the plateau
near the town, known as the Red hills; six miles further westward, and
west of the belt of cretaceous rocks, they are seen resting on Utatir beds
near Tiruvakarai, whilst a few miles further west they completely overlap

+ Schmid u. Schleiden; Ueber die Natur der Kieselho'zer, Jena, 1855, pp. 4, 36.
394 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [ Chap. XVL

the cretaceous beds and rest on gneiss. Fragments derived from
the cretaceous beds and containing cretaceous fossils have been found
near Tanjore. Near RAjdimahendri the Cuddalore sandstones overlie the
Deccan trap, the jurassic rocks and the gneiss. It is safe, therefore, to
conclude that the sandstones cannot be older than upper tertiary. And
the discovery by Mr. Foote, in the Tinnevelli district, of subrecent marine
beds, containing only living species of mollusca, associated with grits which
he believed to represent the Cuddalore beds,! appears to show that an.
even later date should be ascribed to them.

The origin of these rocks is as obscure as their date. Occurring as
they do, parallel with the coast, it is natural to suppose that they are of
marine origin and have been formed near the shore, when the genera]
contour of the coast was the same as it now is, though the level of the
land was somewhat lower. But the complete absence, so far as is known,
of all marine remains is not easy to explain. Coarse sandstones and grits
are usually unfossiliferous, but in beds which have undergone so little
change some casts of shells, at least, would probably be found in the more
argillaceous strata, if they were of marine origin. At the same time it is
difficult to suppose that the western coast of the Bay of Bengal can
have formed part of a river valley in tertiary times, and it is equally im-
probable that stratified grits, sandstones and conglomerates, like those of
the Cuddalore beds, can be a form of subaerial wash.

On the west coast of the Peninsula a series of clays and sandy clays with
lignite beds near their base, known as the Warkalli beds, are found for
about twenty miles along the Travancore coast, from about three miles north
of Quilon to the same distance south of Warkalli, and have been supposed
to represent the Cuddalore sandstones of the east coast. Some doubt may
be expressed as to the correctness of this correlation, for the Warkallj
beds are said to attaina thickness of 200 feet, double the greatest recorded
thickness of their supposed representatives on the east coast, and their
upper surface is said to be formed by the same plain of marine denu-
dation which cuts the gneiss further inland,’ pointing to an older date than
that of the Cuddalore beds. The Warkalli beds are said to be underlaid
by the limestones containing eocene fossils, which have been referred to ina
previous chapter, but though there is no reason to doubt the presence of
these eocene limestones, they have not been examined zz situ by a
competent geologist, and we have no information at present as to whether
or not they are conformable to the overlying beds. If so, it would point
to a greater antiquity for the Warkalli beds than there seems to be any
good ground for ascribing to the Cuddalore sandstones.

' Memoirs, XX, 41, (1883). | 2 W. King, Records, XV, g2, (1882).
Chap. XVI] CAVE DEPOSITS. 305

On the coast of Kathidwdr a subrecent marine limestone, largely used
as a building store in Bombay, is found. It is commonly known as Porebandar
stone from the name of the port whence it is shipped or, using the name
proposed by Dr. Carter, miliolite. The typical miliolite is a finely oolitic
freestone, largely composed of foraminifera, which form the nuclei of the
oolitic grains, but near the sea coast the limestone is not infrequently
mixed with a large proportion of sand. In the eastern part of KAthidwar
these beds are only found near the coast ; further west, however, they form
the whole surface mapped as recent, and extend on tothe tertiary rocks and
the Deccan trap. They attain a maximum thickness of about roo feet,
are extensively false-bedded in thin layers, and, though clearly a marine
deposit land shells, which were doubtless washed down by small streams
or floods, have been found in some of the more impure beds."

There is only one locality in the Indian Peninsula where mammali-
ferous cave deposits have been detected. This is at a place called Billa
Surgam, a few miles north of Banaganpalli in the Karnil district. The
caves are in the limestone belonging to the Karnil series, and situated
at a higher level than the beds of the present drainage, their floor is in-
crusted with stalagmite, beneath which red marl, full of bones of animals,
large and small, is found. These caves were first discovered by Captain
Newbold, whose collections were, however, never described, nor can the
specimens now be found. More recently they have been explored by
the Geological Survey, and the collection of bones, some of which exhibit
traces of having been shaped by man, described by Mr. Lydekker. The
fauna, besides many living forms, contains five species, Viverra karnu-
liensts, Hystrix crassidens, Atherura karnuliensis, Rhinoceros karnu-
liensis, and Sus karnuliensts, which are extinct, though closely allied to
living forins. But the most interesting feature is the occurrence of four
types identical with, or closely allied to, living African forms; these
are Cynocephalus, sp. Eguus asinus, Hyena crocuta, and Manis cl.
gigantea.” ;

The older alluvial deposits are well represented in the valleys of the
Peninsula. Leaving the deposits of the Indo-Gangetic plain for separate
description, the various older alluvial deposits of the peninsular rivers

1 Your, Bo. Br. Roy. As. Soe. V. 513, (1857); | Four. As. Soc. Beng., XIII, 610, (1844); R. B.
Geological papers on Western India, Bom-| Foote, Records, XVII, 27, 200, (1884); XVIII,
bay, 1857, p. 750; Memoivs, XXI, 126,| 227, (1885); R. Lydekker, Records, XIX, 120,
(1884). (1886); Pal, Indica, series x, IV, pt. 2

3 For further details, see T. J. Newbold: | (1886).

’
396 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT (Chap. XVI

deserve notice, both on account of the area they occupy, and of the organic
remains they have yielded. ‘The tivers of the Peninsula may be divided
into two main groups; the first comprises the Narbadé and Tapti, which
flow westward and drain the central portion of the Peninsula; the second
includes the Mahdnadi, Godavari, Kistna, Penner, Cauvery, and several
minor streams which flow eastwards into the Bay of Bengal.

In the first named of the two groups one striking peculiarity is
noticeable. Extensive alluvial plains, composed of clays and gravels,
exist in the valleys of the Narbad4 and Tdpti. In the Narbadd vallev
the principal plain extends from a little east of Jabalpur to Hardd, a dis-
tance of more than 200 miles, and varies in breadth from 12 miles to
35. There is a smaller plain further down the river, extending for about
80 miles from Barwai to the Harin Pal south of Bagh, but it is compara-
tively ill marked, the alluvial deposits are, so far as is known, much less
deep, and no mammalian remains have been found. In the T4pti valley
there is a large plain in Khdndesh, extending east and west for
about 150 miles, and terminating to the eastward close to Burhdnpur.
This plain lies chiefly to the north of the river, and is probably in places
as much as 30 miles wide, but its limits have not been accurately deter-
mined. It appears to be connected by a narrow alluvial belt to the south-
east with the plain of the Purna,a tributary of the T4pti, draining a great
portion of Berar. The Purna plain is at a higher level than Khdndesh
and is about 100 miles long, and in places 40 miles broad, its eastern ex-
tremity being near Amrdoti, so that the whole length of the combined
Tdpti and Purna plains is about 240 miles. The Tdpti and both the
Narbadd plains are closed on the west by rocky and hilly country, through
which the river has cut a channel with a rapid descent, and in the case of
the Narbada4, as will be explained presently, it is ascertained that the allu-
vial deposits of the upper basin extend to a considerable depth beneath
the level of the river bed at the point of exit, so that the plain lies in a
great rock basin,

In the valleys of the eastward flowing rivers, such as the Godavari,
Kistna and Cauvery, there are no such broad and well defined alluvial
plains as in the drainage areas of the T4pti and Narbadd, ‘There are
numerous extensive alluvial flats in many places, but they are far inferior
in extent tothe Narbad4 and T4pti plains, and they appear to be chiefly
due to the river having worn a broad valley through soft, or easily disin-
tegrated rocks, This is especially the case on the Godavari and its tribu-
taries, the alluvial portions of the river valley being in the Gondwana
rocks, or else in the Deccan traps, whilst the river traverses rocky gorges,

1 This is not quite certain however, the | but thz fall from one to the other cannot
ground not having been properly surveyed, | be much more than 100 feet, to judge by
There is a considerable amount of rock ex- | the railway levels,
posed im the rivers between the two plains,

#
Chap. XVI.] ALLUVIUM OF THE NARBADA, 397

through the metamorphic rocks forming the various barriers, at the
places where the valley leaves the softer formations. On the Narbadd and
TApti it is otherwise, the rocks underlying the alluvial areas, so far as
is known, are of the same kind as those cut through by the rivers at
their exit from the plains. It is not improbable that the formation of
these well defined plains in the Narbadd and ‘ldpti valleys, and the ab-
sence of similar flats on the Godavari and Kistna, may be due to the rise
of the Indian Peninsula in posttertiary times having been, as already
suggested, greater or more rapid to the westward than to the eastward.

Partly in consequence of mammalian bones having been discovered in
considerable quantities, and partly because the geology of the neighbour-
ing country is of so much interest and variety as to have attracted the
notice of many geologists, the alluvial deposits of the Narbadé4 valley have
received far more attention than similar formations on the banks of the
other Indian rivers.' The great plain already mentioned as extending
from Jabalpur to Hardé is chiefly composed of a stiff, reddish, brownish
or yellowish clay, with numerous bands of sand and gravel intercalated.
Kankar abounds throughout the deposit, and pisolitic iron granules are
of frequent occurrence in the argillaceous beds. Occasionally pebbles and
sand are found cemented together by carbonate of lime, so as to forma
hard compact conglomerate. This rock is especially developed at the base
of the alluvial deposits, and is often found forming a coating to the under-
lying rock, not only in the Narbadd but in many other river valleys,
The clay is frequently quite devoid of stratification, but it appears never
to attain any great thickness without sandy layers intervening. The river,
in many places, cuts through the clays, sands, and gravels to the under-
lying rock, usually belonging to the transition series, and the section of
old alluvial deposits on the banks of the stream never greatly exceeds 100
feet in depth, this being about the usual difference in elevation between
the bed of the Narbadd and the géneral surface of the alluvial plain in the
neighbourhood of the river. But in a boring which was made at Stkakheri,
north of Mohpdni and south of the Gddawdrd station on the Great Indian
Peninsula Railway, a depth of 491 feet was attained, without the base of
the alluvial deposits being reached; another bore-hole was made through
alluvial beds close to Gidawé4rd station to a depth of 251 feet. Through-
out the thickness of nearly 500 feet, no change of importance was
detected in the alluvial formations. By far the greater portion of the
beds traversed consisted of clay with calcareous and ferruginous grains,
sand and pebbles being found occasionally throughout. The bottom

. For description of the Narbadé alluvial | VI, 227, (1869); Records, VI, 49, (1873);
deposits see Memoirs, Il, 279, (1860); | VIII, 66, (1875).
398 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. C Ohap. XVL

of the bore-hole was in lateritic gravel, and it is possible that rock was not
far distant.

The evidence thus obtained of the depth to which the alluvial deposits
of the Narbada valley extend proves that they fill a rock basin, for the bed
of the Narbada river, at the point where it leaves the alluvial plain near
Handid and commences to run through the rocky channel which extends to
Bdrwai, is not more than 200 feet below the level of the surface at Gdéda-
ward and Sdkakheri, and the valley is surrounded by higher rocky ground
in every other direction. A slight prolongation of the alluvial basin to the
south-west in the direction of Hard4, the prevalence of alluvium in parts
of Nimdr, and the circumstance that there is a great break by which the
railway traverses the Sdtpura range, immediately east of Asirgarh, may
indicate that the upper Narbada formerly joined the Tapti in Khdndesh,!
and that the lower part of the valley of the former river, as it now exists,
is due to changes of level in the later posttertiary period.

The surface of the Narbaddé alluvium is undulating, and evidently
denuded by the action of rain and streams. There is a slight slope of the
surface to the westward throughout the plain, the elevation of the railway
station at Hardd, at the western extremity of the alluvial tract, being 947
feet above the sea, whilst Sohdgpur station is 1,103 feet, Narsinghpur
1,185, and Jabalpur, which is, however, on rock a little above the
plain 1,351. The fall of the surface in 200 miles is probably about 3:0
feet.

Different views have been put forward as to the marine, lacustrine,
or fluviatile origin of the Narbadé alluvial deposits, but, before considering
these, it will be well to give a list of the organic remains hitherto identified.
They consist of bones and shells, and the following species have been
determined :—

VERTEBRATA, ?

MamMaLia— MamMatia,—contd.
Rhinoceros unicornis.

Orsus namadicus.
Elephas namadicus.

Bubalus paleindicus.

Leptobos fraseri. t » tnsignis.

Bos namadicus. t » ganesa.

Cervus (? duvaucelli). REpTiLia—

Sus, sp. a :
ae Daye Pangshura flaviventris.

Hippopotamus pal@indicus, | Batagur, ct. dhongoka.

Trionyx, cf. gangetius.

o namadicus.
| Crocodilus, sp.

t Equus namadicus.

1 The greatest elevation on the G. I. P. | Narbadd.
Railway between the Narbad4 and Tépti ® Lydekker, Pal, Indica, series x, Ill,
valley, is 1,245 feet above the sea, or only 300 | (1884-86). The species marked with a dagger
feet above Hard in the alluvial plain of the | are found also in the Siw4liks.
Chap. XVI. ] NARBADA FOSSILS. 399

MOLLUSCA.!
GasTEROPODA== Planorbis exustus.
Melania tuberculata. * » comp essus?
Paludina bengalensts. LaMELLIBRANCHIATA—
33 dissimalis. Unio corrugatus? var.
* Bythinta cerameopoma., n» ‘indicus.
* ” pulchella. » sp. near U. shurtlefianus.
Bulimus insularis. » marginalis.
* Lymnea acuminata. Corbicula, sp. near C. striatella.

The only trace of man hitherto found in these deposits consists of a
chipped stone scraper or hatchet discovered by Mr. Hacket 7 situ near
the village of Bhutra, eight miles north of Gddaward.*_ The material is Vin-
dhyan quartzite, and the form similar to that of some of the implements
fo und inthe lateritic deposits of Southern India, and in the postpliocene
formatious of Europe.

The only form identical with existing Indian species is Rhdnoceros
unicornts, originally described under the name R. namadicus, but accord-
ing to Mr. Lydekker the bones are not distinguishable from those of the
living species? LElephas namadicus is allied to the existing Indian ele-
phant, Bubalus palzindicus is very close to the living Indian wild buffalo,
and the deer is a near relation to, if not identical with, the bdrasingha
(Cervus duvaucelli), Onthe other hand, Elephas insignis and Hippo-
potamus namadicus belong to extinct subgenera, the first being found,
and the latter represented by a nearly allied species, in the pliocene Siw4lik
rocks. Hippopotamus palzindicus and Bos namadicus are not nearly
allied to any Indian living species, the first belongs to a genus now only
found in Africa, whilst the second, although having some characters in
common with the living wild cattle of India, Bos (Brbos) gaurus, differs from
the latter in many important particulars, and appears to be quite as closely
connected with true taurine oxen belonging to the type of Bos taurus.
Bos namadicus, indeed, cannot be classed in the subdivision Bzdos. The
relations of the remaining mammals are less distinctly made out, the speci-
mens on which the species are founded being for the most part fragmentary.

The only reptile clearly identified is Emys tectum, which is considered
identical with a living Indian form. It is very singular that only frag-
mentary remains of crocodiles occur, for they abound in the Siwdlik
rocks and a species is common in the Narbadé at the present day. The mol-
lusca appear to be the same as species now living in the area, and all the

' Memoirs, Il, 284, (1860); Records, VI, | mens having been preserved in the Geological
54, (1873). The nomenclature in this list | Museum.
is that adopted in the first edition of the 4 Records, VI, 49, (1873); two figures of
Manual. The species marked with an asterisk | the implement are given.
are not determined with certainty, no speci- 2 Pal Indica, series x, I, p. viii, (1880),
400 GfOLOGY OF INDIA—PLEISTOCENE AND RECENT

CChap. XVI.

commonest forms now known to occur in the river valley are represented,!
except some minute species of land shells. Their absence is not surpris-
ing, because land shells for the most part float, when washed away, and
are left on the surface, where they decompose, instead of being preserved
in alluvial deposits. :

The examination of the molluscan remains in the Narbad4 clays and
gravels completely disproves the idea of a marine origin, but it has been
considered by some observers that the deposits are lacustrine.? This view
was principally based upon the uniform appearance of the clay and the
absence of stratification. But this very uniformity and want of strati-
fication are common characters of undoubted river deposits, and may be
observed on the banks of most large streams, whilst the frequent deposi-
tion of pebble beds throughout the clays could not have taken place in
the still waters of a large lake. The bones too are isolated and broken,
sometimes even being rolled, whereas, if deposited in a lake, different
bones would in all probability be found together, because away from the
margin there could be no current in the lake of sufficient force to trans-
port bones divested of flesh, and any mammalian remains deposited in
the bottom of the lake must be derived from floating carcases or portions
of carcases. Moreover, the Chelonia and fresh water mollusca are all
forms which inhabit either rivers or shallow marshes in river valleys, and
it is improbable, if so great a change took place in the area as would
be involved in the replacement of lakes by a river valley, that a greater
difference would not be produced between the tortoises and fresh water
shells formerly inhabiting the waters and those still living.

The fact of the alluvial formation occupying a rock basin shows,
however, that a considerable upheaval of land must have taken place to
the westward, and it is possible that this upheaval may for a time have
given rise to a lake, and the lower beds may consequently be lacustrine
even through those from which the fossils were obtained are alluvial. If
the Narbadé has really been diverted from its original course as suggested
above this could only have happened through a movement of elevation
sufficiently rapid to pond back the drainage and produce a lake.

The alluvial plains of the T4pti valley require but brief notice.§ In

' The only exception of any importance
is Melania spinulosa, but that is not by any
means so generally distributed a form as
M:tuberculata. The absence in the Narbadd
deposits of Melania variabilis and M. spinu-
losa, the latter of which is included in Mr.
Theobald’s lists of living Narbad4 species
(Memoirs, 11, 287), was noticed by Dr. Fal-
coner, Quart. Four. Geol. Soc., XXI, 382,

(1865) ; but it is extremely doubtful whether I.
variabilis does exist in the Narbada valley or
its neighbourhood. The occurrence of M,
lyrata included in Mr. Theobald’s list, loc. cit.,
is also very dcubtful.

2 Memoirs, \1, 283, (1860).

® For a few additional details, see Memoirs,
VI, 276, 286, (1859); and Wynne, Records,
II, 1, (1869),
Chap. XVI. ] TAPTI ALLUVIUM. gor

their principal characters they resemble the Narbadd plain, but the depth of
the deposits is unknown, no deep borings having been made, As in the
Narbada valley, the river now runs at a considerable depth below the alluvial
plain and is evidently cutting its channel deeper. The whole basin is com-
posed of the Deccan trap, and the Tdpti cuts its way out to the westward
through the same formation. No remains of mammalia have hitherto been
detected in the alluvium, but they will probably be found if sought after;
the few mollusca found, as in the Narbadé plain, belong to recent fresh
water species inhabiting rivers.

The difference in elevation between the Tdpti and Purna plains is not
accurately known, nor are the levels of different parts of the plains well
determined, the only data available being those furnished by the railway.
The height above the sea at Bhusdwal, just south of the alluvial flat, near
the eastern extremity of Khandesh, is 677 feet. This cannot be much
ahove the flood level of the TApti river, for the rail level at the bridge over
the Tapti, only about six miles distant, is 685 feet. At Malkapur, close to
the western extremity of the Purna alluvial plain, the level is 816 feet, at
Akola 917, at Murtazdpur 986, and at Badnera south of Amraoti, 1,093.
The last locality, however, is some miles distant from the south-eastern
edge of the alluvium, and none of the railway stations are out in the allu-
vial plain, as in the Narbadd valley.

The only peculiarity of the Purna alluvia: deposits, which deserves notice,
is the occurrence of salt in some of the beds at a little depth below the sur-
face. Throughout an afea more than 30 miles in length, extending from the
neighbourhood of Dahih4nda (Dhyanda), north of Akola, to within a few
miles of Amrdoti, wells are sunk for the purpose of obtaining brine in several
places on both sides of the Purna river, The deepest wells are about 120
feet deep. They traverse clay, sand and gravel, and finally, it is said, a
band of gravelly clay, from which brine is obtained. No fossils have
been found in the clay and sand dug from the wells. The occurrence of
salt in the alluvial deposits of India is not uncommon, and it is impossible
to say, without further evidence, whether it indicates the presence of marine
beds. The absence of marine fossils in all known cases is opposed to any
stch conelusion, but still it is not impossible that the land may have been
1,000 feet lower than it now is in late tertiary, or early posttertiary times,
and this difference in elevation would depress the Purna alluvial area
beneath the sea level.

It has already been mentioned that the alluvial deposits of the Godavari
do not occur in distinct basins, like those of the Narbad4 and T4pti. This
river in general has but a slight fall, and forms a broad alluvial plain where it
traverses softer beds, whilst it cuts a steeper slope through harder rocks.
There is an exception to the latter rule in the gorge above Rdjdmahendri.
402 GEOLOGY OF INDIA —PLEISTOCENE AND RECENT. [ Chap, XVI

Extensive alluvial areas occur along the upper part of the Godavariin the
Bombay presidency and the adjoining portion of the Nizam’s dominions,
and similar tracts are found on the Pengangd, Wardha, and Waingangd,
tributaries of the Godavari, in Berar, NAgpur and Chanda.

_The composition of these deposits differs in no important particular
from that of the Narbad4 and Tdpti alluvium, The gravels are chiefly
composed of rolled agates and fragments of basalt derived from the Deccan
traps, which are the prevailing rocks in the upper part of the valley.
Silicified fossil wood in all sizes from smal) fragments up to trunks ro
and 15 feet long! is abundant along the west margin of the Chikidla sand-
stones, from near the Wardha to the Godavari near Enchapalli, and is found
less abundantly from here on to Albaka on the Godavari. The greater por-
tion of the alluvium in all cases consist of brown clay with kankar. In the
Wardha valley beneath the clay and calcareous conglomerates some fine
sandy silt, light brown or grey in colour, occurs west of Chdnd4, and
contains salt, with a considerable proportion of sulphate of magnesia ®
(Epsom salts),

Mammalian bones have been found, sometimes it is said in large num-
bers, in the Godavari valley, but very few appear to have been preserved,
and the only species identified is Ele¢has namadicus Bones of Bos and
other animals occur, and it appears probable that the fauna is similar to
that of the Narbadd valley. From the gravels near Mungi and Paitan
(Pytun) on the road from Ahmednagar to Jalna, Mr. Wynne obtained an
agate flake, * apparently of human manufacture, thus affording a second
instance of traces of man occurring in the pleistocene river gravels of
the Peninsula.

The most important localities at which bones have been observed are
the neighbourhood of Méngi and Paitan already mentioned, and one or
more places on the Pengangé or its tributaries in the neighbourhood of
Hingoli.® At one spot near Hingoli bones are said to have been found in
immense quantities, but unfortunately they were not preserved.

The valley of the Kistna resembles that of the Godavari in. many
respects. There are similar plains of alluvial clay with beds of sand, gravel
and calcareous conglomerate, but none of these plains appear to be
of great extent. Beds of gravel have been observed in many placesat a
height of 60 to 80 feet above the present course of the river and its
tributaries.®

>? W. King, Memoirs, XVIII, 208, (1881). VIL, 477, (1838); Carter, on the authorit
: Hughes Memoirs, XIII, 92, (1877). of Dr. Bradley, Four, Bo. Br. Roy, As. 76
Falconer, Quart. ¥our. Geol. Soc., XX1, 381,| V, 304, (1854); Newbold, Four. Ray. As. Soc.

(1865), Memoirs, VI, 232, (1860). VII, 246 i
* Fora description by Dr. T. Oldham and (1869). " REDE ease Mena Re
figures, see Records, I, 68, (1868). § Newbold, Your. Roy. As. Soc., VIII, 247,

5 Capt.O, W. Gray, “Mad. Four. Lit, Sci., | (1846); Foote, Memoirs, XIL, 237, (1875).
Chap. XVLJ KISTNA ALLUViUM. 403

The only important mammalian remains hitherto found in the alluvial
deposits of the Kistna and its tributaries consist of portions of the cranium
and mandible of a Rhznoceros, and some bovine teeth and jaws, found on
the Gatparba near the town of Gokdk The bovine remains have not
been determined but the RAznmoceros has been described under the name
of R. deccanensts by its discoverer, Mr. Foote ;* the species differs widely
from all living forms, and does not appear to be very nearly connected
with any known fossil Indian species. Some fresh water shells of living
species were found with the bones.

It was probably from some part of the upper drainage area of the Kistna,
also, that Colonel Sykes obtained the teeth of a trilophodont Mastodon
described by Falconer ® under the name of M7. pandionts.

Large numbers of chipped quartzite implements of human manufacture,
and belonging to the same type as that discovered in the Narbada alluvium,
have been found in various gravels in the southern Maratha country on the
Malparba and other affluents of the Kistna.* The relations between
the ossiferous gravels and those containing the implements are, however,
somewhat obscure.

Nothing of importance is known concerning the older alluvial deposits
of the remaining rivers in the Indian peninsula,

It is in the Mahdnadi, Kistna, and Penner valleys that the principal
diamond gravels are found, frequently at heights considerably above the
present stream level. The pebbles in the gravels are composed of various
kinds of metamorphic and transition recks.

Throughout the east coast of the Peninsula, from the delta of the Ganges
to the neighbourhood of Cape Comorin, with the exception of a few miles
near Vizagapatam, there is a belt of alluvial deposits, varying greatly
in breadth, but nowhere exceeding about fifty miles, In places the hills
approach the sea, léaving only a comparatively narrow belt of sandy fore-
shore, as south of the Chilk4 lake in Orissa and again near Pondicherri,
whilst broad alluvial plains extend inland for many miles, near the mouths
of the great rivers Mahdnad{, Goddvari, Kistna, Cauvery, etc., where there
is actually a slight projection beyond the general coast line, owing to the
quantity of sediment deposited, although the strong currents which sweep
up and down the coast prevent any great seaward extension of the deltas.

To the northward the east coast alluvium joins the older alluvial de-
posits on the western side of the Ganges delta, and the two resemble each
other closely in mineral characters. The coast alluvium consists chiefly of

1 Memoirs, XII, 232, (1876). 4 Foote, Memvirs, XII, 241, (1876).
2 Pal, Ind., series x, I, pt. i, (1874). 5 Newbold, Sour. Roy. As. Soc., VII, 226,
3 Palzontological Memoirs, London, 1868, (1853).

I, 124.
404 GEOLOGY OF INDIA—-PLEISTOCENE AND RECENT, { Chap. XVI

clays with kankar and, near the hills, pisolitic nodules of iron peroxide, the
latter being in places sufficiently abundant to render the deposit a kind of
laterite gravel. Gravels and sand also occur, frequently more or less
mixed with ferruginous concretions, and there is, in many localities, an
apparent passage between the ferruginous gravel of the ailuvium and the
low level form of laterite, but in other places this older alluvium rests un-
conformably upon the low level laterite, which has been shown, by the
occurrence of palzolithic implements, to be itself of posttertiary age.

The surface of the coast alluvium is usually quite flat near the sea and
in the river deltas, but towards the hills it is more uneven, and the surface
has undergone a considerable amount of denudation, evidently from being
at a higher level.

At Madras and Pondicherri, shells belonging to recent species have
been found at depths of from 5 to 20 feet beneath the surface, or con-
siderably above the present sea level. Farther south also, near Porto Novo
in the lower valley of the Vellar,! a bed of estuarine shells is found above
the present flood level of the river, and consequently at a considerable
height above the sea Similar deposits of shells have also been noticed
near Cuddalore and Tanjore.”

The shells, as a rule, are estuarine forms, such as now live in the creeks
and backwaters of the coast,® but in several cases true marine species have
been found. The subfossil shells near Madras are so abundant in places
that they have been collected for burning into lime.

Another place where estuarine shells have been observed is close to
the Chilk4 lake insouthern Orissa. The forms found were Cytherea casta
and Arca granosa, and the deposit containing the shells is now at elevations
of from 20 to go feet above the level of the highest tides.

The thickness of the alluvium has been tested at Madras by a boring
which went through it, and struck the crystalline rocks at 55 feet from the
surface.* Further south, at Pondicherri, the thickness of the alluvium is
much greater, one boring having been put down 550 feet without reaching
its base, The alluvial deposits of Pondicherri are both interesting and im-
portant, in that they yield a supply of artesian water at various depths below
the surface,° and in one boring, at Bahur, a bed of lignite, 10°65 m. (35 feet)
in thickness, was struck at a depth of 73°38 m. (240 feet) from the surface.§
The lignite is too impure to be of commercial importance, but since. it

HF. Blanford, Memoirs, IV, 192, (1853). | tilis, Arca granosa, Cythera casta, C. mere-
? King and Foote, Memoirs, IV, 254, (1864). | tréx, Ostrea, a large species.

8 The following are the most characteristic | 4 Newbold, four. Roy, As. Soc., VIII, 248,
species. They are seldom, if ever, found in (1846).

the open sea, but they are always met with in| 5 For details see W. King, Records, XIII, 113,
backwaters, and at the mouths of rivers, and | jy4, (1880). :

many of them occur in creeks of daltas near| 6 Geological Survey, MS. Records.
the sea:—Potamides telescopium, P. fluvia-
Chap. XVI] SMOOTH WATER ANCHORAGES, 405

must have been formed at or near the surface, it is interesting as evi-
dence of an amount of subsidence corresponding to the depth at which it
was found beneath the surface,

Evidence of subsidence to a less degree is again found in a submerged
forest at the western end of Vdlimukam bay in the Tinnevelli district. The
forest, or rather so much of it as can be seen, is described as about half an
acre in extent, ly.ng at or just below high water mark; the stumps have
a diameter of one to one and a half feet at the base of the hole, and are
surrounded by black mud containing remains of twigs and detached
branches, An incised bone pendent was found, which appeared to have
been washed out of this mud, showing that the forest flourished since the
advent of man.

The trees of this forest could hardly have flourished at sea level
or on the ccast, so that there has certainly been some subsidence in
this neighbourhood, but indications of a contrary movement are found close
by in the occurrence of Potamddes and other littoral marine shells in clay

above high water level, showing that this clay must have been elevated
since it was formed.'.

Before proceeding to the description of the recent accumulations on the
west coast of the Peninsula it will be well to notice the remarkable
smooth water anchorages of Aleppi and Narakal. These are mud
banks of about four miles in length, whose position varies in the course of
years within the extreme limits of about eleven miles. The sea bottom
on these banks is composed of a very soft mud, which readily mixes
with the sea water, and smooth water can always be found over the
mud banks, though open to the full force of the south-west monsoon,
however tempestuous the sea outside may be. It was this peculia-
“rity which first attracted attention, and rendered them important to the
navigators of a coast where there arc no sheltered harbours, and the
accounts which have been written from time to time constitute a
tolerably extensive literature.2 According tothe most recent investigation
of the subject, these smooth water anchorages owe their origin to a bed of
very soft, fine grained, greenish clay, containing foraminifera and diato-
macez, which underlies the soft recent sandstones of the surface of the
narrow strips of land separating the sea from the backwaters of Travan-
core and Cochin. When the water levelin these backwaters is raised
by the monsoon rains, this mud is forced outwards, and rises in cones
and ridges along the shore and under the sea, and once it has become

1 Foote, Memoirs, XX, 83, (1883). ature will also be found. A more recent in-
2 A good account of these mud banks by Dr. | vestigation by Mr, Lake is printed in Records,
W. King is published in Records, XVII, 14, | XXIIL, 41, (1890).
(1885), where an account of the previous liter-
406 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [Chap. XVI.

thoroughly mixed with the sea water the waves of the open sea are
smoothed off and reduced in size over the mud banks.

‘This result appears to be due to two separate causes. In the first place
the mud contains an appreciable proportion of oily matter, and the action
of oil in stilling stormy waters is now well known, but the second cause
appears to be much the more important. The large quantity of impalpable
mud mixed with the water increases its density, and, consequently, the
waves, on entering this denser water, decrease in size and are retarded.
Moreover, as the proportion of mud is much less at the surface than lower
down, the lower part of the wave is retarded more than the upper, and the
wave may actually break if the increase in density be sufficiently rapid,
or merely be obliterated if it is sufficiently gradual. This action is
intensified by the large amount of fresh water falling on the sea as rain and
poured out by the rivers, which floats on the surface in such quantities that
ships may often replenish their stock of fresh water by dipping over the
side of the vessel with a bucket. It is doubtless due to the greater density
of the deeper Jayers of water, owing to the smaller proportion of salt and
mud in the upper layers, and the consequent retardation of the lower
portions of the larger waves that they are broken up, while the film of oil
derived from the mud causes the smaller wavelets to be smoothed off.

There is no such continuous plain of alluvium along the western shore
of the Peninsula as onthe east coast. The ground between the Sahyédri
range and the sea, where not hilly, consists geneially of a gentle slope
towards the coast, composed of rock, covered in many places by laterite.
The coast itself is rocky in parts, and the alluvial deposits are chiefly
confined to the neighbourhood of the small streams, which can from the
Western Ghats to the sea, or of the backwaters, or lagoons, which have
been cut off by banks of sand along the coast. The backwaters are of

’ considerable extent in Travancore and Malabar, but they are wanting
farther north and on the coast of the Bombay presidency. The alluvial
valleys between the hills are unimportant south of Bombay itself, although
they gradually increase in exterit to the northward,

Alluvial plains, evidently of comparatively recent formation, connect
the hills of Bombay and Salsette island, a few creeks alone remaining
to show the position of the marine channels which formerly existed. Farther
north these plains gradually increase in extent, until they merge into
the alluvial flat of Gujarat, ‘

At Bombay the alluvial deposits consist of blue and yellowish brown
clay, The former varies in thickness from a few inches to several feet, its
upper surface being at present about one or two feet below high water
level. It is very salt, and contains small grains and nodules of kankar, and
occasionally plates of gypsum ; it is frequently penetrated by mangrove
Chap. XVI J ALLUVIUM OF THE WEST COAST, 407

roots, which are usually riddled by Zeredo borings, just as in the mud of
tidal creeks, and at one spot large masses of oyster shells have been found
init. The yellowish brown clay appears to be the older of the two deposits.
Its surface is frequently above the sea level, it abounds in larger masses
of kankar, and it has occasionally yielded estuarine shells, Placuna,
Ostrea, etc, That these alluvial deposits are estuarine, and precisely
similar to the aud now deposited in the creeks and backwaters of the
coast, or on the shores of Bombay harbour, is shown by the similarity
of mineral character and by the organic remains, both vegetable and
animal, found in the clay.)

Some very interesting indications of subsidence were found in the ex-
cavation of the Prince’s dock at Bombay. A large number of tree stems
and roots were found in the blue clay, many in the position in which they
originally grew and some of the stumps were 30 feet below high water
level. The evidence of subsidence here is unmistakeable, but the littoral
concrete, seen on the west side of the island, must have been furmed at a
lower level than it now stands at. The elevation on one side of the island
and depression on the other could not have been contemporaneous, so that
we have clear proof of oscillations of level similar to, but of greater extent
than, those Dr, Buist recognised many years ago.”

It is evident that Bombay harbour is the last remaining inlet out of
many which formerly indented the Bombay coast, and that this harbour
is gradually silting up and being converted into dry land, The process,
however, is slow, and it may be ages before its progress is such as to
affect the trade of Bombay, but, unless depression takes place in the area,
or means are devised for checking the deposition of mud, there can be
no question of the ultimate result. Except at Bombay, little has been
recorded concerning the alluvium of the western coast south of Damdn,
and that little presents no features of interest,

In the neighbourhood of-the rivers Tdpti and Narbad4 there is, how-
ever, a broad and fertile alluvial plain? near the sea, resembling in
some of its features the alluvium of the east coast. Commencing te
the southward near Damén, this plain covers the greater portion of the
Surat, Broach, and Ahmadabdd districts, and continues as far as the Rann,
where it joins the area of recent deposits connected with the Indus valiey.
Near Surat this plain is about 30 miles in breadth, and near Baroda it is
60 miles wide.

The alluvium of eastern Gujarat consists of brown clays with kankar,
resting upon sands and sandy clays with occasional gravels. The surface

1Buist, Trans. Bom. Geog. Soc, X, 181, | ? Trans. Bo. Geog., Soe, X, 177, (1857).
(1852); Carter, four. Bom. Br. Roy. As. Soc, | § Memoirs, V1, 233, (1869); Records, I, 30,
IV, 204, (1853). | (1868); VIIT 49, (1875).
408 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [Chap. XVL

is covered with black soil to the southward, though not in the district
of Ahmaddbéd, and is frequently flat over considerable areas, but in parts
of the country the ground is undulating, evidently in consequence of
having been denuded by rain action, The deposits appear to have been
chiefly estuarine or marine, and have probably been raised, as on the east
coast, but no fossils have been found. The Gulf of Cambay is said to be
gradually silting up, and there can be very little doubt that it was formerly
part of a broad inlet leading from the Rann, then an inland sea, to the
ocean, and that the remainder of the inlet has been converted into the
alluvial plains of Ahmad4bdd, Broach, Surat, and north-eastern Kathidwar,

In north-eastern KAthidwdr, on the borders of the Rann, there is a
large alluvial tract,! continuous with the alluvium of Ahmedabad, and
similar in character. Between Kdathidwdr and Ahmedabad, in the line of
depression between the head of the Gulf of Cambay and the Rann of Cutch,
there still exists a large shallow lake of brackish water, called the Nal, about
twenty miles in length by three or four broad. In the neighbourhood of
this marsh shells of a form of Cerzthium (probably Potamides telescopium
or P. fluviatilis) are found, showing that estuarine conditions have pre-
vailed at no distant period, and tending to confirm the probability that
the depression between Kathidwdr and Ahmedabdd is an old marine inlet,
silted up in recenttimes. The distribution of black soil in the neighbour-
hood of the Nal will be noticed presently.

Along the south coast of Kathiawér there is very little alluvium, its place
being taken by a calcareous grit, with marine shells, which is evidently of
late formation. A glance at the map will show that this coast is exposed
to the full action of the currents which sweep along the shores of the
Peninsula, so that it is unlikely that any accumulation of sediment would
take place. A patch of recent deposits has been mapped at the western
extremity of the Kathid4wdr peninsula, but along its north-western coast the
Deccan traps extend down the sea shore. The belt of alluvium reappears
in Cutch,? where it is from three to ten miles broad, there being only one
place where rocks come down to the shore. ‘Ihis is in the Gulf of Cutch.

The alluvial plain of Cutch consists of a brown loam, resting upon mottled
clay, with kankar and grains of quartz.

An agglutinated calcareous shelly grit is found, a little raised above the
sea level, in several places on the west coast of India. This deposit, which
was called littoral concrete by Dr. Buist,? consists of shells, corals, pebbles,
and sand, cemented together more or less thoroughly by carbonate of
lime, and sufficiently compact in places to be employed as an inferior kind

' Rogers, Quart. Four. Geol. Soc, XXVI, | 2% Wynne, Memoirs, 1X, 81, (1872).
118, (1870); F, Fedden, Memoirs, XXI,130, | ° Trans. Bo. Geog. Soc., X, 179, (1852); Four,
(1884). Bom, Br, Roy, As. Soc. 1V, 206, (1853).
Chap. XVI] LITTORAL CONCRETE. 409

of building stone, The best known locality is in Bombay island, where the
shelly grit forms the flat ground of the Esplanade and part of the surface
on which the fort was built, the same deposit is also found at Mahim and
other places in the island, resting sometimes upon rock, but more often
upon the blue alluvial clay, described a few pages back. The same form-
ation is found to the southward at Malwd4n,! and northward here and there
as far as Daman, where it was observed by Mr. Wynne, apparently in
process of formation.* Near Bulsdr, a little north of Damén, the littoral
concrete was observed to be stratified, the strata dipping at a low angle
towards the sea.

In western Kdthidwdr the same formation is much more widely deve-
loped. It here assumes the character of an earthy calcareous grit, is
usually of a dark ashy colour, and contains marine shells and corals. Occa-
sionally it attains a thickness of 60 feet, and it rests unconformably on
the denuded surface of the miliolite. The fossils found in the calcareous
grit, so faras is known, are all species now living on the neighbouring
coast, but no thorough comparison has ever been made.

There can be very little doubt that the shelly calcareous grits of the
Bombay and Kathidwér coast are truly marine, not estuarine, and that they
are the result of a littoral accumulation of the sand and pebbles found on
the shore, together with marine shells and corals. The beds may have ori-
ginally been sand spits or beach deposits, very little, if at all, above high-
water mark, and consolidated by the cementing action of carbonate of lime
after being raised. In any case there appears to be evidence of a rise in
the land, trifling at Bombay, but greater in Kathidwar.

Indications of local deposits, supposed to have been formed in lakes,
have been noticed on the Nilgiri hills of Southern India * and in the southern
Mardth4 country,! and have been supposed to indicate changes of level.
No fossils have been found in these deposits, nor does the evidence in
either case amount to clear proof of the former existence of lacustrine
conditions, although the probabilities are in favour of this view.

It would be beyond the scope of the present work to enter into the
question of Indian soils. Consisting as they do of the surface of the
ground altered by the action of the air and rain, by impregnation with
organic matter, and by the results of agricultural processes, they necessarily
vary with every difference in the underlying formation, whether it be one
of the older rocks or of the more recent unconsolidated deposits. There are,
however, two forms of superficial formations which, having received repeated

' Memoirs, XII, 243, (1876). 2H. F. Blanford, Memoirs, 1, 243, (1858).
2 Records, 1, 32, (1868). -4 Foote, Memoirs., XII, 228, (1876).
410 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [ Chap, XVI.

notice in Indian geological works, require a few remarks to be devoted
to them, and one of the two, the vegur, or black soil, is a very remarkable
substance. The red soil also requires notice, because it has been so
frequently mentioned in geological treatises.

The somewhat ferruginous soils Common on the surface of many
Indian rocks, and especially of the metamorphic formations, would pro-
bably never have attracted much attention but for the contrast they
present in appearance to the black soil. They have only been noticed,
as arule, in papers relating to the black soil country in the western and
southern portions of the Peninsula. The commonest form of red soil is a
sandy clay, coloured red by iron peroxide, and either derived from the
decomposition of rock ¢z s¢tu or from the same products of decomposition
washed to a lower elevation by rain. The term is, however, frequently
used in a very vague sense, apparently to distinguish such soils as are not
black, and hence many alluvial soils may be comprehended under the
general term. In very many cases, too, it appears ta have been applied
in Southern India to thick alluvial beds of sand or sandy clays, which
are in fact ordinary river or rain-wash deposits.

The regur of Peninsular India, called black soil from its colour, and
cotton soil from its suitability to the cultivation of cotton, occupies the
surface of a very large portion of the country, and Newbold considers that
at least one-third of Southern India is covered by it. The name
is a corruption of the Telugu regada, or of cognate words in affined
languages. ?

Kegur, in its most characteristic form, is a fine dark soil, which varies
greatly in colour, in consistence, and in fertility, but preserves the constant
characters of being highly argillaceous and somewhat calcareous, of
becoming highly adhesive when wetted (a fact of which any one who has
to traverse a black soil country after a shower of rainbecomes fully aware)
and of expanding and contracting to an unusual extent under the respec-
tive influences of moisture and dryness. Hence, in the dry season the
surface is seamed with broad and deep cracks, often five or six inches across

1 The following are some of the principal Carter, Your. Bom. Br. Roy. As. Soc. V;

writers who have described regur:— 329, (1854).
Christie, Edin, Phil. Four, VI, 119,| Theobald, Memoirs, 11, 298, (1860); X,
(1829); VII, 50, (1829); Mad. Four. Lit. 229, (1873).
Sci., 1V, 469, (1836). H. F, Blanford, Memoirs, 1V, 183, (1852).

Voysey, Four. As. Soc. Beng., Il, 303, (1833). | King and Foote, Memoirs, IV, 352, (1864).
Newbold, Proc. Roy. Soc., IV, 54, (1838.; | W. T. Blanford, Memoirs, VI, 238, (1869) ;

Four. As. Soc. Beng. XIII, 987, (1844); Records, VIII, 50, (1875).
XIV, 229, 270, (1845); Four. Roy. As, | T. Oldham, Records, IV, 80, (1871).
Soc., VIIL, 252, (1846). Foote, Memoirs, XII, 251, (1876).

Hislop, Four. Bo. Br, Roy, As. Soc, V,
61, (1853).
REGUR OR COTTON SOIL,

Chap. XVI] 4ul

and several feet deep. Like all argillaceous soils, regur retains water, and
consequently requires less irrigation than more sandy ground; indeed, as
a rule, black soil is never irrigated at all in the western Deccan, Nagpur,
and Haiderdbad. When dry, it usually breaks up into small fragments; on
being moistened with water it gives out an argillaceous odour, It is said
to fuse, when strongly heated, into a glassy mass, but this is not invari-
ably the case, and is probably dependent on the proportions of iron and
lime present.

The chemical composition ‘of regur has not received much attention.
From the few and partial analyses ! which have been made the proportions
of iron, lime, and magnesia seem to vary, and there appears always to hea
considerable quantity of organic matter combined. ‘Ihe black colour
appears to be due either to the carbonaceous elements of the soil, or to
organic salts of iron, but the tint varies much, being frequently brownish,
and sometimes grey.

Christie made some experiments to determine the absorbent power of
regur. He first dried a portion at a temperature nearly sufficient to char
paper-; he then exposed to the atmosphere of a moderately damp apart-
ment 2615°6 grains of the dried soil, and found after a few days that it had

1 The following are theanalyses. In neither
case is it stated how the analyses were made,
nor which ingredients were determined by
loss. In the first, by Dr. Macleod and pub-
lished by Captain Newbold (four. Roy. As.
Soc., VIII, 254), a complete analysis of a dried
sample appears to have been made, but the
locality from which the specimen was derived
is not stated :—

In the other analysis by Mr. Tween (Memoirs,
IV, 361), undried scil was used, and the com-
ponent parts were only cetermined in the
soluble portion. The residue in all con-
sisted chiefly of magnesia and alkali; in At,
B1, B2, there were traces of sulphuric acid.

A and B were from near Seoni, C from
Indore, D from BarwAni, and E from Burhdn-
pur; Seoni and Barwdni are in the Narbadé

 

 

 

 

 

Silica. ‘ ‘ , - 482 valley, and Burhdnpur in the Tdpti.
Alumina : 3 20°3 Ax, A2 represent the surface soil and
Carbonate of lime . “ » 160 subsoil taken from the same locality, A1 being
Carbonate of magnesia . . 102 the surface, A2 from § feet below surface. The
Oxide ofiron . . , ‘ 1-0 two marked Br, Ba, are, in like manner, the
Water and extractive . 2 43 soil and subsoil (3 feet deep) from one locality,
coe while C, D, and E are the soils taken from
100'0 only a few inches below the surface, Br is
— considered the best quality of soil:—
A B
Cc D E
1 2 1 | 2
Insoluble . . . . . e a . 62°79 47 61 628 63'7 68°61 57°91 61°80
Organic matter , . e es : 9°2 84 9° 87 72 38°97 7°65
Water. we oa os . 84] 7:6 8:2 65] 9% 99 7°35
Oxide of iron, . é ~ ‘ é é wv 15°9 10°9 118 6°76 436 5°7
Alumina... g . oy Cis 75 86 7'6 8"4 5°81 8°75 7°67
Carbonate of lime , e i ° é 1'2 11°89 5 13 1°57 4°28 8°53
100) Ico’ | 100 100° 99°35 98'90 | 98°70

 

 

 

 

 

 

 
412 GEOLOGY OF INDIA—PLEISTOCENF AND RECENT. [Chap. XVI,

gained 147'1 grains. He then exposed the same sample to an atmosphere
faturated with moisture, and found that the weight increased daily, till the
end of a few weeks, when it was found to be 2828-4 grains, The soil had,
therefore, gained 212°8 grains, or about 8 per cent.

As a tule, the purest beds of regur contain no pebbles, although this
soil usually abounds in kankar. Fragments of chalcedony or zeolite are,
however, often found in the black soil, where it is derived from tke decom-
position of basalt, and in Southern India regur occasionally contains debris
of the metamorphic rocks, sandstone or limestone, on which it rests.

Where uncultivated, black soil plains usually support but few trees,
and those, as a rule, of no great size, but the principal product is grass,
commonly growing to a height of three or four feet, but sometimes con-
siderably higher. The growth of grass on the uncultivated plains of India
is, however, greatly promoted, and the trees injured or killed, by the
universal practice of burning the grass annually in the dry season, so that
it is probable that the plains of black soil would support forest if left to
themselves.

The fertility of this soil is so great that some of the black soil plains
are said to have produced crops for 2,000 years without manure, without
having been left fallow, and without irrigation. On the other hand, some
varieties of black soil, occurring near the coast of Southern India, are com-
paratively infertile. .

_The typical appearance is only presented by this soil near the surface
of the ground ; if the regur is more than about 6 to 10 feet deep, it usually
passes down into brown clay with kankar. It is never, except where it
has been carried down and re-arranged as a stream deposit, met with at any
depth beneath the surface.

The distribution of black soil in the Indian peninsula is of some im-
portance, because it affords a clue to the origin of the formation. Regur
is found everywhere on the plains of the Deccan trap country, except in
the neighbourhood of the coast. A very similar soil is found locally
in the basaltic Rdjmahdl hills, but with this exception nothing of the
kind appears to be known in Bengal or the neighbouring provinces.
In Southern India, however, tracts of biack soil are found scattered
throughout the valley of the Kistna, and occupying the lower plains and
flats of Coimbatore, Madura, Salem, Tanjore, Rdmndd, and Tinnevelli.
There is but little on the Mysore plateau. Some occurs on portions of the
coast plain on the eastern shore of the Peninsula, and the great alluvial
flat of Surat and Broach in eastern Gujarat consists of this soil. The
soils of Ahmadabad are light coloured, but regur occupies the surface of
the depression lying between Ahmaddbdd and Kéthidwér, and connecting
the head of the Gulf of Cambay with the Rann of Cutch.!

1 Rogers, Quart. Jour. Geol, Soc, XXIV, 118,( 1870).
Chap. XVL] ORIGIN OF REGUR, 413

In many cases there cannot be a question that regur is simply derived
from basalt by surface decomposition, and it is not surprising that
numerous observers, from Christie and Voysey to Carter and Theobald,
should have contended, and should still contend, that all cotton soil is
derived from disintegrated trap rocks. Throughout the immense Deccan
trap area, the passage from decomposed basalt into regur may be seen in
thousands of sections, and all the alluvial valleys, most of which contain
black soil, are filled with deposits derived from the «disintegration of basaltic
rocks. More than this, the boundary of the trap is approximately the
boundary of the black soil over enormous areas; where the latter is found
beyond the trap boundary, volcanic rocks may very probably have existed
formerly, and have disappeared through disintegration, or the soil have been
washed down from the neighbcuring trap hills. This is admirably seen
around Ndgpur and Chanda in the Central Provinces, where regur occurs
everywhere upon the trap, but is never seen upon the metamorphic rocks a
few miles to the eastward, except where there is reason to suppose it has
been transported, as in the alluvial flats of rivers which flow from the trap
country. Again, whilst nothing resembling regur is found in the meta-
morphic region of Bengal, Behar, Orissa, Chutid Nagpur, Chhatisgarh, and
the neighbouring provinces, soils, undistinguishable from those of the
Deccan traps, are found in the basaltic Rajmahdl hills, and a similar
formation has also been observed in Pegu,! derived from the decomposition
of basalt. It has been urged that basalt may have been more widely spread
in Southern India than is now the case, and that, where none is now found,
its disappearance is due to its having been converted, by disintegra-
tion into, regur.

This view cannot, however, be accepted. In the first place, as was
shown by Newbold, basalt generally disintegrates into a reddish soil, quite
different from regur in character. This reddish soil may be seen in places
passing into regur, but the black soil is, as a rule, confined to the
flatter ground at the bottom of the valleys or on flut hill tops, the brown
or red soil occupying the slopes. Again, the masses of black soil in the
valleys of the Godavari and Kistna might be due to the alluvial deposits
having been derived from the trap rocks, through which both rivers flow
in the upper part of their course, but hundreds of square miles in the
basins of the Penner, Pdl4r, Cauvery, and other rivers still farther to the
south are composed of precisely similar regur to that of the trap area.
There is no reason for supposing that the Deccan trap ever extended to
the valleys of the rivers named, or can there be any reasonable doubt that
the alluvial flats contained in these valleys are maiuly formed from the
detritus of metamorphic rocks.

Captain Newbold considered? all regur to be of subaqueous origin in

1 Theobald, Memoirs, X, 229, (1873). |? Your. Roy. As. Soc., VIII, 256, (1846).
4t4 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [Chap XVI,

India, and compared it to the deposits in tanks, and to the mud of the
Nile. Mr. H. F. Blanford suggested! that the cotton soil of Trichinopoli
had accumulated in lagoons or backwaters near the sea, and he showed
that in one place, near Pondicherri, regur was actually being formed
in a nearly dry lagoon separated from the sea by a sand spit. Messrs,
King and Foote, on the other hand, considered? it more probable that the
Trichinopoli regur was a fresh water deposit accumulated in marshes. It
has since been shown ® that a complete passage takes place in the neigh-
bourhood of Surat between the deposits formed in tidal estuaries and the
regur of the surrounding country, and it appears probable that much of
the black soil of eastern Gujarat may have been originally a marine or
estuarine (brackish water) formation, Qn the other hand, Hislop* object.
ed to the theory of formation by deposition in water, and he appears to
have been the first to suggest that regur may really be of subaerial origin
and due to the impregnation of certain argillaceous soils by organic
matter. This appears to be the most probable theory ; there can be no
doubt that some forms of regur originate from the decomposition of basalt
tz sétu, others from the disintegration of other argillaceous rocks, whilst
other varieties again were originally alluvial clays formed in river valleys,
or deposited in fresh water marshes, estuarine flats, or salt water lagoons.
The essential character of a dark colour appears due in all cases to the
admixture of organic matter, and perhaps the presence of a small quantity
of iron. It is far from improbable that most of the black soil flats of India
were covered with luxuriant forest, before the vegetation was annually
exposed to the effects of fire. ‘he increased dampness of the soil, the
protection from denudation by rain, and the supply of decomposing vege-
table matter may have contributed to the formation of the more fertile
forms of regur. That the process of regur formation is purely superficial,
and that it is due to surface action of a past time, is well seen in many
of the regur plains with a slightly undulating contour. In such places the
earth is black on the flats above, where the superficial layer has not been
washed away. brown where the wash of rain has swept away the surface
soil, and the black soil washed from the sides of the hollows has frequently
accumulated towards the lower portion of them.

The abrupt termination of regur in places at the edge of the trap
country is simply due to the change from an argillaceous soil to a sandy
one. The basalt appears generally to decompose into a highly aluminous
substance, the met rorphic rocks, on the other hand, produce sand to a
large extent. At the same time it should be stated that it is not quite
clear why argillaceous deposits should have become regur in Southern
India, whilst nothing of the kind is known in Bengal, except in the basal-

1 Memoirs, \V, 191, (1863). 3 Records, VIM, 50, (1875).
2 Memoirs, \V, 357, (1864). 4 Four. Bom, Br, Roy. As, Soc. V, 61; (1857).
Chap. XVI J BLOWN SAND, 415

tic region of the Rajmahdél hills. A dark coloured soil certainly forms in
the marshes of eastern India, but it has not the character of regur, and
no cotton soil has been noticed in the dense forests of Chutid N4gpur and
Bastdr, nor, except on the surface of basalt, in the forest-clad plains of
Burma. It is doubtful whether true regur occurs on the Malabar coast
between Bombay and Cape Comorin, and the marshy soils on the top of
the Sahyddri range do not form cotton soil. The black soil plains appear
to be almost confined to those parts of India which have a moderate rain-
fall, not exceeding about 50 inches, but it is impossible to say whether
this is a necessary condition.

It may then be stated that regur has been shown on fairly trustworthy
evidence to result from the impregnation of certain argillaceous forma-
tions with organic matter, but that the process which has taken place is
imperfectly understood, and that scme peculiarities in distribution yet
require explanation,

True peat forms in the hollows on the Nilgiris and some of the
other mountains in Southern India, such as the Shevaroys,! at elevations
above 4,000 feet, and its formation is due, as in temperate climates, to the
growth and decomposition of a moss, In the marshes of the Gangetic
delta an inferior kind of peat is also formed by the decomposition of
various aquatic plants, and especially of wild rice.4 The peat like beds
found so widely distributed in the neighbourhood of Calcutta at a little
depth below the surface appear to be derived from the decomposition of
forest vegetation, A scmewhat similar substance has been obtained from
beneath a marsh in Oudh.?

Sand drifted by the wind forms low hillocks on many parts of the
Indian coast. A series of parallel ridges of sand hills along the shore
of Orissa has been supposed to mark successive positions of the shore line.
A similar tract of blown sand is found north of Orissa in the Midnapur
district, and southwards at intervals throughout the whole of the east coast.
The sand is, of course, derived from the sea shore and blown up into ridges
at right angles to the prevailing wind, with their longer slope to windward
and a shorter and steeper surface to leeward. Smaller patches of sand are
sometimes found on the banks of backwaters. The sand hills frequently
extend for two or three miles inland from the coast, and in such cases the
inner ridges are covered with a peculiar vegetation, amongst which the
cashewnut tree (Anacardium occidentale) and a screw-pine (Pandanus) are
conspicuous, and in some cases between the parallel ridges coinciding in
direction with the coast the ground is flat, and even occasionally marshy,

1 Foote, Memoirs, XII, 252, (186). | ® Proc, As. Soc, Beng., 1865, p. 85s
2 Four, As. Soc. Beng., XXIII, 400, (1854).
416 CEOLOGY OF INDIA—PLEISTOCENE AND RECENT, (Chap. XVI.

as in parts of Midnapur. In the latter case it is probable that a lagoon
has existed, which has been gradually silted up, the origin of the lagoon
being due to the formation of a sand spit outside it. As already noticed,
the existence of several parallel sand ridges probably indicates a rise of
land, each ridge coinciding with a former coast line.

On the Malabar coast, sand dunes are equally common, and contribute
greatly to the formation of lagoons or backwaters! by accumulating on
spits of sand. In the northern portion of the western coast about
Bombay no sand hills have been noticed, probably because the detritus
from the trap rocks does not form a suitable material, but further
north again, in Surat and Broach,’ in portions of Kéthidwdr, and in Cutch,
blown sand occupies more or Jess ground in many places in the neighbour-
hood of the shore. ;

Sand dunes in India are not confined to the sea coast, but are frequent.
ly found on the banks of rivers. And the accumulation of blown sand on
river banks is of common occurrence or many of the peninsular rivers,
such as the Godavari, Kistna, and Cauvery. In some instances noticed
by Newbold,® villages have been buried by the sand blown from the river
beds during the dry season.*

One peculiar form of sand hill, known as ferz, is developed to a large
extent along the Tinnevelli coast, and to a small extent in the north-western
parts of Nellore and in the south of Travancore. The sand of which these
hills are composed consists of rounded grains of colourless quartz, stained
red, often bright red, by a thin film of ferruginous stain, which is easily
dissolved by acids. In the Tinnevelli district they owe their origin to the
dense clouds of sand and dust blown by the south-west monsoon off the bare
red soil plains towards the coast, where the wind meets the sea breeze, is
checked, and the sand dropped to form the téris.°

In the extrapeninsular area we find recent and subrecent river gravels
in every valley, but the more extensive accumulations, if we except the
alluvium of the Irawadi river, are all found in rock bound basins of closed or
arrested drainage, which have been formed by differential movements of the
surface during the elevation of the hill ranges among which they are found.

In the dry country west of the Indus there are extensive accumu-
lations of recent deposits, of which only asmall proportion can be regarded
as alluvium in the true sense of the word. Beyond the frontier there are
immense stretches of blown sand and loess in western Balichistdn and
Afghdnistdn, of which very little is known, but it is probable that they

Newbold, four. Roy. As. Soc., VIII, 268, | 4 The principal accumulation of blown sand
(1846). in India, that of the great RAjputdna Desert,

3 Memoirs, VI, 235, (1869) ; IX, 82, (1872). | will be described in the next chapter.

9 Four. Roy, As. Soc. VMI, 269, (1846). 5 R. B. Foote, Memoirs, XX, 87, (1883).

f
Chap. XVI] LOESS OF BALOCHISTAN. 417

are composed principally of the same types of accumulation as are seen in
the smaller valley plains around Quetta.

First among these, as being the oldest, is a series of usually more or less
bright red clays, sands and gravels which, in some of the publications of
the Geological Survey,’ have been regarded as tertiary, and have been
coloured as such on the accompanying geological map, but although it is
possible that they may belong to the newest portion of that period, they
ace so intimately connected with the recent deposits that, as has already
been noticed,® it is more convenient to describe them in this place rather
than in what might be regarded as their more proper place. These deposits
are frequently undisturbed, especially towards the centre of the vallc y plains
and are then difficult to distinguish from more recent deposits, except that the
latter are seldom so deep ared in colour, More usually, however, they
have undergone some slight disturbance, which has enabled the drainage to
cut into them and form an irregular surface dotted with smail hills, devoid
of soil or vegetation, owing to the saline nature of the clays and the steep.
ness of their slopes. Towards the margins of the valleys where these
deposits abut against the hills, they are sometimes tilted up at high angles
of dip, as in the Mashdlak range west of Quetta.

Though they occur in close proximity to typical Siwdliks, no actual con-
tact section has yet been found, but there are certain indications that the
red clays of the valleys are considerably newer than those of the Siwdlik
system, and it is certain that they were deposited after the main features
of the orography had been marked out by disturbance and erosion.’

The most important of the recent deposits of these plains are the
extensive gravel slopes at the foot of the hilis, and the loess.

The great gravel slopes, or dhdman, which everywhere fringe the
foot of the hills, and often reach a width of many miles in this compara-
tively rainless country, form one of the most conspicuous features in the
scenery. of the more open parts of the hill country west of the Indus.

 

 

 

Fig. 21.—Diagram illustrating the theory of the karez. PP=limit of permanent saturation of
subsoil ; K.K =karez,

They occur as great inosculating fans, spreading with a slope of 300 to
600 feet per mile from the mouths of the stream valleys. It is into these

1¢C, L. Griesbach, Memdirs, XVIII, 18, 3 These valley deposits have not yet received
(1881) ; W. T. Blanford, Memoirs, XX, 115, | the detailed attention they deserve. See
(1883). Records, XXV, 36, (1892); see also Memoirs,
2 Supra, p. 319. XVII, 18 (1881); XX, 115, (1883).
418 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [Chav. XVI.

fans that long underground tunnels, known as fares, are driven,
with a slope less than that of the surface, till they pass below the level of
permanent saturation, and, acting as a subsoil drain, carry the water out
to the surface.

The loess deposits consist of a fine grained, usually grey coloured and
unstratified accumulation of wind blown dust, precisely similar to the great
loess deposits of China, which have been described by the Baron von
Richthoven. They vary in size from small patches of a few yards across
to great plains like that of Thal Chotidli. In the Kachi, as the plain
south of Sibi is called, the deposits of the plain appear to be principally
wind blown loess, more or less mixed with true alluvium.

Closely connected with the true loess is a more or less finely stratified
type of deposit, which is formed in the low lying parts of the loess plain.
After every heavy shower the drainage from the higher parts of the plain,
as well as from the surrounding hills, collects in these depressions, whence
it gradually disappears by percolation and evaporation. The water, when it
first collects in these depressions, always carries a large amount of solid
matter in suspension, which is deposited when it comes to rest, the coarser
particles sinking first of all and the finer afterwards. By a repetition of
these floods, a finely bedded accumulation of alternately finer and coarser
grained material is formed, which presents a great similarity to a lacus-
trine formation, though it was not deposited in a lake in the true sense of
the word, but in mere temporary collections of flood water.

On the great plains of Rdwalpindi (known as the Potwdar), Bannu,
and Peshawar, extensive deposits of gravel, sand, and silt exist. Little
is known about the later deposits in the Peshawar and Bannu plains,
but those of the Potwdr present some features of interest. The sur-
face consists of a rather light brown alluvial clay, often containing
kankar, and passing in places into fine silt. Beneath this alluvial de-
posit there is a mass of gravels and sand, sometimes enclosing boulders
of large size. The boulders are not, however, confined to the pebble
beds, many have been observed imbedded in fine silt, and this circum-
stance, together with the great size of many of the blocks found, and the
distance to which they have been transported, has induced several observ-
ers to attribute the transport of the larger masses to ice, whether floating
down a river or ina lake. It has been suggested that the Potwdr
may have been converted into a lacustrine basin in post-tertiary times by
the elevation of the Salt range and the ridges west of the Indus. There
is but little evidence in favour of this view, but still it is not impossible,
for, although the pebble beds underlying the finer silt of the Sohdn valley

For additional details concerning these al- | 122, 140, 223; (1877), XIII, 221, (1880).
luvial deposits of the Potwdr, see Records, X, ;
Chap. XVI] INDUS FLOODS. 419

appear too coarse for lacustrine? deposits, the silt may be, in part at least,
a later deposit.

The posttertiary deposits are quite unconformable to the Siwédlik
rocks, which had been greatly disturbed and denuded before the later
beds were formed. These later beds themselves, however, are occasionally
found dipping at a considerable angle, due, it is said, to original deposi-
tion. The pebble beds ate found around Rdwalpindi and in the neigh-
bourhood of the Indus ; they overlie the Rotds gorge near Jehlam, occur
on some of the Salt range plateaux, and cap the mountain above Kala-
bagh on the Indus. They are found at a considerable elevation above the
‘present river beds, some fragments of crystalline rocks in the neighbour-
hood of the Indus, apparently brought down by the stream, having been
observed 2,000 feet above the river.

The large blocks attributed to ice flotation appear to have been derived
from the Himalayas. They are abundant along the Indus as far up as
Amb on the left bank of the river, in the gorge of the Siran and for some
miles below Attock, around Jhand about twenty miles farther south, and
farther still to the southward near the village of Trap on the lower course
of the Sohdn. Some of the blocks measure nearly 50 feet in girth and
others are even larger. In places such blocks have been found 20 miles
away from the banks of the Indus. :

The Indus, as is well known, is subject to extraordinary floods, due to
a portion of the upper valley becoming blocked by landslips (or according
to some by glaciers) and to the sudden destruction of the barriers thus
formed. Such floods occurred in 1841 and 1858, and have doubtless taken
place in past ages.t In the flood of 1841 the waters of the Kabul river
were checked and forced backwards for twenty miles by the rise of the
Indus, and Drew has shown that the lake in Gilgit, formed by the landslip
in 1840-41, must have been 35 miles long and upwards of 300 feet deep.
Enormous quantities of detritus must be carried down by the violent
floods produced by the bursting of such barriers, and if, as appears
probable, the low temperature of the glacial epoch was felt in India, such
lakes at an elevation of 5,000 or 6,000 feet above the sea would have been
deeply frozen in winter, and large blocks from the river bed and dam might
easily have been imbedded in the ice, glaciers also in the north-western
Himdlayas must have been. more extensive than they now are, and the
formation of lakes dammed up by glaciers was probably of more common
occurrence than at the present day. Shaw? has called attention to the
occurrence of heaps of stone and gravel of all sizes brought 80 miles down

1 For accounts of these floods, see Cunning- | cially Drew, Jummoo and Kashmir Territories,
ham’s Ladak, London, 1854; Montgomerie, Four. | London, 1875, p. 44. Numerous references to
As. Soe. Beng., XX1X, 128, (1860); Shaw, High | other accounts are given by the last named
Tartary, Yarkand, and Kashghar, London, 1871,| writer.

p. 433, ete., and Appendix, p. 481; and espe-! 72. ¢, p. 486.
2E
420 GEOLOGY OF INDIA—PLEISTOCENE AND RECENT. [ Chap. XVIL

the Shdyak, one of the tributaries of the upper Indus in Laddkh, by blocks
of ice, and a similar action on a larger scale on the Indus may easily have
supplied the erratics of the upper Punjab. If the Potwar was a lake, the
dispersion of the erratic blocks is easily understood ; if not, the area over
which the masses of rock are found may be due to variations in the course
of the Indus, and the reversed flow of its tributaries in great floods.

In one locality near Fatehjang a number of land and fresh water shells
were found in silt, apparently the same as that in which boulders are else-
where imbedded. The species found, including Lymunea rufescens, Plan-
orbis exustus, Paludina bengalensts, Bythinia pulchella, Melania tuber-
culata, Bulimus insularts, Opeas gracilis, etc., are the same as are now
common in the country, and it appears doubtful if they would have sur-
vived any very great diminution of temperature. At the same time it is
possible that the beds containing shells may be of later date than those
with boulders.!

The recent and subrecent deposits in the Himdlayas are represented
by lacustrine deposits, moraines, talus accumulations, and more conspi-
cuously, by the river gravels abundantly developed in nearly every valley,
as well as along the outer foot of the range. So far as these require notice
in this work, they will be referred to in the chapter devoted to the Himé-
layan range, but there are three larger expanses of recent and subrecent
deposits in Kashmir, Hundes and Nepdl which, being extensive enough
to be depicted on the accompanying geological map of India, require some
description of their principal characteristics.

The alluvial basin of Kashmir has a length of about 84 miles witha
breadth of some 20 to 25, and is in part occupied by low lying alluvial
deposits, not much raised above the level of the Jehlam river, but prin-
cipally by older deposits forming elevated plateaux on the borders of the
alluvial plain and islands rising from it. These elevated plateaux of.
alluvial and lacustrine deposits are locally known as Karewa, a name
which has been adopted for the deposits of which they are formed.

The Karewa deposits consist principally of sand and shingle, with
some fine grained clayey silt towards the centre of the valley; the beds are
for the most part horizontal or slope with a gentle dip, which is probably
merely the original slope of deposition, but near the flanks of the Pir
Panjal they dip away from the hills at angles rising to 20°.

The best published descriptions of these beds are those of Colonel
Godwin-Austen? and Mr. Drew,? by both of whom they are regarded as of
lacustrine origin, an opinion also adopted by Mr. Lydekker‘ but difficult
to accept inits entirety. It is very probable that some of the finely bedded

1 Theobald, Records, X, 141, (1877). 8 Jummoo and Kashmir Territories, p. 210.
® Quart. Four. Geol, Soc., XX, 383, (1864). | 4 Memoirs, XXM, 72, (1883).
Chap. XVI] VALLEY OF NEPAL, 421

fine grained deposits described by Colonel Godwin-Austen were deposited
in still water, but the frequent alternations of beds cf shingle with sand
and the layers of lignite from one to three inches in thickness, point to
subaerial conditions of formation. Even the presence of true lacustrine
deposits does not prove that the whole of the Kashmir lake basin was ever
occupied by a lake. This rock basin was probably gradually formed by a
deformation of the earth’s crust, and the hollow so produced was filled up
almost, if not quite, as soon as formed. At the present day true lacustrine
deposits are being formed in those places on the northern limit of the
valley where, owing to a deficiency of deposition, hollows have been left
in which water has accumulated, and it is probable that the conditions
have been much the same as at present throughout the geological history
of the Kashmir valley, and that a minor area of true lacustrine deposits has
been accompanied by a greater area where subaerial accumulation of sedi-
ment has been in progress,

It is possible that some of the older beds of the 4avewas may be con-
temporaneous with part of the upper Siwdliks, but the only fossils yet
found, besides undetermined fish scales and plant remains, have been land
and fresh water shells, all apparently belonging to living species.

The only other valley at all comparable with that of Kashmir is Nepal.!
The superficial differences correspond with those that mark the struc-
tural characters of the two regions; both are lengitudinal valleys, lying in
the general strike of the strata, but the clear open oval area of Kashmir
approximately coincides with the elliptical synclinal depression of the
calcareous upper palzeozoic strata. Nepal, on the contrary, is rather a
group of confluent valleys, with high dividing spurs in both directions. On
the prolongation of the strike of the rocks there is a continuation of the
special excavation of the mountain zone, and the rocks of this zone, being
prevailingly calcareous, has suggested the conjecture that the feature is
primarily due to erosion by solution, as may also be the case with Kashmir.
Another cause, however, and the proximate one of the formation of a lake-
basin in Nepal, was probably, in part, a relative rise of the hills on the
south, for here also the bottom beds of the valley deposits have undergone
local disturbance on this side.

These deposits correspond very closely with those of Kashmir. There
is no remnant of a lake, but the other features are alike. An extensive
upland area, known as tdur land, corresponds to the karewa of Kashmir,
and to the b4éngar of the Gangetic plains. It is the surface of the old
deposits, no doubt considerably modified by waste in the central parts,
and by rainwash accumulations near the hills. The streams flow at a
depth of from 50 to 500 feet below this surface, according to position, but

1H. B. Medlicott, Records, VII, 93, (1875)
2E2
422 GEOLOGY UF INDIA—PLEISTOCENE AND RECENT. (Chap. XVI.

here, as in Kashmir, they are now, for the most part, subject to overflow,
and thus form the alluvial valleys, known as fho/as, corresponding to the
khédir \and of the plains. Beds of serviceable peat, much used for brick
and lime burning, occur at various levels in the valley deposits, and there
is also a blue clay, extensively used for top dressing the fields, whose
fertilising virtue seems to be due to the phosphate of iron (vivianite)
freely scattered through it in blue specks. No fossil remains have as yet
been found in any of these deposits.

The subrecent deposits of Hundes occupy an area of some 120 miles
long by from 15 to 60 miles broad in the upper valley of the Sutlej, which
now flows in a deep and narrow gorge, not much less than 3,000 feet deep,
cut through the horizontal deposits it had formed at an earlier period of
its history.

Our knowledge of these beds is almost entirely dependent on the
description of the two brothers Richard and Henry Strachey. It seems
possible, from an observation of the latter of these, that the deposits are in
part of lacustrine origin, for he mentions that, in the central part of the
valley, the cliffs exhibit throughout their height a fine homogeneous clay,
with but little gravel in it. There is some inherent probability inthe sup-
position that part of these deposits were formed in local accumulations of
water, but there is nothing to show that the whole of them might not be of
subaerial origin, as it is almost certain that the bulk of them might have been.

But the chief interest of these deposits attaches to the mammalian
fossils they contain. These have long been known, though it was only
within late years that their derivation from the horizontal gravels was
definitely determined by Mr. C, L. Griesbach.! The earlier specimens,
brought across the frontier as curiosities by the Tibetan traders, had been
regarded as tertiary by the late Dr. Falconer, who considered that the
beds had undergone considerable elevation since their formation. A
more recent revision of the fauna by Mr. Lydekker* renders the reten-
tion of this opinion, itself improbable on account of the horizontality of
the deposits, impossible. With the exception of Wippothertum, which was
determined by Mr. Waterhouse from specimens brought by Sir R. Strachey
though it is questionable whether the materials were sufficient for its
determination, only living genera are known; of these Bos, Ovis (?),
Capra, and Eguus are genera still living in the highlands of Tibet.
Hyezna is not at present known in Tibet, though there is no reason why
it should not formerly have ranged into high altitudes, and, besides, the
correctness of the determination is open to question. There remains the
genus RAznoceros, which points to a warmer climate and a lower altitude
than that in which the remains are found. It is, however, not impossible

' Records, XAII, 91, (1880). | * Records, XIV, 178, (1881.)
Chap. XVI.) EASTERN HILLs. 423

for a Rhinoceros, especially one of small size, to have lived on the bushes
which grow in the neighbourhood of many of the Tibetan rivers, while the
doubtful evidence of this genus is more than outweighed by the fragment
of a skull figured by Royle,!' which agrees so closely with that of Pantholops
hodgson# that there can be little doubt of at least generic identity, and
Pantholops is a genus peculiar to the most elevated and coldest portions of
Tibet. It is consequently more probable, so far as the paleontological
evidence goes, that the subrecent deposits of Hundes were formed at or
near the elevation at which they are now found than that they were formed
at a much lower level and subsequently elevated without undergoing any
disturbance.

In the hills east of India there are some rock basins occupied by allu-
vial deposits, of which the best known is that of Manipur. About 50 miles
long by 20 broad, of an irregular shape, with many small hills rising like
islands from the alluvial plain, it is not surprising that it should be gener-
ally regarded as a filled in lake. There are, however, no real reasons for
supposing that any large proportion of the valley was ever occupied by
deep water, There are no terraces round it such as would have resulted
from the lowering of the outlet of the lake during the long period occu-
pied by its filling up. “The deposits are all of ordinary alluvial type, and
the courses of the streams show that the present surface is the result of the
gtadual subaerial formation of an alluvial plain, ‘The elevation of the sur-
rounding hills is, geologically speaking, of comparatively recent date, prob-
ably not dating further back than the latter portion of the tertiary period,
and it is probable that the origin of the Manipur basin was gradual, and that
the active erosion of the surrounding hills, due to abundant rainfall, caused
it to be filled up as fast as it was formed, with the exception of insigni-
ficant areas that partially escaped sedimentation, and were occupied by
shallow lakes.

The upper waters of the Chindwin (Kyin-dwin)| river drain a number
of alluvium filled valleys, the largest of which, on the upper part of the
Chindwin itself, is known as the Hukong valley, while, in the hills west of
the Chindwin, there are the Kubo valley and those south of it. No details
are at present known regarding these alluvial spreads of Upper Burma, as
no detailed geological investigations have yet been possible, but in Lower
Burma more extended investigations have been made, and it is possible
to distinguish posttertiary deposits of two distinct periods.

Along the margin of the Irawadi and Sittaung alluvium, there is a
broad, but interrupted, belt of undulating ground, clearly distinguished

‘Illustrations of the Botany, etc., of the | fig. 1. The original specimen cannvt now
Himalaya Mountains. 4° London, 1839, pl. III. | be found.
424 GEOLOGY OF INDIA —PLEISTOCENE AND RECENT. [(Chap, XVI

from the flat alluvial plains near the river both by the greater inequality
of its surface and by its more sandy character. This tract is locally
known as Eng-dain, or the country of the Eng tree (Dipterocarpus
grandiflora’, but the same name is naturally applied to the very similar
sandy tracts occupied by the pliocene fossil wood group, so that the
popular distinction does not precisely coincide with the geological limits
of the formation.

The Eng-dain tract is composed chiefly of gravel, derived in a large
measure from the neighbouring hills, but partly from a distance. A por-
tion of the deposits, like the diddar,the edge of the Ganges valley, may
simply be the detritus washed from the surface of the hills by rain and
small streams to form a slope at the base of the range, but in Pegu, as
in other countries with a heavy rainfall, this slope is inconsiderable, and a
great portion of the alluvial gravels are simply stream and river deposits.
Similar beds of sand and gravel are found in many places underlying the
argillaceous delta deposits of the Irawadi, and are evidently of more
ancient origin.

Besides the fringe, of variable width, formed by the gravels along the
edge of the older rocks, large tracts of the same older alluvial deposits are
found in places isolated in the delta, occasionally forming ground raised
to a considerable height above the flat country around. One such tract,
about 20 miles long from north-east to south-west, by 10 miles broad,..-
occurs east of Nga-pu-tau and south of Bassein; another, of about the same
dimensions, lies south-west of Rangoon. These areas may be ancient
bhéngar deposits, or they may be caused by local upheaval.

Except in the immediate vicinity of the river channel, there is no
important expanse of alluvial deposits in the valleys of the Burmese rivers;
the beds of all, immediately above their deltas, are formed in places by
older rocks, and there is no such continuous alluvial plain as is found
along the course of the Ganges and Indus. Small tracts of alluvium occur,
as usual, every here and there, but the wide undulating plains in the
neighbourhood of the river in Upper Burma are largely composed, not
of river alluvium, but of the pliocene fossi] wood deposits.

Compared with the Gangetic and Indus deltas, those of the Irawadi
and other Burmese rivers convey an idea of imperfection and backward-
ness, as though the latter were of more modern growth than the former,
and had made less progress towards the formation of a great fertile plain.
The Salwfn cannot be said to have any delta at all, and in the Irawadi
delta, as has already been mentioned, elevated tracts, both of rock and of
the older alluvial deposits, occur in the neighbourhood of the sea. Con-
sidering the size of the river, the Sittaung delta, if the alluvial plain ex-
tending to the northward beyond Taung-ngu (Tonghoo) be included, is
Chap. XVI IRAWADI DELTA, 425
proportionally more extensive than that of the Irawadi, but still the broad
Gulf of Martaban extends into the very mouth of the Sittaung river.

The Irawadi delta extends from the Rangoon river to the Bassein
viver, and the head of the delta may be placed near Myanaung.? The
first important distributary—that forming the head of the Bassein river—
leaves the main river a little above Henzdda, but water overflows in
floods some miles above Myanaung, and finds its way to the sea by the
Myit-ma-kha Khyaung, the origin of the Rangoon river. The various
tivers and creeks of the Irawadi delta are said to be far less liable to
change than those of the Ganges and Indus, but it must be remembered
that the authentic history of the latter rivers, and especially of the
Indus, extends much farther back than does that of the Irawadi. The
general surface of the delta near the sea, with the exception of the
higher tracts already mentioned, differs but little in elevation from that
of the great Indian rivers and Mr. Theobald considers that at least 2,000
square miles must be below the level of high spring tides, Large marshes,
or jhils (‘‘ eng” in Burmese), are found occupying the depressions between
the raised banks of the principal streams, and the whole region, especial-
ly in the neighbourhood of the sea, consists of a network of the tidal creeks.
Little appears to be known as to the progress of the delta seaward.
judging by the contour of the coast, it would appear that the Irawadi,
owing to its far greater size, and perhaps to the larger proportion of silt
transported by its waters, had pushed its delta seaward far beyond the Sit-
taung. The Salwin traverses for the most part an area of hard metamor-
phic rocks, and probably brings down but little detritus, so that the con-
version of the Gulf of Martaban into land, if it is ever to be effected, must
depend largely upon the deposits from the [rawadi.

The alluvial plain and delta of the lower Irawadi consist mainly of
a clay? very similar to that found in the Gangetic plain, but containing
much less lime, and consequently poor in kankar. The colour is gene-
rally yellowish brown, sometimes reddish, owing to the presence of
peroxide of iron. The proportion of sand varies, and is greater on the
whole than in the Gangetic alluvium. A few thin layers of sand occur
interstratified with the clay, and a band of dark blue or carbonaceous
clay, a few inches in thickness, has been noticed in several localities.

1 Mr. Theobald considers Min-gyi, 13 miles
below Myanaung, the apex of the delta; and
taking Puriam point, east of the Bassein river,
and Elephant point, west of the. Rangoon
river, as the two lateral angles, he estimates
the distances from Min-gyi to Elephant and
Puriam points as 129 and 176 miles, respec-
tively, the two points being 137 miles apart,
—Records, Ill, 21, (1870).

2. Mr. Theobald considers this clay marine
or estuarine, but no fossils have been found in
it and his main arguments, founded on the
similarity between the clays of the Irawadi
and Gangetic deltas, are of course favourable
to the Auviatile origin of the Irawadi clay, if
that of the Ganges be also of fresh water
origin,—Records, III, 17, (1870).
426 GEOLOGY OF INDIA— PLEISTOCENE AND RECENT. C Chap. XVI

The clay, in many places towards the head of the delta, is seen to rest
upon pebbly sand, and the latter is frequently found beneath the clay in the
delta itself, wells being sunk through the argillaceous surface formation
to the porous stratum beneath. In the absence of any borings, however,
it is impossible to say what the nature of the beds at a depth below the
surface may be, and it is not clear whether the sand is the underlying
formation throughout, or whether it is merely intercalated between beds
of clay.

On the surface of the clay, in the immediate neighbourhood of the
river, deposits of silt and sand are found in some places, and resemble
the khkddar deposits of the Ganges vailey. No extensive area, however,
is covered by these sandy beds. They form a narrow belt along the river
channel above the influence of the tide, and occupy a rather larger area
around Pantanau. The deposits of the Sittaung alluvial plain closely
resemble those of the Irawadi.
CHAPTER XVII.

THE INDO-GANGETIC PLAIN.

Area and elevation-—Fluviatile origin of the Gangetic plain—Subrecent marine conditions in
the Indus valley—Character of Indo-Gangetic alluvium—-Fossils in the alluvium—General
features of the Indo-Gangetic plain-—The Brahmaputra valley — The delta of the Ganges
and Brahmapvtra—The plains of upper Bengal and the North-West Provinces — Kalar or
Reh—The Punjab--The lost river of the Indian desert—The lower Indus valley and
delta—The Rann of Cutch—The desert of western RAjpitdna.

The immense alluvial plain of the Ganges, Indus, and Brahmaputra
rivers and their tributaries, the richest and most populous portion of India,
covers an area of about 300,000 square miles, and forms approximately one
fourth of the whole surface of British India, exclusive of Burma. The greater
part of the provinces known as Assam, Bengal (including Behar), the North-
West Provinces, Oudh, the Punjab and Sind, are included in the great plain
which, varying in width from go to nearly 300 miles, entirely separates the
geological region of peninsular India from the Himdlayas to the north, the
Sul4im4n and Kirthar ranges to the west, and the hill regions of Assam,
Tipperah, and Chittagong to the eastward. Owing to the varying extent
to which the surface is raised on the margins of the area by the detritus
brought by rivers from the hills, and the gradation between the finer depo-
sits of the plain and the coarser gravels forming the slope at the base
of the Himalayas, it is difficult to estimate exactly the greatest height of the
plain above the sea. The highest level recorded by the Great Trigono-
metrical Survey between the Ganges and Indus, on the road from Sahd-
ranpur to Ludhidna, is 924 feet,’ and this may be fairly taken as the

! The following elevations of localities in the | BRAHMAPUTRA VALLEY,—contd.

Indo-Gangetic plain will afford some idea of Burdmukh, near Tezpur r . 256
the general height of the surface above the Gauhati . F é a . 163
sea. The figures, except in the case of Raj- Godlpéré_ . ‘ ‘ 3 . 150
mahdl, are taken from the maps and published Gances VALLEY—

sections of the Great Trigonometrical Sur- Bardwin . ; ; : = 0s
vey, with a few additions kindly furnished by R4jmahél_ . : . 68

the Surveyor General, Colonel Walker. At Bewares : : é . 258

all the localities quoted the height is the Allab4b4d . . : y . 319
approximate level of the plain:— Cawnpore . : ; : . 47
BRAHMAPUTRA VALLEY Agra : ‘ ; 7 » 553
Sadiya.. , . . - 440 Delhi . * . . © 715
Dibrugarh. . 7 ° » 348 Meerut ‘ : % é » 739

 

Sibsdgar .. . : : - 319 Sahdranpur : . . + 907
428 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. C Chap. XVIL

summit level at the lowest part of the watershed between the Indus and the
Ganges. There is no ridge of high ground between the Ganges and Indus
drainage, and a very trifling change in the surface might at any time turn
the affluents of one river into the other. It is reasonable to infer that
such changes have taken place in past times, and that the occurrence of
closely allied species of Platanista (a fresh water dolphin peculiar to the
Indus, Ganges, and Brahmaputra) in the two rivers, and of many other
animals common to both streams, may thus be explained.

An idea once prevalent amongst both geologists and naturalists was, that
the great Indian plain had been an arm of the sea in late geological times,!
It is possible that this may have been the case, but there is absolutely no
evidence whatever in favour of such a view, and some facts are opposed to
it. On the southern flank of the Himdlayas, no marine formations have been
discovered of later date than eocene, and even these are unknown, except
in one place east of the Ganges, between the spot where the Jumna leaves
the Himdlayas, and the Gdro hills, or throughout thirteen degrees of
longitude, whilst the later tertiary formations, belonging to the Siwélik
system, contain fresh water Reptilia and Mollusca, and not a single marine
shell has been found in them. In Sind marine beds of miocene date are
found, which become replaced by fresh water beds as they are traced up
the Indus Valley,and in the Salt range the fresh water Siwdliks rest upon the
nummulitic limestone. It is true that it is impossible to tell what beds
may be concealed beneath the Indo-Gangetic alluvium, and marine strata
may exist to an enormous extent without appearing at the surface. It is
also unquestionable that the amount of information hitherto derived from
borings is very small indeed, but so far as that information extends,
and so far as the lower strata of the alluvial plain have been exposed in
the beds of rivers, not a single occurrence of a marine shell has ever
been observed, nor is there such a change in the deposits as would render
it probable that the underlying strata are marine. As will be shown
presently, the lowest deposits known in the plain itself are of postter-
tiary age, and they are certainly fresh water, whilst the tertiary deposits

Inpus VALLEY— 1 Hooker, Himalayan Journals, tst ed.,London,
1854, 1, p. 378; Theobald, Records, III, 19.
Lage : , 2 7 7 eo. Mr. Theobald’s. main pagurdeal
euaben 5 7 2 7 - 806 | derived from the clay at Pattharghatta, near
Heepue 2 ‘ . 7 + 645 Rajmahal, has been shown by a re-examination
babere Le : : : © 798 | ofthe locality to be untenable, the deposit in
Dera: lemnail bas ° : + 595 | question being merely asurface wash, contain.
Multan . . . . + 407 :

ing fragments of bricks amongst other things
ee : : : ? ve (Memoirs, XIII, 224, (1877). Dr. Falconer
ssa . : . ape. considered that the Indu-Gangetic area was

bik
: ; oe Ba ag ie formerly an arm of the sea, but that it had been

en ie : : : * ITO | converted into land before the Siwdlik epoch,
Kotri . . : ° » 66

 

Paleontological Memoirs, 1.29.
erusreey Chap. XVILI FLUVIATILE ORIGIN, 429
are chiefly known to occur on the northern margin of the plain. The
older pliocene deposits of Perim island in the Gulf of Cambay lie, how-
ever, to the south of the alluvial area, and five species of mammals found
in them are also met with in the Siw4liks at the base of the Himdlayas,
so that there was probably land communication between the two areas.
‘The only evidence known in favour of magine conditions having prevailed
during the deposition of any portion of the Gangetic alluvium is the occur-
tence of brine springs’at considerable depths in a few localities. These
springs, however, are not numerous, and without additional evidence it is
impossible to look upon them as proofs of marine deposits. At the same
time it is by no means impossible that the sea occupied portions of Sind
and Bengal long after the plain of upper India wasdry land. With refer-
ence to Bengal there is very little evidence. Mr. Fergusson, in a masterly
essay on recent changes in the delta of the Ganges,’ has brought forward
a quantity of historical data tending to show that the whole Ganges valley
was probably not habitable 5,000 years ago, and that the extension of
human settlements to the eastward from the Punjab has been gradual.
The latter may be conceded, with the reservation that additional evidence
as to the previous want of population is desirable. The Ganges valley
5,000 years ago, like that of the Brdhmaputra valley at the present day, may
have been so swampy as to be ill suited for cultivation, and yet there is
no reason for supposing that the area had recently been covered by the
sea, for the state of the surface may have been due to an amount of
depression sufficient to render the area marshy, but not enough to cause
it to be overflowed by the ocean. That depression has taken place in the
delta is shown by the records of the Fort William (Calcutta) borehole,
to be described presently, but the only known marine beds in the neighbour-
hood of the Ganges delta, those at the foot of the Gdro hills, are of tertiary
age, and probably pliocene.

In the Indus valley some evidence has been obtained of the sea having
occupied part of the areg in posttertiary times.? East of the alluvial
plain of the Indus near Umarkot is a tract of blown sand, the depressions in
which are filled by salt lakes. These lakes are supplied by water trickling
through the soil from large marshes and pools supplied by the flood waters
of the rivers, and it is evident that the depressions amongst the sand
hills are at a lower level than the alluvial plain, and that the salt is

’ Quart. Four. Geol. Soc., X1%, 321, (1863).
There is one ethnological fact which Mr. Fer-

country, but it is at least as probable that the
non-Aryan tribes were indigenous, and that

gusson has not noticed. The population of
Bengal, as any one who has seen much of
Indian races will probably admit, is shown by
colour, physique, and habits of life to contain
a large proportion of the non-Aryan races,
the people of upper India, on the other hand,
having a much larger Aryan element. This
mixed race may have migrated into the

the present Bengali race is due to an admix-
ture of Aryan blood. The point is, whether
Mr. Fergusson has not taken the south-eastern
migration of the more civilised population
amongst uncivilised tribes for the original
peopling of the Gangetic plain,

2 Your. As. Soc. Beng., XLV, pt. ii, 93, (1876)
Records, X, 10, 21, (1877).
430 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. (Chap. XVII.

derived from the soil beneath the sand. To the southward is a great flat
salt tract known as the Rann of Cutch, marshy in parts, dry in others,
throughout the greater part of the year, but covered by water when the
level of the sea is raised by the south-west monsoon blowing into the Gulf
of Cutch and the old mouth of the Indus, and all water which runs off the
land is thus ponded back. The Lini river, which flows into the Rann, is,
except after rain, extremely salt, and salt is largely manutactured from the
salt earth at Pachpadra, close to the Luni, more than 100 miles from the
e lge of the Rann, and nearly 300 from the sea. Both the present condition
of the Rann and tradition point to the area having been covered by the sea
in recent times, and having been filled up by deposits from the streams
running into it, while the occurrence in some of the salt lakes near Umar-
kot, 150 miles from the sea, of an estuarine mollusc Potamides (Pirenella)
/ayard?, common in the salt lagoons and backwaters of the Indian coast,
seems to indicate that these lakes were formerly in communication with the
sca. The enormous quantity of blown sand, also, which covers the Indian
desert, can only be satisfactorily explained by supposing that it was derived
from a former coast line north of the Rann and east of the Indus valley.!

It appears probable that in posttertiary times an arm of the sea ex-
tended up the Indus valley at least as far as the salt lakes now exist, or to
the neighbourhood of Rohri, and probably farther, and also up the Lini
valley to the neighbourhood of Jodhpur, the Rann of Cutch being of course
an inland sea. The country to the westward has been raised by the
deposits of the Indus, and the salt lakes have been isolated by ridges of
blown sand.

It is true that along the western margin of the Indus alluvium later
tertiary (Manchhar) rocks are found, containing remains of mammalia and
precisely resembling the Siwdlik formation, and as there is nevertheless a
probability that the lower Indus valley was an arm of the sea in postterti-
ary times, it may fairly be argued that the sub-Himdlayan Siwdliks are no
proof that the Ganges valley was not an inland sea at the same epoch.
But in the Indus region the representatives of the Siwdliks pass down-
wards into miocene marine beds. In lower Sind the Manchhar formation
itself becomes interstratified with bands containing marine shells, and not
very far to the westward, there is a very thick marine pliocene formation
on the Balichistdn coast, so that there is evidence in abundance of the
sea having occupied portions of the area in later tertiary times, whilst
there is no proof of any such marine conditions in the Ganges plain.

The various deposits of the Indo-Gangetic plain® may be roughly classed

1 A description of this area will be found { the North-Western Provinces, some extracts

at the end of the present chapter. from which were published in Records, Il, 17,

* The authorities for the following account | (1870) ; Mediicott, Sketch of the Geology of

are manuscript reports by Mr. ‘Ihevbald on | the North-Western Provinces, Records, VI, 9,
parts of the alluvial area in Bengal, Bchar, and (1873), and various papers referred to.
Chap. XVID] OLDER AND NEWER ALLUVIUM, 431

under two subdivisions, older and newer; the former consisting of beds
which are undergoing denudation, whilst the latter form the newer
accumulations, the flood and delta deposits now in process of formation.
[It is difficult, if not impossible, to draw any distinct line of separation
between these two subdivisions, unless, as but rarely occurs, they contain
fossils characteristic of their age, but, generally speaking, all the higher
ground is composed of older deposits, whilst the newer alluvium is chiefly
confined to the neighbourhood of the river channels, except in the delta of
the Ganges, and in the Brahmaputra plain. Still, there are large parts, both
of the Indus and Ganges plains, which are flooded every season, and on
these areas newer deposits are formed by the flood waters, Moreover, as
the rivers constantly change their courses, they often sweep away deposits
only a few years, or even a few months old,

The prevailing formation throughout the Indo-Gangetic alluvial area is
some form of clay, more or less sandy. The older deposits generally
contain kankar, the newer deposits do not as a rule, but there are numer-
ous exceptions in both cases. In the Indus valley the alluvial deposits are
much more sandy than in the Ganges valley, and the surface of the ground
is paler in colour, except where marshy conditions prevail. The deposits
of the Brahmaputra valley in Assam are also sandy. In both these valleys
the greater part of the area is cccupied by the newer alluvial deposits
whilst the greater portion of the Ganges plain, except towards the delta
is composed of an older alluvial formation.

The older alluvium is usually composed of massive clay beds of a rather
pale, reddish brown colour, very often yellowish when recently exposed to
the air, with more or less kankar disseminated throughout. In places,
and especially in Bengal and Behar, pisolitic concretions of hydrated iron
peroxide, from the size of a mustard seed to that of a pea, are disseminated
through the clay ; occasionally these nodules attain larger dimensions, some
being found near Dindjpur of the size of pigeons’ eggs. In places
kankar forms compact beds of earthy limestone. Sand, gravels, and con-
glomerates occur, but are, as a rule, subordinate, except on the edges
of the valley, the quantity of sand in the clay decreasing gradually as the
distance from the hills increases. Pebbles are scarce at a greater distance
than from 20 to 30 miles from the hills bordering the plain. Beds of sand-
stone, sufficiently compact for building, have occasionally been found,
but are of rare occurrence. On the whole, there is no great differ-
ence between the alluvial formations of the Indo-Gangetic plain and those
of the Narbada and T4pti, except that the latter are rather darker in
colour, and perhaps less sandy.

The newer alluvial deposits consist of coarse gravels near the hills, and
especially at the base of the Himdlayas, sandy clay and sand along the
course of the rivers, and fine silt consolidating into clay in the delta and
GEOLOGY OF INDIA—INDO-GANGETIC PLAIN.

CChap. XVII.

in the flatter parts of the river plain, In the Ganges delta beds of impure
peat commenly occur. Fresh water shells are of more frequent occur-
rence in the newer forms of alluvium than in the older, the species being
those now living in the rivers and marshes of the country.

432

The only information of importance hitherto procured as to the nature
and depth of the alluvial deposits beneath the surface is derived from
five borings: one, 481 feet deep, at Fort William, Calcutta, within the
delta and close to a tidal river ; the second at Umballa, 7o1 feet deep, at
nearly the highest level of the plain away from the slope of detritus along
the margin; the third, carried to a depth of 464 feet, at Sabzal-kot west
of the Indus, about 21 miles west by north of Rdjanpur and about 400
feet above sea level; the fourth, at Agra, carried toa depth of 481 feet
from a surface level of 553 feet above the sea; and the fifth, and deepest,
having a depth of 1,336 feet from a surface level of about 370 feet, at Luck-
now. All these boreholes were made for the purpose of obtaining water.

The Calcutta borehole is, with the exception of that at Lucknow, the most
important, because it was carried down to a depth of ahout 460 feet below
the mean sea level. The following account of the deposits passed through in
the borehole is taken from the “ Abstract Report of the Proceedings of the
Cemmittee appointed to superintend the Bore Operations in Fort William
from their commencement, December 1835, to their close in April 1840:””—

“ After penetrating theough the surface soil to a depth of about ten
feet, a stratum of stiff blue clay, fifteen feet in thickness, was met with,
Underlaying this was a light coloured sandy clay, which became gradually
darker in colour from the admixture of vegetable matter, till it passed into
a bed of peat, at a distance of about thirty feet from the surface? Beds of
clay and variegated sand intermixed with kankar, mica, and small pebbles,
alternated toa depth of 120 feet, when the sand became loose and almost
semi-fluid in its texture. At 152 feet the quicksand became darker in
colour and coarser in grain, intermixed with red water worn nodules of
hydrated oxide of iron, resembling to a certain extent the laterite of South
India. At 159 feet a stiff clay with yellow veins occurred, altering at 163

' Four, As. Soc. Beng., 1X, 686, (1840). See
also an excellent account by Lieutenant (after-
wards Colonel) R. Baird Smith, Calcutta Four.
Nat. Hist., 1, 324, (1841) and Proc. Geol.
Soc., IV, 4, (1842). From the latter the account
in Lyell’s “ Principles of Geology ” appears to
be chiefly taken, Some additional details will be
found ia the Four. As. Soc. Beng., I, 369, 649,
(1833) ; FV, 235, (1835); V, 374, (1836) ; VI,
234, 321, 498, 897, (1837); VII, 168, 466,
(1838).

2 Eighty feet in the original, but this 1s

almost certainly'a misprint; first, because
Lieutenant Baird Smith mentions in his de-
scription the occurrence of peat between 30
and 50 feet from the surface, whereas from
75 to 120 feet sandy clay is said to occur,
and this agrees with his descriptive catalogue
of the specimens extracted from the borehole,
and with his figured section ; secondly,
because, as will be shown hereafter, a bed of
peat is found everywhere around Calcutta at
a depth of 20 to 30 feet.
Chap. XVILJ CALCUTTA BOREHOLE, 433

feet remarkably in colour and substance, and becoming dark, friable, and
apparently containing much vegetable and ferruginous matter. A fine
sand succeeded at 170 feet, and this gradually beceme coarser and mixed
with fragments of quartz and felspar to a depth of 180 feet. At 196 feet

clay impregnated with iron was passed through, and at 221 feet sand
recurred, containing fragments of limestone with nodules of kankar and

pieces of quartz and felspar; the same stratum continued to 340 feet, and
at 350 feet a fossil bone, conjectured to be the humerus of a dog, was ex-
tracted! At 360 feet a piece of supposed tortoise shell? was found, and
subsequently several pieces of the same substance were obtained. At
372 feet another fossil bone was discovered, but it could not be identified,
from its being torn and broken by the borer. At 392 feet a few pieces of
fine coal, such as are found in the beds of mountain streams, with some
fragments of decayed wood, were picked out of the sand, and at 4oo feet a
piece of limestone was brought up. From 400 to 481 feet fine sand, like
that of the seashore, intermixed largely with shingle composed of fragments
of primary rocks, quartz, felspar, mica, slate, and limestone prevailed,
and in this stratum the bore has been terminated.”

The first and most important observation to be made on the foregoing
facts is that no trace of marine deposits was detected, but on the contrary
there appears every reason for believing that the beds traversed, from top
to bottom of the borehole, had been deposited either by fresh water, or

in the neighbourhood of an estuary, At a depth of 30 feet below the
surface, or about to feet below mean tide level, and again at 382 teet,

beds of peat with wood were found, and in both cases there can be but
little doubt that the deposits prove the existence of ancient land surfaces.
The wood in the upper peat beds was examined by Dr. Wallich and found
to be of two kinds, one of which was recognised as belonging to the sindri
tree (Heritiera littoralzs), which grows in abundance on the muddy flats
of the Ganges delta, the other probably as the root of a climbing plant
resembling Briedelia. Moreover, at considerable depths, bones of terres-
trial mammals and fluviatile reptiles were found, but the only fragments of
shells noticed, at 380 feet, are said to have been of fresh water species.

The next noteworthy circumstance is the occurrence at a depth of 175°
to 185 feet, again at 300 to 325, and again throughout the lower 85 feet
of the borehole, of pebbles in considerable quantities. The pebbles in
the lower portion are especially mentioned as large, and their size is
shown by the circumstance that they impeded the progress of the bore,
and that it was necessary in several cases to break them up before they

1 A ruminant bone, according to Dr. Fale= ) As. Soc. Beng , V1, 234, pl. xviii, (1837); and
oner; Lyell, Principles of Geology, London, ed. | Calcutta Four. Nat. Hist., 1., pl. ix, (1841).
1867, I, p.479. The specimen cannot now be | ? Figured Four. As. Soc. Beng. V1, 321, pl.
found. Figures of this bone are given, Your. | xxi; and Cale, four. Nat, Hist., I, pl. ix.
GEOLOGY OF INDIA—INDO-GANGETIC PLAIN.

(Chap, XVII,

could be extracted, so that it may be fairly inferred that they were at least
two or three inches across (the borehole was six inches in diameter). The
greater part of the pebbles were clearly derived from gneissic rocks, but
some fragments of coal and lignite which were obtained were perhaps
from the Damuda series, though their composition indicates the probability
that they were derived from the tertiary or cretaceous coal seams of the
Assam hills.}

The peat bed, it may here be mentioned, is found in all excavations
around Calcutta, at a depth varying from about 20 to about’30 feet, and
the same stratum appears to extend over a large area in the neighbouring
country.* A peaty layer has been noticed at Port Canning on the Mutla
(Mutlah), 35 miles to the south-east, and at Khulnd, in Jessor, 80 miles
east by north, always at such a depth below the present surface, as to be
some feet beneath the present mean tide level. In many of the cases
noticed, roots of the suzdri tree were found in the peaty stratum. This
tree grows a little above ordinary high water mark in ground liable to
flooding, so that in every instance of the roots occurring below the mean
tide level, there is conclusive evidence of depression. This evidence is
confirmed by the occurrence of pebbles, for it is extremely improbable
that coarse gravel should have been deposited in water 80 fathoms deep,
and large fragments could not have been brought to their present posi-
tion unless the streams, which now traverse the country, had a greater
fall formerly, or unless, which is perhaps more probable, rocky hills existed
which have now been covered up by alluvial deposits. The coarse
gravels and sands which form so considerable a proportion of the beds
traversed can scarcely be deltaic accumulations, and it is therefore
probable that whenthey were formed, the present site of Calcutta was
near the margin of the alluvial plain, and it is quite possible that a portion
of the Bay of Bengal was dry land.®

At Lucknow the deepest of all the boreholes in the Gangetic alluvium
was driven to a depth of 1,336 feet from the surface, or nearly 1,000 feet
below sea level. The beds passed through from top to bottom were of the
same character, alternations of sand and sandy silt, with occasional bands
of kankar, and beyond the mention of coarse sand near the bottom of the
borehole, there are no indications of an approach to the base of the alluvial

434

1 In‘a boring recently sunk at Chanderna-
gore subangular gravel of quartz and felspar
was met with at about 150 feet. The felspar
fragments were extremely abundant and cannot
have travelled any great distance. This in-
dication of the vicinity of rocks exposed at the
surface is interesting in connection with the
hypothesis of the recent origin of the outlet of
the Ganges into the Bay of Bengal,—iufra,

P+ 443.

-

2 Baird Smith, Four. As. Soc. Beng., 1X, 686,
(1840); H.F. Blanford, Four. As. Soc. Beng.,
XXXII, 154, (1864). See also notices of earlier
borings, four. As. Soe. Beng.,11, 369, 649, (1833).

3 But whilst the depression of nearly 500 feet,
probably since tertiary times, is unmistakable
in the neighbourhood of Calcutta, the signs
of elevation within the same epoch in Orissa,
only 100 to 209 miles distant to the south-west,

| are equally distinct.
Chap. XVILJ LUCKNOW BORING. 435

deposits. Like all the other boreholes, it was sunk in searcn of a supply of
artesian water, and apart from its interest in showing the great thickness of
the alluvial deposits, it is important as proving that artesian conditions do
prevail under the Indo-Gangetic plain. After the surface water was shut
out, a water-bearing stratum was struck at 158 feet, whose water stood
at 61 feet from the top of the borehole; at 190 feet another was met with,
and the water stood at 42 feet, at 341 feet the water rose to 24 feet from
the surface, at 750 feet to 13 feet, at 783 feet to g feet, at 975 feet to 2 feet
from the top of the borehole. At ggo feet and 1,040 feet two water bear-
ing strata were struck in which the pressure was less and the water sunk
to 5 feet. from the top of the boring, but at 1,141 feet it again rose to within
18 inches of the top and at 1,189 feet a bed of quicksand was struck from
which water flowed at the rate of 1o gallons per minute over the top of
the casing, itself 24 feet above the mean level of the surrounding plain.!

There is very ‘little of interest in the other three boreholes that have
been sunk in the Gangetic alluvium, except in so far as they bear on the
theory of the origin of the Himalayas, as will be mentioned in the sub-
sequent chapter. *

The Agra borehole, sunk near the southern margin of the alluvium, is the
only one which traversed its whole thickness to the supporting floor of rock.
The total thickness of alluvial deposits passed through was only 481 feet,
composed of sand and sandy clays with some kankar, the uppermost 150
feet being apparently composed to a considerable extent of blown sand, as
opposed to true alluvial deposits.®

Umballa is on the watershed of the Indo-Gangetic plain, between the
Jumna, which flows into the Ganges, and the Sutlej, a tributary of the
Indus. The locality is about 905 feet above the sea, and 20 miles from the
base of the Himdlayas. There is very little of interest in this borehole.
The depth to which it was carried was insufficient to test the thickness
of the alluvial deposits, and it ceased 200 feet above the level of the sea.
No mention is made of any organic remains being found, but their
occurrence could not be anticipated, as they occur but rarely in the alluvial
formations of the Gangetic plain.®

The borehole at Sabzal-ka-kot is only four miles from the base of the
hills, and by far the greater portion of the beds traversed consist of sand
and pebbles, clays being subordinate, although several beds were met with.

The rarity of organic remains, especially in the older alluvial deposits,
has already been referred to, but shells are occasionally found, belonging
to species now inhabiting the rivers and marshes of the country. An
important discovery of mammalian remains was made about 1830 in some.

! For a detailed account and section of this) 121, (1885).

borehole, see Records, XXIII, 261, (180). 3 For detailed accounts, see T. Login, Quart.
2 For detailed section see Records, XVILI.| four. Geot. Soc, XXVIII, 198, (1872).

2F
436 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. [(hap, XVIL

calcareous shoals of the Jumna.!, The bones were chiefly found cemented
together with substances of recent origin, such as fragments of weapons
and boats, into a mass of concrete, chiefly formed of the kankar washed
from the river’s bank, but in two cases the skeleton of an elephant was
found preserved in the clay, In one instance, in which the bones were
clearly i situ, they were found 4} feet above the highest flood mark,
and 80 feet below the summit of the clay cliff formed by the river, and
there appears no reason to doubt that all the specimens found were
originally derived from the clay. The following species have been re-

cognised ? ;—

Semnopithecus, sp. Sus, sp.
Elephas namadicus. Bos (Bubalus’ paleindicas.
Mus, sp. Bos, sp.
Hippopotamus (Tetraprotodon) ' Antilope, sp.
paleindicus. Cervus, sp.

Equus, sp. Fish and crocodile bones.

Three of the species, all that have hitherto been specifically identified,
are found in the Narbada alluvium also, whilst the only genus not now
found wild in India is Hippopotamus ; the species belong, however, to the
same subgenus as the living African animal. The evidence is not
sufficient to justify any decided conclusions, except that the Jumna clays
must have been deposited in the same posttertiary epoch as the Narbadd
alluvium, but so far as the specific identifications go, they tend to indicate
that the Jumna fossils are newer than the Narbadd remains, as the extinct
type Hexaprofodon and the foreign form Bos namadicus have not been
recognised amongst the former. .

Some bones were also found in the Betwd river in Bundelhkand and
the Bugdoti between Mirz4pur and Chandar,’ but they have not been
identified.

Before proceeding further a few words are requisite in explanation
of a word which it will be found necessary to use occasionally in the
following pages and of four Hindi terms applied in the Ganges valley to
particular kinds of alluvial surface which require notice, because they will
be found freely used, and because, with perhaps one exception,’ they have
no precise equivalents in English.

To Anglo-Indians it is quite unnecessary to explain the meaning of
the term kankar, but the explanation may be of sore use to European

'Sergeant E. Dean, four. As, Soc. Beng., IV, 5 Four. As, Soc. Beng., 1V, 571, (1835).

261, (1835). See also Falconer, Quart. Four.| * The exception is khddar, which corre-
Geol, Soc., XXI, 377, (1865); Paleontological | sponds to the English word strvath, The Eng-
Memoirs, II, p. 640. lish term is, however, local ; its exact meaning

2 Several are figured, Four. As. Soc. Beng., Il. | is far from commonly known, and it is only
pl. xxv, (1833); and IV, pl. xxxiii, (1835). used in hilly country.
Uhap. XVI. J FOSSILS IN THE JUMNA ALLUVIUM, 437

students. The original signification of the word is gravel, the term being
applied to any small fragments of rock, whether rounded or not. By
Anglo-Indians, however, the name has been especially used for concre-
tionary carbonate of lime, occurring, usually as nodules, in the alluvial
deposits cf the country, and especially in the older of these formations.
‘The commonest form consists of small nodules of irregular shape, from
half aninch to three or four inches in diameter, composed of tolerably
compact carbonate of lime within and of a mixture of carbonate of lime
and clay without.) The more massive forms are a variety of calcareous
tufa, which sometimes forms thick beds in the alluvium, and frequently
fills cracks in the alluvial deposits, or in older rocks.2 In the beds
of streams immense masses of calcareous tufa are often found forming
the matrix of a conglomerate, of which the pebbles are derived from
the rocks brought down by the stream. There can be no doubt that
the kankar nodules, calcareous beds, and veins are all deposited from water
containing in solution carbonate of lime, derived either from the decomposi-
tion of the debris of older rocks of various kinds, or else from fragments of
limestone and other calcareous formations contained in the alluvium,

Bhibar is the slope of gravel along the foot of the Himdlayas. Com.
pared with the slopes in the dry regions of Central Asia, Tibet, Turkistdn,
Persia, etc., the gravel deposits at the foot of the great Indian ranges
are insignificant, the difference in height between the top and bottom of
the slope nowhere exceeding 1,000 feet. This difference is probably partly
due to the much greater rainfall in India, and to streams being consequently
able to carry away a much larger proportion of the detritus washed from the
surface of the hills, partly also to the circumstance that the recks in the
lower regions of the hills are not subjected to the loosening effects of frost.

Streams issuing from the Himalayan ranges lose a part, or the whole,
of their water by percolation through the gravel in the d4éar region.

1 The following analyses will give a fair idea of the usual composition of nodular kankar:—
(1, Ghdzipur, Prinsep, “Glean, Sci.” I, 278, (1831); 2, 3, 4, RAnigdnj, Dejoux, Re-
cords, VII, 123, (1874) ;--5, Barmuri;--6, Ramnagar ;—7, Sanktoria, all near
Radniganj, Tween, ibid ;—8, 9, Sahdranpur, Thomson, Rurki Treatise on Civil
Engineering, I, p. 115.] ;

I 2 3 4 5 6 7 8
Carbonate of lime . 72° os 72° 56°94 78°5 saan 54 654 663 57°18 77°33
Carbonate of magne-
sia 5 . « 0% ae 3 172 BP ass oes cg Bes as Trace,
Oxide of iron and
alumina - « it Oxide of 9 167° 2 Oszideof ironand 2°77 1'9) (2° 10°32 "73
iron, . alumina,
Water . * . 4 .Clay ov 22" 30° 10°§ Waterand organic 2°7  2°3 4°5
matter,
Silica ‘i c . isz Sand 3? 9°67 7° Insoluble , + 40°6 30°4 27°22 32°50) 13°94,

2 See the account by Captain E. Smith of | Soe. Beng., Il, 622, (1833); also Newbold, Four,
the kankar in the Jumna alluvium, Your. As. | Roy, As. Soc., VIII, 258, (1846).

2F2
438 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. [ Chap, XVIL

“The whole tract in its original condition is covered with high forest, in
which the s4l (Shorea robusta) prevails. At the base of the slope, much
of the water which has percolated the gravel re-issues in the form of
springs; the ground is marshy, and high grass replaces the forest. This
tract is the ¢avdz, a term not unfrequently applied to the whole forest-clad
slope at the base of the Himdlayas, known also as morung in Nepal.

|

Fig. 22.—Diagram illustrating the relations of Bhébar and Tardi.

The alluvial plain itself, in the North-West Provinces especially, is com-
posed of dhdugar or high land, the flat of older alluvium, now at a consid-
erable elevation above the rivers which traverse it, and £hddar or low
land, the low plain through which each river flows. The latter has
evidently been cut out from the former by the streams, It is of variable
width, and is annually flooded. ‘

In the upper provinces the high banks of the rivers are frequently
capped by the hills of blown sand, known in the North-West Provinces
as bhir. This is the extreme form of a rather important element
in the formation of Indian river channels, and the same result in a less
marked form may be traced in a rather sandy, raised bank, along the
course of many large rivers down to the limits of tidal action in the
deltas. In the lower parts of the river plains this bank, which is above the
flood level and is usually selected for village sites, intervenes between
the river channel proper and the marshy ground liable to annual floods
on each side, the communication between the two latter being kept up by
numerous creeks. The origin of the S4%r land, or raised bank, is the
following. During many months of the year, and especially in the hot
season, strong winds arise, frequently of a very local character, and some-
times apparently almost confined to the river channels. ‘I hese, in the dry
season, are plains of loose sand, often two or three miles across and some-
times wider, of which the river usually occupies not more than a fourth.
The wind on the Indus and Ganges frequently blows in nearly the same
direction as the river channel. Such winds are especially prevalent about
midday and in the afternoon, and their effect in transporting the sands of
the river bed is so great that the atmosphere becomes too thick for
objects a few yards distant to be seen. All who have been in the habit of
navigating Indian rivers must have noticed the prevalence of these sand
storms. They are so marked that, where large sand banks exist to windward
of the river, it is often impracticable for vessels to continue their
course, except in the morning before the wind arises, or in the
evening, when the motion of the air has diminished. Much of the sand

 
   

Terai we
SAN RII.

       
Chap. XVII.] ASSAM VALLEY. 439

raised by the wind falls again in the bed of the river, but quantities must
fall upon the banks in the immediate neighbourhood, where the deposit is
retained by vegetation and gradually consolidated into a firm bank. It
is only where the quantity of sand is greater that blown sand hills are
formed, The original raising of the river bank to the flood level is due to
the deposition of silt in a manner which will be explained presently when
treating of deltaic accumulations, but the elevation of the immediate
neihbourhood of the river bed above the reach of the highest floods is
due to the deposit of sand by the wind.

To enter at length into the various peculiarities of land surface! which
are found in different parts of the great plain of northern India would be
far beyond the scope of the present work. A brief account of the principal
characters must suffice, The whole region may be roughly divided into
five great tracts, each possessing marked peculiarities. These are, com-
mencing to the eastward :—

1. The Brahmaputra valley in Assam.
2. The delta of the Ganges and Brahmaputra.
3. The plains of upper Bengal and the North-West Provinces.

4. The Punjab.
5. The lower Indus valley and delta.

The Assam valley is a gigantic £hdédaz, or strath, the greater portion
being liable to flooding and consequently not in a habitable state. ‘There,
are, however, higher tracts here and there, sometimes mere mounds,
rising a little above the general level, and sometimes small plains,? which
may be considered as representing the extensive bhdugar of the Gangetic
plain. Along the foot of the hills are gravel deposits, but they do not
appear to be very extensive.

The quantity of silt carried down by the Brahmaputra is very great,
far greater than in the Ganges. The comparative backwardness of the
river valley, as shown by the small amount of habitable land, is sur-
prising, since it is evident that the river is occupied in rapidly raising its
plain by deposits of silt, and the necessary inference is that the alluvial

1 The following papers may be consulted for | don, 1860, I, pt. ii; Falconer, Quart. Four.

fuller accounts : Geol. Soc., X XI, 372, (1865); Login, Quart.
For Assam-- Memoirs, lV, 437, (1865); VII,| Four. Geol. Soc. XXVIII, 186, (1872); H. B.
155, (1869). Medlicott, Records, XIV, 205, (1881); XVI,

For Lower Bengal and the delta--Mr. Fer- | 205) (1883) ; XVII, 112, (1885).
gusson’s paper, Quart. Your. Geol. Soc., XIX, For the Punjab—Sketch of the Geology,
gat, (1863); also Colebrooke, As. Res., VII, | Punjab Gazetteer, Provincial volume, 1889,
1, (1801) ; and Rennell, Pxil. Trans., LX XI, 87, | p. 22.
(1781). 2 Memvirs, 1V, 438, (1865). One of these
For the piains of Upper India~-Records, V1, | plains is described by Major Godwin-Austen,
9, (1873); Sir P. Cautley, Ganges Canal, Lon- | four, 4s, Soc. Beng., XLIV, pt. ii, 40, (1875).
440 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN, [ Chap. XVIL

plain of Assam, in its present form, is not only of later date than the
Gangetic plain, but absolutely newer than many portions of the Ganges
delta. The difference may be due to a depression of the lower part of
the Brahmaputra valley in Assam, to an clevation of the delta, or to a
great increase in the supply of water. The second theory is distinctly
disproved by the yeneral evidence of subsidence in the delta, and the
third is improbable; the evidence is therefore rather in favour of the
Brahmaputra valley in Assam having been an area of subsidence in a
relatively late geological period. As will be shown presently, there is
some additional evidence in favour of this view within the delta itself.

The limits of the delta, or the places where the rivers first bifurcate and
commence to give off disturbances, are between Rajmahdl and Murshid4bdd
on the Ganges, and on the Brdhmaputra opposite the south-west corner of
the Gdro hills. But for a considerable distance above the actual delta the
rivers flow through a broad plain of low ground, a large area of which is
liable to flooding, and consequently to the deposition of silt. The delta is,
in fact, the natural continuation of the £A4ddar, or alluvial flat in the upper
portion of the river’s course, and this £kddar becomes broader before it
expands into the delta.

By far the best description of the Ganges delta, of the changes it is
undergoing, and of the action of the rivers in raising the land by the
deposition of silt, is that of Mr. Fergusson.2?, He has shown that rivers
oscillate in curves, the extent of which is directly proportional to the
quantity of water flowing down the channel. Thus, the oscillations of the
Ganges where broadest (7,000 feet in the low season) between Monghyr and
Rajmahdl, average g miles in length; where it contains less water, and is
only half the breadth (3,500 feet), between Allaha4bd4d and Chanér, the
oscillations are 3°7 miles long ; in the Bhdgfrath{, where it averages 1,200 feet
in breadth, the length of the oscillations is 1°5 miles ; and in the MAatdbhdngé,
where only 500 feet broad, the length of each oscillation becomes only half a
mile.’ The next point which he notices is well known, the tendency of
rivers to raise their banks, but the explanation is partly novel. When the
whole country is covered with water, moving rapidly towards the sea in
the river channels, and stationary throughout the intervening marshes, the
dead water of the marshes prevents the floods of the rivers from break-
ing out of the channels, and, by stopping the course of the silt charged
water along the edges of the creeks and streams, forces it to deposit the

Bi For a full discussion see Fergusson, Quart. | hati to be only about 100 feet above the
Four. Geol. Soc., X1X, 330, (1863). It should, | sea, instead of 163, the maximum flood-level

however, be noticed that Mr. Fergusson was led | since determined by the Great Trigonome-
by some published barometrical observations, | trical Survey.

now shown to have been insufficient, to suppose | # Quart. Four. Geol. Soc., XIX, 321+54, (1863).
the level of the Brahmaputra valley at Gau- | § Quart. Four, Geol. Svc., XIX, 324 (1863).
Chap. XVII. J GANGETIC DELTA, 44t

sediment it has in suspension. Hence gradually arises a system of river
channels, travetsing the country in many directions, between banks which
are higher than the. intervening flats, and these flats form persistent
marshes, known in the Ganges delta as jhils or bhils.

Each river frequently changes its precise course, the smallest alteration
in its channel having an effect which is felt for many miles above and
below. So that, just as the oscillations of a denuding stream produce
a low alluvial flat between high banks, the curves of a depositing river
gradually form a high alluvial flat, raised above the surrounding country.
In course of time this raised tract is abandoned by the main river for the
lower ground at the side, and the river bed is either filled up by silt, or,
if near the sea, converted into a tidal creek.

The present Bengal delta, therefore, comprises a large area in which
the ground has been raised above the general flood level, through having
been traversed by the main branches of the Ganges in past times. Such
is the case in the country north of Calcutta, The eastern part of the
delta is more backward, the marshes, or j47/s, are more extensive, and the
banks of the streams less consolidated, and this is now the great deposit-
ing area, But large tracts of low country, suchas the salt lake near
Calcutta, are found in the western area also, The remarkable struggle
which takes placebetween the Ganges and Brahmaputra, each tending,
by raising the neighbourhood of its channel, to drive back the other, and
to gain possession of a larger tract of delta, is most vividly told by
Mr. Fergusson, but is too long for extract here. Mr. Fergusson refers
many of the more modern changes in the delta to the upheaval of the ele-
vated tract known as the Madhupur jungle, which had the effect of divert-
ing the Brahmaputra to the eastward into the Sylhet 4z/s, where the silt of
the river was deposited. The result was that scarcely any sediment found
its way to the sea by the Meghn4, the great estuary of all the Sylhet rivers,
and hence the sea face of the delta tothe eastward curves back in the form
of a gulf. The gap was much greater at the commencement of the present
century, but about that time the Brahmaputra having, by the deposit of silt,
greatly raised the portion of the Sylhet jiz/s into which it flowed
changed its course completely in the course of a few years, and instead
of flowing to the east of the Mddhupur jungle, cut out a new channel to
the west of the raised tract. Since its change, of course, the Brahmaputra
has been brought much nearer to the main stream of the Ganges, and
the two rivers are now depositing silt so rapidly on the eastern sea face
of the delta, that great changes are taking place, and new islands are
rapidly forming, whilst the western portion of the deltaic coast line,
through which but a small portion of the flood water of the great rivers
finds its way to the sea, has undergone but little change since it was
first surveyed in the last century.

1 The former term is Hindi, the latter Bengali.
442 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN, CChap. XVII

In the sea outside the middle of the delta there is a singularly deep
area, known and marked on charts as the “Swatch of no ground,” in
which the soundings, which are from 5 to 1o fathoms all around, change
almost suddenly to 200 and even 300 fathoms. This remarkable depres-
sion runs north and south, and has been referred to a local sinking, but
it appears more probable, as has been shown by Mr. Fergusson, that the
sediment is carried away from the spot, and deposition prevented, by the
strong currents engendered by a meeting of the tides from the east and
west coasts of the Bay of Bengal. Mr. Fergusson also shows that, so
long as the Bay of Bengal has preserved its present form,! the meeting of
the tides must have favoured the formation of a spit of sand along the
present position of the Sundarbans, as the lower portion of the Ganges
delta is called, and that any great deposit of silt to seaward of the
present line is impeded by the fine sediment being washed away by the
tidal currents, and deposited in the deeper parts of the Bay.

In spite of all that has been written on this subject the origin of the
swatch of no ground has by no means been cleared up. A very similar
depression has been shown to exist in the bed of the shallow sea off the
Indus delta and the cause in both cases has probably been the same, a
combination of an excess of subsidence with a deficiency of sedimentation,
the latter due to the action of surface currents in sweeping away the silt-
laden waters. It is not in accordance with what we know to suppose that
at such depths as we are there dealing with, there can be any currents of
sufficient velocity to account for the depression by actual erosion.

The chief point in the above theory to which exception might be taken
is the question of whether the elevation of the Madhhupur jungle is suffi-
ciently recent to account for the changes in the course of the Brahmaputra
river, This tract of country is composed of a red, iron stained, clayey soil,
in which accummulations of pisolitic concretions of oxide of iron are found
and worked in places as an iron ore. The clay is of the same type pre-
cisely as the older alluvium of the Brahmaputra and lower Ganges valleys.
There seems little room for doubt that this is really a region of special ele-
vation, for were we to suppose that it is part of the old surface of the delta,
left standing at its former level while the surrounding area was depressed,
the height of its ground level, which rises to 100 feet above the general
level of the delta outside the Madhupur jungle, would necessitate a much
greater extension of the delta into the Bay of Bengal than there seems
any ground for supposing to have ever been the case. The steeply scarped
western face and the gentle fall to the level of the delta on the east show

! This is probably not so old as pliocene, | since the close of the SiwAlik epoch, that
because such gigantic disturbance has taken | the shape of the northern part of the Bay of
place throrghout the Assam hills and Avakan, | Bengal must have changed greatly.
AGE OF THE GANGETIC DELTA.

Chap, XVIII

that the elevated area must once have been more extensive than it now is
and that the western half has been washed away by the rivers that im-
pinged upon it, while the deeply eroded undulating nature of its surface
shows that it has been raised above the flood level of the rivers and so
subject to denudation for too long a period to make it probable that
the diversion of the Brahmaputra river to the east was entirely due its
elevation.

Whether the elevation of the MAdhupur jungle was anterior to, or con-
temporaneous with, the depression, which it is difficult to suppose has not
taken place in the area occupied by the Sylhet jhils, it is. impossible to
say, but the latter would in itself have been sufficient to account for the
diversion of the Brahmaputra to the east of the Mddhupur jungle, and was
probably its principal determining cause.

An interesting point to determine in this connection is the date of
origin of the delta of the Ganges. Reference has already been made to
the fact that the plateau of the Assam range forms structurally part of
the Peninsula of india, and to the presence of rocks of peninsular type
north of the Brahmaputra valley. Moreover, the upper tertiary deposits
south of the Assam range differ from those at the foot of the Himdlayas,
so far as the latter are known, in being partly of marine origin, and as
will be shown further on, the formation of the depression occupied by
the Gangetic alluvium was most probably an integral part of the oper-
ations which resulted in the elevation of the Himalayas. In the same
way it is probable that the transverse depression, through which the Gan-
getic drainage now finds its way to the sea, may have been formed at the
close of the tertiary period parz passu with the elevation of the Tipperah
hills,

There is some direct evidence in favour of the more recent origin of
the Gangetic outlet in the presence of closely allied species of dolphins
in the Ganges and Indus rivers, of a very different generic type from the
cetacean inhabiting the Irawadi. These two species must be descended
from a common ancestor which acquired a fresh water habitat, and the
‘differentiation of the Indus and Gangetic species have arisen from a sub-
sequent separation of the drainage areas.! The changes in the course of
the drainage over what is now the watershed region, which will be referred

443

1 The occurrence of allied farms of porpoise | gained ‘access in communication with the

or dolphin in the Ganges and Indus, and the
circumstance that the peculiar genus living in
these rivers is unknown elsewhere (the ceta-
cean inhabiting the Irawadi being of a very
different generic type) have attracted the
attention of naturalists already. The ova and
young of fish are not difficult of transport, and
a very trifling accident might place a pool of
water to which tle fish of one river have

other stream. Crocodiles and river tortoises
can live for a long time out of water, and have
considerable powers of migration on land,
but dolphins are confined to the rivers, and
could neither live in a shallow. pool, nor
traverse dry land. The existence, therefore,
of closely allied species, doubtless derived
from a common ancestor, in two distinct
rivers, is avery striking fact. Mr. Murray
GEOLOGY OF INDIA—INDO-GANGETIC PLAIN,

(Chap. XVII.

to further on, though they opened water communication between the Indus
and Ganges rivers, probably did so only in the torrential region, which is
not frequented by the dolphins, and the difference existing between the
two species indicates a more prolonged separation than could have been
the case had there been migration from one drainage area to the other,
when they were put into communication with each other by the wanderings of
the rivers near the present limits of the two drainage areas. We are conse-
quently driven to suppose, either that the two closely allied species originated
independently of each other, which is extremely improbable to say the least,
or that the great bulk of the Himalayan drainage once found its way to the
sea by a single delta, instead of two, and this must have been either at the
head of the Arabian sea, or of the Bay of Bengal. The indications of the
sea having extended up thé Indus valley within the recent period, and the
absence of any similar indications in the delta of the Ganges, make it
probable that the former was the original outlet of the drainage, and that
the formation of the gap between the Rdjmahal and Garo hills, and of
the Gangetic delta, is geologically of recent date.

444

On the western edge of the delta in Bengal there is a large area of
older alluvium, whose surface is slightly undulating, evidently in conse-
quence of denudation. This tract, which is continuous with the alluvial
area of the east coast, comprises the greater portion of the country to the
westward of the Bhdgirdthi and Highli, and probably owes its comparative
elevation to the deposits fromthe Mor, Adjai, and Damodar rivers,

The great plain of Northern India is the area of an alluvial deposit older:
than that of the delta, and the greater portion of the area is composed of
bhéngar land, through which the rivers cut their £hddar valleys at depths
of from 50 to 200 feet below the general level. The b4déngar surface, as
a rule, is nearly flat, but is much cut up by ravines in the neighbourhood
of the rivers,

The question as to whether the great rivers are on the whole, raising their

(Geographical Distribution of Mammals,
p. 213) proposed an ingenious hypothesis to
account for the phenomenon. He considered
that the plain of upper India was once an
arm of the sea, that it was cut off by the rise
of the coast in Sind and Cutch, and gradually
converted into a brackish, and then a fresh
water lake, discharging itself by the Ganges,
that meantime the marine dolphins inhabiting
the sea had gradually become adapted to the
changed conditions, and had in fact become
Plataniste, He then supposes that the
Ganges was cut off from the lake, which over-
flowed again, and this time into the Arabian

 

| Sea, the dolphins of the Ganges and Indus

being specialised during the change. It would
be unnecessary to refer to this hypothesis,which
of course is little more than a suggestion, but
for the large amount of support the idea has
received from naturalists. It is of course
foreign to the purpose of the present work to
discuss the genesis of Platanista, but, as will
be shown, the geological phenomena of. the
Indo-Gangetic plain do not bear out Mr.
Murray’s hypothesis, which, it should be stated:
was never proposed as a geological theory, but
merely as illustrative of the possible mode of
origin of allied species. *
Chap. XVILJ DISTINCTION OF BHANGAR AND KHADAR. 445

beds by a deposit of silt, or cutting their channels deeper, has been much
discussed without leading to any definite conclusions. The abrupt scarps
by which the dkdnzgar is not unfrequently terminated, and the defined
limits of the 4Addar, clearly prove that the latter has been at some time: or
other an area of denudation, but it is not easy to tell whether, at the pre-
sent time in any given stream, the tendency is to raise or lower the
general khddar level. It is also by no means so evident, as might at
first sight be supposed, whether the 4hdngar land generally is an area of
denudation or of deposition, although this can, as a rule, be easily seen
in each particular area. Thus the minor hill streams from the lower ranges
of the Himalayas between the Sutlej and Jumna must deposit sediment, for
they cease within the area, whilst betweenthe Jumna and the Ganges nu-
merous streams rise inthe d4dugar, and they must be denuding agents.
In the neighbourhood of the khddar, bhdngar land is frequently cut into
by ravines, which prove conclusively that the surface of the country is being
washed away, but all such marks of rain action cease at no great distance
from the low ground, and the principal secondary streams, instead of run-
ning from the upland dhdngar by the nearest route, at right angles, or
nearly at right angles to the main river, usually pursue a nearly parallel
course down the middle of each doéé; or triangular area between two
principal streams.

As the velocity of the rivers where they leave the hills is much greater
than in the alluvial plains, there must, so long as diminution takes place in
velocity of the water when the river is carrying as much earthy matter as
it can transport, be a continuous deposition of detritus, and a gradual
raising ofthe area flooded by the stream. This is the case even in the larger
rivers which carry a considerable body of water at all times, while the
effect of the small streams, which dry up more or less for a great’portion
of the year, but are converted into muddy torrents charged with coarse
sediment during the heavy rains of the summer monsoon, is necessarily to
raise the surface of each dod, especially in the neighbourhood of the
hills, and to produce floods from which finer sediment is deposited on the
surface of the dzdngary land. Whether the addition thus produced is, on
the whole, greater than the wasting of the surface from rain is a question
which it is impossible to decide throughout a great part of the country.

One question, which presents itself, is the necessity of accounting
for the rivers now cutting their channels at a level considerably below
that of the alluvial dh/ngar flat, because this flat must, at all events
in the neighbourhood of the khddar, have been deposited by streams
from the same drainage area, at a period when, the main river ran
at a comparatively higher level. The change may be due to a general

1 A Persian word, meaning ‘ two waters,’ and { well as to the land intervening between them.
applied to the confluence of two rivers, as
446, GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. (Chap. XVIL
elevation of the upper Gangetic plain, or toa depression in the deltaic
region. Of the former there is no evidence, of the latter, as shown by
the result of the Calcutta borehole, there is ample proof, and it is there-
fore quite possible that in early posttertiary times, when the animals
lived, whose remains are found in the Jumna alluvium, the area of the
Ganges delta had been raised to a considerably higher level than it occu-
pies at the present time. Colonel Greenwood has shown! that the deposit
of silt in river valleys must take place backward, that the lowest portion
of the slope must be first raised, and that the check thus given to the flow of.
water will cause silt to be deposited, so as toraise the alluvial plain further
up the course of the river, and if no change of level takes place, the gra-
dual elevation of the Ganges delta by silt deposit will ultimately react on
the higher portions of the valley until the rivers once more deposit allu-
vium on the high d4éugar land, provided always that this has not been
raised so muchas to render the slope too great for the rivers to be
depositing agents.

One point of interest has been explained by Mr. Fergusson in the
paper so often mentioned. A glance at the map will show that the
Ganges from Allahabad to Rajmahdl, and the Jumna from Delhi to Alla-
hdbdd, flow close to the southern margin of the great alluvial plain. This
is due to the enormous quantity of silt brought down by the Himdlayan
rivers, and the comparatively small supply furnished by those streams
which debouch into the Ganges valley from the southward. The northern
portion of the plain has consequently been raised, and the main drainage
of the whole forced to find its way as close to the hills of the southern
margin as it can. During’ this process the courses of the tributary rivers
running from the northward have been driven westward, and the con-
fluence of these tributaries with the main stream of the Ganges has been
shifted upwards along the course of the main river, owing to the tendency
of the streams to deposit silt in the neighbourhood of the delta.

The dhdébar slope of gravel along the foot of the Himdlayas, although
evidently of comparatively recent formation, has frequently, to the east-
ward, been cut into terraces by the streams from the hills.? This is a
necessary consequence of the streams cutting deeper channels in the rocks
of the hilly ground. It is curious to note, however, that to the westward
the dhdbar is being raised instead of being cut through by streams. It is

' Rain and Rivers, 2nd ed., London, 1866,
pp. 173, etc.

? Hooker, Himalayan Journals, rst ed., Lon-
don, 1854, I, p. 378 (larger edition). Dr. Hooker
very naturally, writing forty years ago, when
the study of river action was in its infancy, and
when nearly all great deposits and all extensive
denudations were supposed to be marine, attri-

buted the gravel to a beach deposit, and the
valleys to marine denudation, There has
been since a great revolution in those por-
tions of geological dynamics which treat of
the action of rivers and the sea, and especially
in the views held, by English geologists at
least, on the comparative amount of work
done by the two agents,
Chap. XVII. ] REH OR KALAR. 447

not known how far this difference is due to the greater rainfall to the
eastward, and to the streams being consequently able to carry away the
gravel as they cut down their bed in the rock, whereas weaker streams afe
prevented from cutting back their channels by their inability to wash
away the gravel they have already deposited. There have doubtless
been alterations of the gradients of the stream beds through recent
upheaval or depression of the surface, and that these would have as much
influence on the present action of the streams where they cross the dhabar
zone as the rainfall.

In connection with the surface of the upper provinces another peculiar
local feature requires explanation. Many tracts of land in the Indo-
Gangetic alluvial plain are rendered worthless for cultivation by an
efflorescence of salt, known in the North-West Provinces as reh,
and further west as a/ar (kullar), The name dsav, meaning barren, is
frequently applied to land thus affected. The salt varies in composition ;
it consists chiefly of sulphate of soda mixed with more or less common
salt and carbonate of soda; it is only found ‘in the drier parts of the
country, being unknown in damper regions, such as Bengal,

The zsar plains have existed for an unknown time. Where the reh
or kadar is abundant, the water in the upper stratum is impregnated to
an extent that is productive of serious injury to the health of the popula-
tion. To a greater or less extent this pollution of the water near the
surface is general throughout Upper India, yet sweet water is obtainable,
in the worst veh tracts, at depths below 60 to 80 feet.

It is consequently clear that the impregnation of the soil is superficial,
and as the upper deposits are demonstrably of fresh water formation, they
must originally have been comparatively free from impurities. Still all soils
contain some sait, and all the water draining from soils is impregnated to
a certain extent. The salts forming veA or kalay appear to be the: refuse
products, and to consist of such substances, resulting from the various pro-
cesses involved in the decomposition of rock, or of detritus derived from
rock, and the formation of soil, as are not assimilated by plants. Unless
these salts are removed they must accumulate, and the natural process of
removal is evidently by rain water, percolating through the soil and carry-
ing off any injurious excess of the rejected salts. Ifthe amount of water be
sufficient, and through drainage exists, there will be a constant dilution and
renewal of the subsoil water, but if the water reaching the subsoil can only
be dissipated by evaporation during the dry season, salts will accumulate in
such subsoil water, and as this water is brought to the surface by capillary
action, and evaporated, the salts held in solution will be left- as an
efflorescence on the surface of the ground,

That the composition of veh does not differ greatly from that of the
448 GECLOGY OF INDIA—INDO-GANGETIC PLAIN. [Chap XVII

salts produced by the decomposition of such rocks as have contributed by
their disintegration to the formation of the alluvial plains of Indlia, is shown
by the composition of the river water! running from the Himdlayas, the
mountains from which the detritus, now forming the plains of India, was
originally derived.

In the case of Upper India it is easy to understand how the destruc.
tion of the conditions necessary for cultivation has been established,
and it is by no means improbable that a similar process has, in other
parts of the world, changed countries, once fertile and populous, iuto
barren deserts. The whole country is treeless. For a great part of the
year a scorching sun and a parching wind dry up the moisture in the
ground, rendering it hard and impervious to water. When the rains of the
monsoon season fall, a large proportion of the water runs off the surface,
and the earth is unable to absorb more than a portion of what remains.
Thusa great part is evapcrated without penetrating the ground. The little
that does percolate through cracks, and in a zig-zag way, through the
more porous layers to the upper water stratum, is no more than sufficient
to replace what has been dissipated by evaporation, fed by capillary
action,

This more or less complete want of water circulation in the subsoil
must have been gradually producing its effects in Upper India throughout
many generations. The natural process is so slow that it would escape
notice were it not that from time to time larger tracts of land become
barren. A disturbmg cause has, however, been introduced in the form
of great irrigation canals, Their immediate effect is to raise the level
of the veh polluted subsoil water, and thus to produce a great increase of
evaporation, with the natural result of more veh being left on the sur-
face, and more land being thrown out of cultivation. It is impossible
to enter at length into the subject here, but it may be stated that, as all
canal water contains salts in solution, whilst rain water contains none, the
only change in conditions, so far as the concentration of salts in the soil is
concerned, by. the addition of canal irrigation, unless facilities for drainage
of the subsoil water are also provided, must be the addition of all the
refuse salts contained in the canal water to those which would be pro-
duced on the surface by the simple action of rain and evaporation,

South and west of Delhi and west of Agra, brine is obtained in places
from wells in the alluvium. No particulars have been recorded which

In several analyses of river and canal water | in veh. See Sel. Ree. Govt,, India, D. P. Wy
from the Ganges and Jumna, the proportion | No, XLII, p. 47, (1864), An able and de-
of sulphate at soda varied from Orog14 tO] tailed account of the origin, composition and
0°4325 part in 10,0003 chloride of sodium | mode of concentration of the reh salts by Mr.

from 0'0023 to o'15 part. The proportion of | W. Center, m.s., will be found in Records,
the two to each other is similar to that found | XIII, 253—273, (1886).
Chap. XVILJ CAUSE OF REH, 449

explain the occurrence of salt in these localities. The case is similar
to that already mentioned in the Purna valley in Berar. The distribu.
tion of the salt producing ground appears irregular, and this is in favour
of the salt being derived from springs in the rock beneath the alluvium.

The plains intersected by the five great rivers which combine to form
the lower Indus are not, asa rule, simply divided into dhdéngar and khddar
like the plains of the North-West Provinces. Owing probably to the
greater fall in the Punjab rivers, their deposits are very sandy, and this
character tends to diminish the pluvial denudation of the surface by allow-
ing the water to sink into the soil. The action of winds upon the sand
of the river, the formation of bhzr land, and the elevation of the ground in
the neighbourhood of the river banks above the intervening tracts, through
the deposition of blown sand, are exhibited in the Punjab to a greater
extent than in the Gangetic plain.

To the south-east the limits of the Punjab alluvium are difficult to
trace, owing to the manner in which both alluvium and rock are concealed
by blown sand, The same is the case throughout the eastern margin of
the Indus alluvial plain in Sind.

The ancient geography of the Punjab is far better known than that of
most parts of India, partly because the civilisation of north-western India
is older than that of other parts of the country, but still more because of
the accurate descriptions given by Greek writers of the Indian campaigns
of Alexander the Great. If is consequently possible to form some idea of
the principal alterations which have taken place in the course of the last
2,000 years, in the channels of the great Punjab rivers, but our best guide un-
fortunately fails us at the most critical point. Alexander never penetrated
to the eastward beyond the land of the five rivers, and there is but little
except vague tradition to tell whether the present tributaries of the Indus
have ever flowed into the Ganges, or those of the Ganges into the
Indus. Yet it is certain that in no part of the great Indo-Gangetic plain
have more important changes taken place since the dawn of history than
in the neighbourhood of the watershed between the Indus and Ganges.

An inspection of the map accompanying this chapter will show a
dried up river channel, which can be traced from the neighbourhood of
Sirs4 into connection with the eastern Narra in Sind, and local tradition
states that this was formerly occupied by a flowing river. At present this
charnel is dry, except in its upper part, where it periodically carries, for
a greater or less distance, the flood waters of the minor streams which
drain the outer Himalayas between the Sutlej and the Jumna. The origin
of the channel is situated at the junction of the alluvial fans of the Sutlej
and Jumna, as is shown by the course of the minor drainage channels, and
there are abandoned river courses leading from it in the direction of the
450 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. [ Chap, XVII.

debouchures of these two rivers from the hills. There can be no room for
doubt that, within the period .known geologically as recent, this river
channel carried a flowing stream to the sea, and there is some evidence,
apart from oral tradition, to show that its drying up took place within the
historical period.

The Muhammadan historians of the eleventh and twelfth centuries
uniformly speak of the combined Sutlej] and Bids rivers, now known as
the Garrah or Sutlej, as the Biyah, a nomenclature which is also employed
in the Hindu annals of Jaisalmer. This retention of the name of the
smaller of the two rivers for the combined waters, where there is no
superior sanctity to recommend it, shows that the rivers must have received
their actual names at a period when the Sutlej did not join the Bids,
but pursued an independent course, and the subsequent abandonment of
the illogical nomenclature may be held to show that the alteration of the
course of the Sutlej] which took it into the Bids did not take place much
before the eleventh century. Previous to this change it doubtless flowed
down what is now the dry river bed known as the Hdkra or Wandan, and
there is some evidence, though far from conclusive, that it followed this
course as late as the eleventh century of our era.

The traditions of the Hindus point to a time when a large and sacred
river, known as the Saraswati and described as ‘chief and purest of rivers
flowing from the mountains to the sea,’ pursued its course through the
eastern Punjab. The modern Saraswati is an insignificant stream fed by the
drainage of the outer hills alone, becoming nearly dry in the hot season and
losing itself in the sands of the Rdjputdna desert. It is absurd to suppose
that the language of the Vedas could have been applied, or that any con-
ceivable alteration of the rainfall could have made it applicable, to the
Saraswati of the present day, and the most reasonable explanation is,
as suggested by Mr. Fergusson, that the Saraswati was in fact the Jumna,
which, in the Vedic period, pursued a ee course to the sea, probably
down the dry river channel just referred to.* It is certainly a suggestive
fact in this connection, that when the Brdhmaputra changed its course
through Bengal about the commencement of the present century and flowed
west of the Madhupur jungle to join the Ganges, the new channel was named
Jamuna, a word etymologically identical with Jumna. On similar principles
the old Saraswati, when it broke eastwards to join the Ganges, may have
assumed the name Jamuna or Jumna for its new course, and if this ex-
planation be correct, the Hindu legend that the Saraswatf joins the
Ganges at Praydg or Allahdbdd, is unwittingly a true statement of fact.

This bringing of the change in the course of the Jumna river, which

. ' See an interesting but anonymous article | C. F. Oldham; also four, As. Soc. Beng, LV,
in the Calcutta Review, Vol. LiX, pp. 1 29, | pt. ii, 322-43, (1887).
(1874) understood to be by Surgeon-Major | 2 Quart. Four, Geol. Soc., XVX, 348, (1869).
Chap. XVIL ] THE LOST RIVER OF THE INDIAN DESERT. 45

has indubitably taken place, down to so recent a date is interesting, for the
change must have occurred previous to the present distinction of £Addar and
bhéngar, and if this distinction has been produced since the Aryan invasion
the question naturally arises whether it may be due, not to movements of
elevation or depression, but to the clearing of the land from forest, and the
extension cf cultivation in the plains, and more especially in the hills,
which, by allowing the rain to flow more quickly off the surface, would in-
crease the erosive power of the rivers when in flood, and cause them to cut
down their channels into the plains over which they formerly flowed.

The surface of the Indus alluvium in upper Sind differs but little from
that of the Punjab, a considerable portion of the area is annually flooded,
and the whole drainage of a great river being here, as in Assam, confined
to a comparatively narrow tract, some permanent marshes of large size exist,
The two most important marshy tracts are along the western edge of the
valley from near Jacobébdd to the Manchhar lake near Sehwan, and along the
eastern edge from Khairpur to below Umarkot. The latter is the channel
considered by some the ancient course of the Sutlej. In the neighbour-
hood of the Indus the ground is rather higher, having evidently been
raised by the deposit of silt, aided doubtless by-the action of the wind on
the sands of the river bed.

Along the edge of the Kirthar range, west of Sind, there is a well
marked d4é4ar slope of gravel, but the breadth seldom exceeds one to
two miles except where rivers run out of the range. This gravel slope
is absolutely barren, and, like other features in Sind geology, is more
conspicuous on account of its barrenness.

There is one singular feature in the Indus valley to which nothing
parallel is to be found in the Gangetic area. The river between Sukkur
and Rohri has cut its way through a low range of limestone hills, surround-
ed on all sides by alluvial deposits. The eastern Narra, fed by the flood
waters of the Indus, traverses an alluvial tract eastward of the hills. In
fact, the circumstance that the flood waters of the Indus, both to the east
and west, traverse plains at a lower level than the river bed, is shown
by the course of the canals, and great fears have been entertained that
the Indus may desert its present channel and break out to the westward,
through the plain in which Jacobdbdd is built, into the line of marshes
already mentioned. The curious features of the tract are not even con-
fined to the present river course, for at Aror, four miles south-east of Rohri,
there is another gap in the limestone range, said, on what is believed to
be good historical evidence, to have been the bed of the river rather more
than nine centuries ago.! At that time the main stream is supposed to have

1 Cunningham, Ancient Geography of India, London, 1871, I, pp. 257, 264, etc.
2G
452 GEOLOGY OF INDIA—INDO-GANGEIC PLAIN, [Qhap, XVIL

traversed Sind considerably to the east of its present course ; it passed by the
old city of B:Ahmandbdd, and then probably ran southward by the Purdn, an
old river bed still existing, to the Kori creek, which was the principal mouth
of the river, The Indus is said to have deserted its old bed at Aror for
its present channel between Sukkur and Rohri, in consequence of an
earthquake about A D. 962, and as Brdhmandbéd was also, in all probability
destroyed by an earthquakel at some period prior to A.D. 1020, it is not im-
possible that the two events were due to the same cause. The Indus is said
to have deserted Brdhmandbdd at the time when the city was destroyed. All
the details preserved, however, are so much mixed up with mythical inci-
dents that but little dependence can be placed upon them, and nearly all
the circumstances mentioned are more or less open to dispute, It is
questioned, for instance, whether Aror was ever situated on tke Indus,
and it is contended that Bukkur, a fortress on an island in the river oppo-
site Rohri and consequently in the channel now cut through the limestone
range, existed before the ninth century. Certainly, the channel through
the hills at Aror is very narrow, and it is possible that it was never tra-
versed by the main stream of the river, though the configuration of the
ground supports the hypothesis that some stream has cut through the hills
at the spot. Again, it is contended that Sehwan, the ancient Sindoména,
was always on the Indus, and that consequently the main stream of the
river must have run in ancient times where it flows now. But, on the other
hand, Alexander is said to have left the river, and marched to the neighbour-
hood of Larkhdna, and thence to Sehwan, from which place he “ marched
back to the river.”? It may be fairly concluded that important changes
have taken place in the course of the river, without feeling certain that
the precise nature of these changes has becn correctly ascertained.

The accumulation of fluviatile deposits in the Indus plain, and the
consequent elevation of the surface, is well seen in the neighbourhood of
Umarkot, where, as has already been mentioned, the flood-water from the
Narra trickles through the sand hills forming the limit of the Indus allu-
vium, and fills large hollows between the ridges of sand. The level of
the bottom of these hollows must have been, in all probability, at least as
high as the general surface of the Indus plain at no distant date,

During the floods, water leaves the Indus, and its tributary the Sutlej,
as far up as Bahawalpur, and flows southward by the eastern Narra, which
must be regarded as a distributary, although its waters now seldom reach
the sea. The true head of the delta, however, is generally considered to

1 Bellasis, Four. Bombay Br. Roy. As. Soc. | while he places ¢ e re 4
V, 413, 467, (1853). ‘There is some doubt | seven mnileg to CN aA el nc
regarding the exact position of Brdhmandbdd. | known as Depar. our. Ro As. Soe now
According to Generai Haig the ruins usually | series, XVI, 285 (1884) ee
known by that name are those of Mansurah, | # Arrian: “ Anabasis,” VI, 16
Chap. XVIL.1 DELTA OF THE INDUS. 453

be a little above Haidarébad, where the Phuleli stream leaves the river.}
The channels of the delta frequently change, more frequently perhaps
than in the case of the Ganges. The sea face is, in all probability, deter-
mined by marine currents, and it is improbable that any great change is
likely to take place through the deposit of sediment.

The eastern part of the Indus delta now receives but little water from
the river. It is said that a large area of country in the neighbourhood of
the Kori mouth was depressed during the earthquake of 1819,? and that
the great size oi: the Kori creek is due to the depression. A very large
area north-west of the Kori creek is covered with salt, sometimes a foot
or even more in thickness, deposited from sea water.

In the neighbourhood of the sea the soil is usually argillaceous and
firm, but in the upper part of the delta the whole surface is composed
of loose micaceous sand with but little clay, and the rivers consequently
have urusual facilities for changing their channels. The littoral portion
of the deltais so low that a broad tract of country is always overflowed at
spring tides, whilst the bottom of the sea in the neighbourhood of the cuast
is so shallow, and the slope outwards so gradual, that large vessels cannot,
in many places, come within sight of the land, A tract of country of variable
width, but in places several miles -road, along the sea face of the della, is
annually flooded by the rise of the river, the water being kept higher than
it would otherwise be by the influence of the south-west monsoon.

Reference has already been made to the Rann of Cutch, and it was
pointed out that this tract of country is evidently an old marine gulf now
silted- up. A brief description of the area and its peculiarities may,
however, be well added to the accoun: of the Indus delta, which it adjoins
to the eastward.

The Rann? consists of an immens2 marshy salt. plain, scarcely above the
sea level and stretching for 200 miles from east to west, and in places

1 A very good description of the Indus delta | buted to the earthquake. It would be well
has been given by Lieutenant T.G. Carless, | however, that the spot should be examined by
Four. Roy. Geog. Soc., VII, 328, (1838), re- | an experienced geologist, as the vagaries of
printed in Sel. Rec. Bombay Govt. XVII, 461- | false bedding (or oblique lamination) in
500, (1855). See also a memoir by Assistant | sands and silts deposited by the strong
Surgeon J. F. Heddle, (bid, p. 403). For the | currents of an estuary, are very likely to
ancient changes in the delta of the Indus see { mislead any one unaccustomed to the peculiar

also Cunningham, Ancient Geography of | appearance of these depnsits.

India, p. 283, etc. ® For a fuller description of the portion north

2 It is stated by Carless, Four. Roy. Geog. | of Cutch by Mr. Wynne see Memoirs, IX:
Soc., VIII, 366, (1838), that the alluvial form- | 14, (1872). See also Burnes, Travels in
ations exposed on the bank of the Kori creek | Bokhara 2nd ed., London, 1835, I, p. 316;
opposite Kotasir are, with the exception of | Grant, Geol. Trans., 2nd series, V, 318,
the uppermost layers, broken up in confused | (1840); Frere, Your. Roy. Geog. Soc, XL,
masses, and inclined to the horizon at an angle | 181, (1870); Rogers, Quart. Four. Geol. Soc
of 30 or 4odegrees. The disturbance isattri- | X XVI, 118, (1870).

° 2G 2

 
GEOLOGY OF INDIA—INDO-GANGETIC PLAIN.

(Chap. XVII.

nearly 100 from north to south, From the south-eastern extremity a low
alluvial tract, dividing Ahmaddbdd from Kathidwar, and including an ex-
tensive brackish water marsh called the Nal, connects the Rann with the
head of the Gulf of Cambay. A very trifling depression, probably not
amounting to 50 feet, would convert Kathidwér into an island, and even a
smaller amount of sinking would suffice to isolate Cutch completely ;
indeed, it is now an island during the prevalence of the south-west mon-
scon, when the sea, raised by the wind, dams back the water brought into
the Rann by the various rivers which drain into the flat from Rajputdna,
Gujardt, and Cutch, in the same manneras the level of the creeks is raised
in the Indus delta. At this time portions of the Rann are seven feet under
water, but the average depth does not exceed five feet. The inundation
lasts from July to the end of November, and portions of the surface, espe-
cially a tract to the westward near Sindri, depressed by the earthquake of
181g, are constantly covered with water. Below this water there is, in
places, a bed of salt, sometimes as much as three or four feet in thickness,
There can be little doubt that the Rann was a gulf of tne sea within
recent times. Not only do the traditions of the country all agree with
this view,' but the present condition of the surface, an immense flat of
sandy mud, can only be explained by supposing that the tract is the site of
an inlet, now silted up. The barren condition of the surface is due to
flooding by salt water at one season, and hot dry weather at other times;
the soil is consequently too salt to support even the vegetation, such
as mangroves, which will grow in ordinary sea water. Unless further
depression takes place, the surface must be gradually raised by the silt
brought in by rivers, and the tracts which support vegetation must extend.

454

The depression of an area of 2,000 square miles around the fort of Sindri
in the western part of the Rann, at the time of great earthquake of 1819, has
been described so often,’ that it appears unnecessary to repeat the account
here. In this case the circumstance which enabled the changes of level
to be accurately estimated was the fact that the whole of the tract affected
was very nearly at the sea level, and so close to the sea that it was

2 There is some historical evidence also. ; No mention of any sea north of Cutch appears

When Alexander the Great sailed down the
Indus he passed through the great eastern
branch, then the main stream of the river, but
now dry, to the Kori mouth. Near this mouth
he came to a great lake (Arrian: ‘* Anabasis,”
VI, 20). Mention is also made of a great lake-
like expanse of water in this direction by some
Mahomedan historians. Sir B. Frere also
states on apparently good traditional evidence,
that Virdwah, in Nagar Pdrkar, north-east of
the Rann, was a seaport from 500 to 800 years
ago, Four, Roy. Geol. Soc. XL, 195, (1876)

 

to have been made by the Chinese travellers
of the seventh century; Cunningham, Ancient
Geography of India, IJ, p. 302.

1 An account is given in Lyell’s Principles,
ed. 1868, II, pp. 97-104, and has been copied
into many text-books. For a very full descrip-
tion by Mr. Wynne see Memoirs, IX, 29-47;
(1872). Mr. Wynne doubts whether the Allah
Bund was really raised, and suggests, with
much probability, that the appearance of ele-
vation was due to the depression of the ground
around Sindri, south of the Allah Bund.
Chap. XVIL. J 455

flooded immediately. A further depression is said to have taken place
in 1845 in the same neighbourhood.!

At first the effect of the depression in 1819 was to produce a great sheet
of water, navigable by boats of some size, but this has gradually silted
up, and Mr. Wynne, on visiting the ruins of Sindri in January 1869, found
that the greater portion had been filled up to nearly the level of the Rann,
and that but a small shallow pool remained around Sindri itself.

THE RANN OF CUTCII,

Though not, strictly speaking, part of the Indo-Gangetic alluvial plain,
this will be the best place to notice that yreat accumulation of blown sand,
in the tract between the Indus and the Ardvallis, which is known as the
Indian desert. The name implies a greater degree of barrenness and soli-
tude than is actually the case. Shrubs and grass tufts are scattered thinly
over nearly the whole area, small trees are not infrequently met with, and it
supports large numbers of sheep and cattle, and a hardy population, civilised
enough to build cities and palaces and wells of hundreds of feet in depth.

Over the whole of this area sand hills are scattered more or less thickly,
but the great accumalation of blown sand forms a strip along the north of
the Rann of Cutch, from which two arms run, one northwards by Umarkot
and then turning north-east and running north of Jaisalmer to Bikaner ;
the other running north-eastwards between Balmer and Jodhpur and co-
alescing with the first about Bikaner. The central area of Jaisalmer,
Balmer and Pokaran is rocky, with comparatively few and scattered sand
hills. ,

The sand hills are of two types. One of these, admirably delineated
in the Trigonometrical Survey maps, is of the ordinary type of sand
dune. Its longer axis is at right angles to the prevailing direction of
the wind, and it presents.a long gently sloping face to windwards, up
which the sand grains are driven, and a steep face to leeward, down
which they roll, whose slope coincides with the angle of repose of the dry
sand.

The other type is one which is not noticed in the text-books. It is very
largely developed in the Thar district of Sind to the north of the Rann of
Cutch, and appears, equally with the first type, to owe it sform to the

'Nelson, Quart. Four. Geol. Soc., I, 103,
(1846), Before quitting the subject of the great
alluvial region of Northern India, it may be as
well to point out that by far the greater portion
of the earthquakes, and especially of the more
severe shocks felt in India, occur in the immedi-
ate neighbourhood of the Indo-Gangetic plain,
and especially near the deltas of the great rivers,
The earthquakes are, as a rule, felt much more

severely on the rocky ground around the allu-
vial plain, than in the plain itself. When
depression takes place, as in the case of Sindri
in the Rann, the shock may be but slightly
felt at the locality principally affected, although
towns in Cutch, onrocky ground, at a distance
of several miles, are thrown down; but this
is in accordance with experience elsewhere.
456 GEOLOGY OF ItNDIA-—INDO-GANGETIC PLAIN CChap. XVIL

prevailing winds, Throughout the area mentioned, the sand is heaped into

 

 

 

 

 

$ 4 3 2 r Oo Miles.
Fig. 23.— Sand hills of the transverse type; after the Topographical Survey of Rajputana.

long narrow ridges, running about north-east and south-west to north
north-east and south south-west, with a steep slope on either side, the
crest gradually rising in height to the north-eastern extremity, which is

 

 

 

 

ea, itt ao ye

2 4 3 2 { O Biles,

Fig. 24.—Sand hills of the longitudinal type ; alter the Sind Revenue Survey.
usually the highest point of the ridge, and beyond this is a steep slope
downwards, coinciding with the angle of repose of the sand.

The mode of formation of this type of sand hill is not very easy to
Chap, XVIL J SAND HILLS OF THE INDIAN DESERT. 457

explain, but from the fact that, where the two types are found together, the
longer axis of the one is at right angles to that of the other, we may con-
clude that in both cases the form is decided by the direction and force of
the wind, the longer axis being parallel in one case, and in the other trans-
verse, to its prevailing direction. The steep slope of repose at the northern
end of these sand hiils shows that they are formed of sand grains which
are driven along the surface of the ground by the wind, and not of those
light enough to be carried in suspension, so that no theory of accumulation
under the lee of bushes, will account for the facts, and we must look else-
where for an explanation, If one of the transverse type of sand hills
be examined, it will be seen that the windward slope is by no means a
uniform plane, but is composed of long narrow ridges, parallel to the
direction of tle wind, with intervening depressions, probably kept open
by a concentration of the wind in them and a consequent increase of
transporting power, if not an actual development of power of erosion. It
seems probable that the longitudinal type of sand hill is due to an ex-
aggeration of this effect, by which the depressions, instead of being com-
paratively shallow, and causing mere saddles in the general ridge, are
carried almost, if not quite, to the base of the accumulation. However this
may be, the restriction of the longitudinal type of sand hill to the seaward
and western margins of the desert appears to show that they are connected
with a greater wind force than the transverse type.

The height of these sand hills is considerable, They frequently exceed
100 feet, ranging to 200 feet, and, according to Sir Bartle Frere, 400 to 500

_ feet in the southern part of the desert. The size of these sand hills and
the area they cover imply an accumulation of blown sand which it is not
easy to account for.

It appears difficult to believe that all the sand found in the desert can
have been derived from the Indus. The surface of the Rann at present is
too muddy to furnish any large supply. The sand consists of well rounded
quartz grains, mixed with smaller quantities of felsparand hornblende, and
is uadistinguishable from the sand of the sea coast except that the grains are
better rounded, as is always the case with wind blown sand. That found
in the bed of the Indus is also very similar in character. The most probable
theory appears to be that the Rann of Cutch and the lower portion of the
Indus valley were, as has already been shown to be probable on other
grounds, occupied by the sea in posttertiary times, and that the sand of
the desert was derived from its shore. ‘Ihe most sandy tracts are on the
edge of the [ndus valley, along the northern margin of the Rann, and along
the depression of the Luni valley, and these portions of the country were
all probably situated on the coast. The form of the rocky hills around
Rdlmer and Jaisalmer shows that they have been shaped by subaerial,
not by marine denudation, and it is probable that the more elevated central
458 GEOLOGY OF INDIA—INDO-GANGETIC PLAIN. [ Chap, XVIL

portion of the desert was land, whilst the Indus valley, the Rann, and the
Lini valley were occupied by sea.

The accumulation of sand in the desert region is evidently due to the
low rainfall and to the consequent absence of streams, the effect being
intensified by the accumulation of sand and the porous nature of the re-
sulting surface. In other parts of India the sand blown from river chan-
nels or the sea coast is either driven by the wind into other river channels,
or it is swept into them again by rain. There are sand hills in abundance
in the alluvial plain of the Indus, but they attain no great size, because
the sand is always swept sooner or later into some stream, by which it is
carried away towards the sea.

Besides the occasional sand hills of the Indus valley in Sind, there are
some much larger tracts in the Punjab, repeating, on a smaller scale, the
phenomena of the Thar and the Rajputdna desert. The most important
of these is in the Sind-S4égar Dodb between the Indus and Jehlam, but
there is a barren tract in the RachnaDodb between the Chendb and Ravi,
and sand hills occur in places also in the Bari Doab between the Ravi and
Sutle}.
 

 

 

 

 

 

 

 

 

 

 

 

 

THE INDO-GANGETIC ALLUVIUM

Scale 1 Inch =128 Miles or aitaw

100

NOTE —The shaded portion represents rock.

Boundaries of the rock area in Rajputana, Nepal, and North of the
Brahmaputra are approximate.

 

 

 

 

oRHOTAK Ja
>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Keg. No. 140, Dir., Geol. Survey.—Sept, 92.—1,500.
PHOTO-LITHOGRAPHED FROM AN ORIGINAL MAP SUPPLIED BY THE GEOLOGICAL SURVEY OF INDIA,

at the Survey of India Offices, Calcutta, Oct. 189°.
CHAPTER XVIII.

THE AGE AND ORIGIN OF THE HIMALAYAS.

Geographical limitation of the Himdlayas— Physical geography—Evidence of the tertiary
deposits as to the age and elevation of the Himdlayas—Revd. O. Fisher’s theory of moun-
tain formation—Age of the Himalayas.

In dealing with the great question of the date of commencement, cause,
and history of the elevation of the Himdlayas, the first point to determine
is the exact signification in which this name shall be used. The Himdlayas
in a general sense are well understood to be the great system of mountains
which rises to the north of the alluvial plains of upper India, and forms the
southern margin of the highlands of Tibet, but the limits of the range at
either end are difficult to define, for it becomes continuous with the moun-
tain ranges between India and China on the one hand and those north of
Afghanistan on the other, and though it is easy to regard these as distinct
ranges, once the change of general direction is well established, the ab-
solute continuity of each with the Himalayas, where the junction takes place,
shows that the elevation of the whole was part of the same great series of
movements of the earth’s crust. It is, however, necessary to adopt some
definite geographical limits to the Himélayas, and those used here will be the
lines along which the strike of the chains of hills, and of the rocks they are
composed of, takes a sudden bend. On the west this line may be taken to
run through the hills west of the valley of Kashmir, from where the Kara-
koram range bens into the Hindu Kush to where the Jehlam leaves the
hills. On the east neither the geology nor the geography of the hills is suf-
ficiently well known to define the limit of the Himdlayas, but it may be
presumed to run in from the neighbourhood of Sadiyd in a north-easterly
direction.

The descriptions of the orography of this great system of mountains vary
very much according to the idiosyncracy of the writer, and the particular
meaning he may have attached to tlie term ‘mountain chain,’ The earliest
of the writers on the Himdlayan mountains, Captain Herbert, regarded the
spur on which Simla is situated as the natural termination of the main
snowy range, giving as his reason that it was the watershed between the
drainage of the Indus and the Ganges. But this system of classification
would lead to endless confusion and completely obscure the true relations
460 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [ Chap. XVII.

of the various chains of which the Himdlaya mountains are composed, for
the rivers, without exception, cut at some point of their course througn
a zone of special elevation, whether we look to present contour or to geo.
logical structure, and it is impossible on any rational ground to separate the

portion of the range on one side of the river valley from its continuatioa
on the other.

The most popular of the views regarding the physical geography of the
Himdlayas is probably that proposed by Clements Markham, which regards
them as consisting of three more or less parallel ranges known as the
northern, central, and southern, respectively. This view was most beauti-
fully illustrated in the map attached to his edition of the travels of Bogle
and Manning, where three long ranges are depicted, stretching across
the map from east to west. The view appears to have a certain resem-
blance to the truth, and cannot be absolutely disproved owing to our igno-
rance of the geography of the greater part of the Himalayas and to the in-
definiteness of the term mountain range, but our knowledge of Himdlayan
geography is sufficient to show that the orography of the Himdlayas is by
no means so simple or well defined as it is represented on the map just re-
ferred to.

In the north-western portion of the Himalayas, where alone the geogra-
phy is known with any degree of completeness, four principal ranges are
commonly recognised. The most northerly and most elevated of these,
which appears to bend round into the Hindu Kush at its north-westerly
extremity, is the Mustagh or Karakoram range, whose culminating peak,
28,265 feet high and the second highest in the world, was formerly known
as K,, but is now often named after its discoverer Godwin-Austen.

South of, and more or less parallel with, this comes the Laddkh range,
which may be regarded as commencing near the junction of the Shdyak
(Shyok) and Indus rivers and running thence south-eastwards along the north
side of the Indus valley. This range, which has a most marked indivi-
duality both geographically and geologically, is breached by the Indys at
about 150 miles from its north-westerly termination in 79° of east longitude.
The range continues as far as Hanle, forming there the south, instead
of the north, side of the Indus valley, but its further continuation is imper-
fectly known.

The Zanskar range appears to owe its existence quite as much to the
accident that it forms the watershed between the Indus and Chenab drain-
age, and has consequently been less denuded than the regions on the north
and south, as to any special elevation it has undergone.

The outermost of the principal ranges is that which is known as the

1“Narrative of the Mission of George | Manning to Lhasa,” 2nd eliiion, Loador, 1879.

Bogle to Thibet and of the Journey of Thomas
Chap. XVIII. } OROGRAPHY OF THE HIMALAYAS. 461

Pir Panjal south of the valley of Kashmir, and as the Dhdoladh4r' south of
Chamba. Though the unity of these two ranges is obscured by their being
broken through by the Chendb and Ravi rivers, their geological structure,
so far as it is known, seems to show that they are in reality part cf one and
the same range. Much might, however, be said in favour of the view which
would regard them as two separate ranges, écheloned along the northern
margin of the plains of the Punjab.

Nothing definite can be said of the south-easterly continuation of the
ranges. The Pir Panjal, Dhdoladhdr and Zanskar ranges may be regard-
ed as coalescing and becoming continuous with the great range of snowy
peaks, while the Lad4kh and Karakoram ranges coalesce to continue as
the range of mountains which runs north of the great longitudinal valley,
of the upper Indus, Sutlej, and Sanpo rivers. It may, however, well be
doubted whether either of these ranges has a real continuity along the
whole length of the Himdlayas, and it is altogether more probable that,
whether we regard them structurally or according to the accidents of the
existing contour of the ground, they consist of a series of comparatively
short ranges overlapping each other at their extremities. The final
classification of the minor ranges of the great Himdlayan system of moun.
tains must wait for a more detailed geological and geographical knowledge
than is at present available.

Though it is impossible to give any definite idea of the detailed oro-
graphy of the Himdlayas it is possible to divide the mountains into oro-
graphical regions sufficiently distinct from each other, even if their exact
boundaries are somewhat indefinite. The innermost of these is the up-
land of Tibet, characterised by great elevation and a dry climate with
its concomitant of very extensive accumulations of detritus in the
valleys. :

The drainage of the southern portion of this region, except that of the
comparatively small area which is drained by the upper Sutlej river, escapes
into the Brahmaputra and Indus valleys at the extremities of the Himalayan
range, But by far the greater portion has aclosed drainage, or such as
escapes finds its way into extra-Indian rivers.

South of the Tibetan region rises the great zone of snowy peaks whose
drainage, from both northern and southern slopes, finds its way, in a more
or less directly-transverse direction, on to the Indo-Gangetic plain. The
watershed of this transverse drainage lies to the north of the zone of highest
peaks, which is repeatedly interrupted by the deep valleys of the rivers
traversing it.

South of the snowy peaks comes a zone of lower hills, seldom rising

* On the accompanying map the name | range which runs south-eastwards from Dal!-
Dhdoladhdér has been misplaced. It is the | housie, south of the Chamba valley.
462 GEOLOGY OF INDIA—AGE AND OKIGIN OF HIMALAYAS. CObap. XVIII.

much over 12,000 feet above the sea, which has been distinguished as
the lower Himdlayas. These lower Himalayas in many places graduate
into the main snowy range, so that it is difficult to draw a definite distinc-
ton between the two; yet, they can be recognised as a fairly well marked
feature of the range, forming a belt of hills some 50 or 60 miles broad,
between the high mountains of the central range and the low hills of
the sub-Himdlayas. West of the Sutlej the lower Himdlayas cannot be
recognised as a distinct feature, the high ranges of the Dhdoladhdr and
Pir Panjal rising directly from the sub-Himdlayan zone, but the inner por-
tion of what has generally been regarded as the sub-Himdlayan zone in
this region rises to greater altitudes than where the lower Himalayas are
typically developed, and should possibly be regarded as the continuation
of this feature.

The sub-Himdlayas, which have been referred to in the last paragraph,
form the outermost zone of the hills. They are usually marked by an abrupt
drop in the average height of the hills, they are exclusively composed of
tertiary and principally upper tertiary deposits, and except in the region
west of the Sutlej, seldom rise over 4,000 feet.

There can be no doubt that this sudden drop in the average height of the
peaks, between the lower and sub-Himdlayan region, is principally due to the
sub-Himdlayan region having been subjectedto a smaller amount of elevatory
movement than the lower Himalayas, though it is doubtless also due in part
to the greater softness of the rocks they are composed of, and their greater
proximity to the lowlands of the plains, but it is not so easy to determine
whether the distinction between the central ranges and the lower Himalayas
is due principally to differences in the amount of upheaval they have
undergone or to denudation. Doubtless both have co-operated. The
bottoms of the river valleys near the plains being at a lower level than fur-
ther into the heart of the mountains, and the average slopes at which the
hillsides stand, which depend on the readiness of the rock to disintegrate
and the amount and distribution of the rainfall, being probably less on the
average in the lower than in the central Himdlayas, the peaks could naturally
not rise to the same altitude. This does not, however, seem to be a sufficient
explanation of the facts, and it is only natural to suppose that the belt of
mountains which contains the highest peaks in the world must have been an
area of special upheaval, while there are some features in the profile of the
main river valleys which support this conclusion, These valleys all penetrate
the hills to within 10 miles of the line of highest peaks without rising more
than 4,000 to 5.000 feet above sea level, but as they cross this line there is a
sudden rise of the river bed which carries it up.to 9,000 to 10,000 feet within
a few miles. Above this the gradient falls again and, in’ the Tibetan
region, the average slope dues not seem to be more than a few feet in each
mile of channel. This sudden rise in the river beds as they cross the line of
Chap. XVIII] ORIGIN OF THE TRANSVERSE VALLEYS, 463

highest peaks seems to show that this has been a region of greater and
more rapid upheaval than those to the north or south, and that the rivers

have not yet been able to cut down to the level they will ultimately
reach,

It has already been mentioned that all the principal rivers draining
from the Himdldyas have their sources to the north of the line of highest
peaks, and that they cross this zone of special upheaval in deep valleys.
The old explanation of this feature was that the valleys were great frac-
tures in the range, through which the rivers found their way. This view has
never been specifically disproved in the case of the Himalayas, but it has
been so frequently shown to be incorrect in other cases where it was main-
tained, it is so generally discredited, and moreover the shape of the valleys
is so palpably due to subaerial erosion that it is unnecessary to devote
further attention to it here, and we may accept these transverse valleys as
having been entirely produced by the action of rain and rivers.

It will be shown further on that the sub-Him4ldyan ranges are composed
of the disturbed and upheaved deposits laid down by the same rivers which
now traverse them. In this case it is evident that the rivers are older
than the hills they traverse, and that the gorges have been gradually cut
through the hills as they were slowly upheaved. In the same manner it
might be supposed that the rivers, originally draining from the north of
what is now the line of greatest elevation, were able to keep their valleys
open by cutting them down, at a pace sufficient to prevent the upheaval
producing an actual reversal of drainage, but in many cases the drainage
area to the north of the line of highest peaks appears to be too small to have
given sufficient erosive power to the stream to allow of this explanation.
In the first edition of this Manual it was suggested ! that these transverse
river valleys may have formerly extended further to the north, draining .
a larger area of country beyond the snowy range than they now do, and
that owing to the greater depth of the gorges of the Indus, Sutlej, and Di-
hing or Sanpo, their upper waters were cut off by a gradual encroachment
of the longitudinal valleys of these rivers on the transverse drainage. The
present writer is unable to accept this view. He regards it as altogether
more probable that the first effect of the commencement of the upheaval
of the Himalayas, was to establish a pair of longitudinal valleys along its
northern face, whose drainage escaped round the extremities of the upheaval,
and that in the first instance the whole of the drainage north of what is
now the line of highest peaks escaped by these rivers. As the mountains
were upheaved the gradients of the rivers flowing directly to their southern
margin became steeper than those of the longitudinal valleys north of the
main range, the erosive power of the streams increased, and they were able
to cut back through the line of maximum upheaval and rob part of the

1 p. 677.
4064 GEOLOGY OF INDIA~—AGE AND ORIGIN OF HIMALAYAS. [Chap. XVIII.

drainage which originally flowed east and west to the gorges of the Indus,
Sutlej, and Sanpo

The few geological investigations which have been made along the
southern margin of the Tibetan highlands have not been sufficiently de-
tailed as regards the distribution of the recent deposits and forms of the
valleys, to decide this question with certainty, but there is one specific
observation, recorded by General Strachey,! which points to the conclusion
that the explanation given here isthe correct one. He records that the
subrecent deposits of the Sutlej valley in Hundes extend right up to the
crest of the Niti pass, and that a detached portion of it is to be seen two or
three miles south of the crest. The mere fact of its extending up to the
crest of the pass shows that there must originally have been higher ground
to the south; in other words, that the original watershed of the Sutle] must
have run further south than it now does, and the occurrence of an outlier
in what is now the southern drainage area, if confirmed, gives a still
further, though unnecessary, proof of the encroachment of the southern on
the northern drainage areas.

The same is indicated by the shape of the valleys which drain in either
direction from the watershed. So far as can be gathered from the admirable
maps of northera Kumdun and Garhwédl, and from the accounts of travel-
lers, the slopes on the southern side of the passes are much steeper than
on the northern; the erosion of these slopes would consequently be more
rapid, and as it progressed the watershed would gradually be forced north-
wards.

The most conclusive evidence, however, seems to be that derived
form the subrecent deposits of the Sutlej valley in Hundes. These show
that the Sutle] was followed approximately its present course during a
period sufficient, firstly, for the formation of a deep rock valley, secondly,
for the accumulation in this of over 3,000 feet of subrecent deposits,
and, thirdly, for the re-excavation of gorges, 3,000 feet deep, through
these same accumulations, There can cons2quently have been no pro-
gressive cutting back of the head waters of the Sutlej during all this
period.

From a stratigraphical point of view the Himdlayan mountains may be
divided into three zones, which correspond more or less with the orographical
ones. The first of these is the Tibetan, in which marine fossiliferous
rocks are largely developed, whose present distribution and limits are toa
great extent due to the disturbance and denudation they have undergone.
Except near the north-western extremity of the range they are not known
to occur south of the snowy peaks. The second is the zone of the snowy

1 Four. Roy. Geog. Soc., XXI, 63, (1851).
Chap. XVIII.) STRATIGRAPHICAL ZONES 465
peaks and lower Himalayas, composed mainly of crystalline and metamor-
phic rocks and of unfossiliferous sedimentary beds, believed to be prin-
cipally of palazozoic age. The third is the zone of the sub-Himdlayas, com-
posed entirely of tertiary, and principally of upper tertiary deposits, which
forms the margin of the hills towards the Indo-Gangetic plains, and has
so intimate a connection with, and so important a bearing on, the history
of the elevation of the Himalayas that it will require a more detailed. notice
here than the others.

The stratigraphy and palzorto‘ogy of the rocks composing this terti-
ary fringe have been referred to in a previous chapter, but it will be neces-
sary to recapitulate part of what has been written, and to add some further
details which are important from the present point of view.

The classification which will be adopted is the following :—

+

Upper Siwdlik.
Upper tertiary pper Siwahk

Reais : Middle Siwdlik.
or Siwalik series

Lower or Nahan Siwdlik.
Kasauli group a ;
\ Dagshdi group Murrce beds.
Usubéthu group.

Lower tertiary
or Sirmur series

The lowest of these groups consists everywhere of marine deposits,
clays, shales with some limestone, and a few bands of sandstone. It passes
upwards with perfect conformity into a series of interbedded sand:tones
and clays. The latter, almost always red in colour, prevailing in the lower
part, the former in the upper, so that there is a gradual increase in the
average coarseness of the débris from below upwards, a feature even more
conspicuously displayed in the sections of the upper tertiaries.

The distribution of these rock groups is noteworthy. There is a long
narrow outlier in western Garhwdl just east of the Ganges, in which only
the marine Subdthu beds are found, A larger area is found further west
in the Simla hills, where all three groups are represented. For a part of
its length this exposure is in direct contact with the Siw4lik series along
the great fault, which will be referred to further on, but along its western
half it is separated by a narrow strip of pretertiary slates. At the
western extremity of this outcrop of lower tertiary rocks, which belong by
position to the lower Himdlayas rather than the sub. Himdlayas, they run
down into a narrow strip, which, stretching along the south face of the
Dhdoladhdr, connects them with the larger area of lower tertiaries in Jammu.

The upper tertiaries are, like the lower, divided into three groups. The
lowest of these, known as Ndhan consists of clays and sandstones. the former
being mostly bright red in colour and weathering with a nodular structure,
the latter firm or even hard, and throughout the whole not a pebble of hard
rock is to be found.
466 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [Chap, XVIII.

The middle Siwd4liks consist principally of clays, and soft sandstones, or
sand rock, with occasional strings of small pebbles, which become more
abundant towards the upper part, till they gradually merge into the coarse
conglomerates of the upper Siwdliks. It must be understood that this
classification, being dependent on lithological characters, not on the pale-
ontology of the beds, is not strictly accurate, and it is certain that the dif-
ferent stages must more or less overlap each other on different sections.
Any classification on paleontological grounds is unfortunately impossible
at present, as most of the fossils have been obtained through native col-
lectors, and their localities are not known with certainty. But this is
unimportant for our present purpose, as it seems certain that the three suc- -
cessive lithological stages do represent successive periods of time, though
part of the conglomerate stage on one section was certainly represented
by a part of the sand rock stage on another.

In the north-west of the Punjab, beyond the Jehlam, the whole of the
tertiary rock groups are said to form one conformable system from base to
summit.! Further east their relations are less simple and at first sight
somewhat perplexing. The true meaning of the anomalies was long ago
pointed out by Mr. Medlicott, * but have been illustrated in so much greater
detail by Mr. Middlemiss in his account of the sub-Himdlayas of Kumdun
and Garhwal® that it will be well to turn to this region for illustrative
sections. Here there is normally a perfectly conformable transition from
the Ndhan group to the middle Siwdlik sandstones, and again from these
to the upper Siwdlik conglomerates, This conformable succession, which
is exhibited by many sections, is illustrated on two of the sections repro-
duced on the accompanying plate, but it isnot invariable. Many sections,
as No 3 onthe plate, show the upper Siwdlik conglomerates resting un-
conformably on the eroded edges of Ndhan sandstones, and this peculiarity
of unconformable contact between two members of a conformable system
finds its most striking exemplification in the short section reproduced in
figure 25.*

West of the Ganges the country has not been examined in the same
detail, but it is certain that the same feature exists. In the neighbour-
hood of Ndhan the Ndhan and upper Siwalik groups are in contact along
a line of fault, but the latter contain many boulders derived from the
sandstones of the former, showing that they had been elevated and exposed
to denudation at the time that the upper Siwédlik conglomerates were
being deposited. Beyond the Sutlej, on the other hand, it was found im-
possible to draw any boundary between the two groups, so gradual was
the transition.’

1A, B. Wynne, Records, X, 112, (1877). * Page 468.

2 Memoirs, III, pt. ii, ( 864). 5 Memoirs, III, pt. ii, Chap. IV, (1864) ; see
3 Memoirs, XXIV, pt. ii, (18g0)., also Records, XIV, 169, (1881),
Chap. XVIII. J MAIN BOUNDARY FAULT. 467

Throughcut this eastern area just referred to, the upper and lower
tertiaries are nowhere found in superposition. ‘They occur on opposite
sides of a great fracture, marking the limits of the sub- -Himdlayan
region, and it is at present uncertain whether any keds whose age would
place them with the Sirmur series conformably underlie the Ndhans in
this region.

The examination of the sub-Himdlayas of Jammuhas beén even more
cursory than that of the country further east, and it is at present uncer-
tain whether the same relations, as exist further east between the diflerent
groups of the Siwdlik series, may not be found to prevail between the lower
and upper tertiaries. The unconformity between the two is proved by the
presence of boulders of lower tertiary sandstone in the upper Siwdlik
conglomerates, but the conformity is not equally well proved. The map
accompanying the only published account ! of this region appears to in-
dicate a conformity between the Siwdliks and the Murree beds of the small
inliers at Naoshera, and between the Punch and Jehlam rivers, and on the
whole it is probable, especially if we bear in mind the asserted conformity
of the whole sequence on the further side of the Jehlam, that the relations
of the lower and upper tertiaries are the same apparently contradictory
ones, of conformity on one section and unconformity on another, as are
exhibited by the groups of the Siwdlilk series.

There is but one explanation possible for the known facts, that this
great thickness of deposits, whose unity of lithological type, no less than
the special sections showing conformity between its subdivisions, prove
that they belong to one rock series, must have been deposited during a
period of disturbance, so that while a continuous sequence of conformable
deposits was Leing laid down in one place, in another they were disturbed,
elevated and exposed to denudation.

Along the whole length of the Himdlayas, wherever the junction of
the Siwdliks with the pretertiary rocks of the Himdlayas has been seen,
it is a great reversed fault. To the west of the Bids a similar reversed
fault forms the boundary between the lower tertiaries and the secondary
and palzozoic rocks of the Himdlayas, and inthe intermediate area, where
the lower tertiaries rise up and form part of the lower Himdlayan area
between the Sutlej and the Jumna, this great fault forms, for part of its
ccurse, the boundary between the Sirmur and Siwdlik series,

The fauit is, however, not a mere boundary fault in the ordinary sense
of the term, that is, the fault is not of a date subsequent to the deposition
of the whole thickness of the series whose boundary it forms, nor did this
ever extend, in its full development, far to the north of the line of fault.

1 Records, 1X, 55, (1876).
2H
468 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [ Chap. XVIII

Mr. Middlemiss’ section, reproduced in the woodcut below of itself proves
this, for it is seen that the great boundary fault was fully developed previous

 

Fig. 25.--Section at the head of the Sara valley, eastern Kumdun, showing”
overstep of the main boundary fault by the upper Siwdliks.

to the deposition of the upper Siw4lik conglomerates, which rest on the
eroded Ndhan sandstones and overstep the fault on to the pretertiary slates,
A somewhat similar section appears to be found in the sub-Himdlayas of
the Darjfling district,! but has not been worked out so completely.

Apart from these special sections there are some considerations of a
general nature, which are in reality of greater importance, It will be seen
that this line of faulting forms an absolute boundary for the Siwdliks in
all the range east of the Sutlej, and in the hills west of the Réavi it simi-
larly forms an absolute boundary for the tertiary rocks, no outlier being
found to the north, and only a few small inliers to the south of it. Now,
if the many thousand feet of tertiary strata found south of the fault had
been laid down in a continuous sequence previous to its formation,
they must have extended far to the north of it, and it is almost impossible
to understand how they could have been so completely removed as to leave
no trace of an outlier.

It was Mr. Medlicott who first, in 1859, pointed out the meaning of this,
—that the fault is in effect an original limit of deposition, to the north of
which the Siwdliks never extended. In order to lay special stress on this
peculiarity, and to take the ‘main boundary,’ as he named it, out of the
category of ordinary boundary faults, he described it as a cliff, against
which the Ndhan sandstones were deposited, and an original boundary of
deposition, only slightly modified by subsequent faulting. Subsequent
investigations have shown that this description requires some modification,
but have fully established the conclusion that the main boundary is not
merely a boundary of present distribution, but in effect an original limit of
deposition. In order to understand the grounds on which this conclusion
is based and was originally reached, it is necessary for a while to leave the
tertiaries and consider the submontane deposits of the present day.

1 P,N. Bose, Records, XXIII, 244, (1890).
Chap. XVII J RELATIONS OF SIWALIKS TO RIVERS, 469

The outer margin of the hills at the present day is everywhere fringed
with a band of gravel deposits forming the “ dAd¢dar,” or gravel slope of the
foot of the hills. The extent and constitution of this varies with its position,
Opposite the debouchures of the great rivers draining the central portions
of the Himdlayas it reaches a great development, and is composed almost
entirely of boulders of hard crystalline and metamorphic rocks, which have
mostly been well rounded in their long journey down the river valley. Boul-
ders of limestone are somewhat rare, while the softer varieties of slate and
sandstone are almost absent, having been unable to withstand the severe
treatment they received. In the stretches intermediate between the great
rivers the nature of the gravel varies according to the rocks exposed within
the drainage areas of the streams ; where these drain only from the outer hills
of Siwdlik conglomerate, rounded boulders of hard rock will be found,
elsewhere there are seen tragments of limestone, sandstone, or slate, which
are often subangular, owing to the shorter distance they have travelled
and the smaller degree of abrasion they have undergone, and are always
‘less rounded than the hard boulders of the great rivers.

If we now turn from the submontane deposits of the present day to the
upper Siwdliks we find a remarkable resemblance between them. Not
oaly are the upper Siwdliks so similar to the recent deposits in general
character that they have, not without reason, been compared to an elevated
portion of the plains, but there is precisely the same connection between
their composition and the existent lines of drainage. In the sub-
Himdlayas of Kumdun, there is a great development of the upper and
middle Siw4liks, and especially of the conglomerates, where the Ramgang4
and Kosi rivers issue from the hills. Further west, where there are no large
streams draining from the interior of the hills, the whole Siwdlik zone
becomes constricted and only the Ndhan group is seen. Between the
Ganges and Jumna the upper Siwd4lik conglomerates again attain a great
thickness, and are composed of well rounded boulders of hard rocks, pre-
cisely similar to the débris brought down by these rivers at the present day.
West of the Jumna the conglomerates die out to a great extent, and those
which are seen consist of fragments, to a large extent subangular, of the
older tertiary sandstones, and of the formations found in the outer pact
of the Himdlayas of this region. Where the Sutlej] debouches from the
hills there are at least 4,000 feet of coarse conglomerates, but in a parallel
section, only seven miles off, there is only about 500 feet of them, in the
middle of over 3,000 feet of brown sandy clays.! The same features have
been noticed in the case of all the other great rivers, that the upper Siwdlik
conglomerates attain a great thickness in their neighbourhood and are
composed of waterworn boulders of hard rocks, while inthe intermediate
country they are generally represented by brown clays undistiaguishable

1H. B. Medlicott, Records, 1X, 57, (1876).
2H2
470 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. (Chap. XVIII.

from the recent alluvium, or if conglomeratic the pebbles are of local
debris.

There is but one explanation possible of these features, that the
Himalayan range already existed at the time when the upper Siwdliks
were being deposited, with very much the same boundaries as at present,
with the principal features of its drainage already established, and with
an elevation comparable to that of the present day. The Siwéliks formed,
therefore, the northern fringe of a series of alluvial deposits, whose southerly
extension must be looked for beneath the undisturbed deposits of the
Cangetic alluvium.

But the Siwdliks now form low hills, in which these once horizontal
deposits have been disturbed, elevated, and exposed todenudation. There
has consequently been a southerly advance of the margin of the hills since
the upper Siwdlik age.

The vast thickness of Siwdlik deposits, whose upper division alone
attains many thousands of feet in thickness, all of which were formed sub-
aerially, and even now, after the elevation they have undergone, only reach
a very few thousand feet above the sea, can only have been formed in an area
which was gradually subsiding as the deposits were heaped up. We must
conclude then that the plain country south of the hills, where the conditions
are so similar to those under which the upper Siwdliks were formed, and
where immense masses of debris have been heaped up without raising it very
much above the alluvial plain to the south, is an area where considerable
subsidence has taken place during the recent period.

There are of course no sections showing the actual nature of the
boundary between this area of subsidence on the one hand, and the region
of recent elevation occupied by the outer hills of the Siwdlik zone on the
other, but there are some considerations of a general nature which, apart
from any reasoning from analogy, indicate its nature.

The steady sweep of the boundary along the length of the Himéalayas,
the absence of any deep re-entering angles or outlying patches, show
that it is in the main a structural feature, and that only its details
bave been shaped by denudation and sedimentation. Nowhere are the
upper Siwd4lik conglomerates found passing conformably beneath the recent
deposits of the foot of the hills, and the section of the outermost ridge is
always an anticlinal, whose southern half shows an increasing steepness of
dip in a southerly direction, The beds actually in contact with the syb-
montane gravels may be uppermost Siwdliks or belong to the lower part
of the Nahan group, the dip may be moderate, vertical, or even inverted,!
rarely the whole southern half of the anticlical may have been denuded

! Examples of all these cases may be found | must be borne in mind that the underground
in the sheets of sections published by Mr, | section south of the junction of the Siwdliks
middiemiss, Memoirs, XXIV, pt. ii, (t8c0). It! with the recent deposits i is purely conjectural,
Chap. XVIII. J BOUNDARY BETWEEN HILL AND PLAIN. 47t

away and covered up by recent deposits, but where it is seen there is
usually a rapid increase in the steepness of the southerly dip near the
margin of the hills, These indications of a line of special bending of
the strata close to the southern edge of the hills help out the suppo-
sition that the actual demarcation of the two contiguous areas—one of
elevation, the other of subsidence—may be of the nature of a fault, one
side of which has been raised and the other depressed, the depression being
oe up as fast as made by the abundant debris brought down from the
ills,

Though there are no sections showing the nature of the junction between
the undisturbed recent deposits of the plains and the upraised Siwéliks
the sections in the sub-Himdlayas throw some light on its nature. Wher-
ever the Siwdlik zone attains any considerable width it is found to be
traversed by one or more reversed faults of great throw, running more or
less parallel to the outer boundary on the one hand and the main boundary
of the Siwdliks on the other. These faults all show an ascending section
on the outer (southern) side and the dip usually flattens towards the fault,
where the uppermost beds are seen in contact with strata of a very much
lower zone. Moreover, the older beds are invariably thrown into an anti-
clinal immediately north of the fault, while the southern half, when it is
present, shows an increasing dip as the fault is neared, exactly as is the
case in the anticlinal of the outermost ridge, and occasionally, as in that case,
the southern half is cut out.}

The relations of the rocks on either side of these great faults are so simi-
lar to what we have inferred is probably the case between the Siwdliks of
the outermost ridge and the deposits of the sabmontane region of the
plains, that it is natural to regard each fault as marking a former limit of
the disturbed tract, and the successive faults as indicating a step by step
southerly advance of the outer margin of the hills, According to this
hypothesis the great main boundary would mark approximately the south-
ern limit of the Himalayas at the commencement of the Siwdlik period,
north of which the upper tertiary deposits did not extend to any great
distance or in any great thickness.

We are not, however, confined to the conclusions that may be drawn
from direct observations in dealing with this problem. It may also be
attacked from the purely physical and mathematical side, as has been
done by the Rev. O, Fisher in his great work on the physics of the earth’s
crust.! Mr. Fisher adopts the hypothesis that the solid crust of the earth is
of limited thickness and rests on a magma of greater density, whose condi-
tion is actually or virtually that of a fluid, As the central core of the earth

1 Physics of the Earth’s Crust, tst ed., 8° London, 1888, pp. 114-41.
472 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [ Chap. XVIII.

cools down by the conduction of heat away from it, the outer crust is
left partially unsupported by the consequent contraction, exposed to a
greater strain than it is capable of bearing, and yielding along lines of
weakness, is thickened both upwards and downwards from a zone some-
where in the thickness of the crust, above which the material will on the
whole be forced upwards, and below it downwards. This zone, called
the neutral zone, is placed, for reasons unnecessary to enter into here,
as its exact position is not of great importance, at three-fifths of the
thickness of the crust from its upper surface. If the subjacent magma
had a density one and two-thirds as great as that of the overlying crust the
apward protuberance would be supported by the buoyancy of that portion
which had been thrust downwards into the magma. But such a great
disparity cannot exist, and the extra weight of the elevated tract will,
consequently, bear the crust downwards on either side as is indicated
in the diagram, fig. 26, till sufficient of the lighter solid material is de-
pressed into the denser magma to provide the requisite buoyancy.

As soon as an elevated tract is formed denudation will commence, and
as it is extremely unlikely that the protuberance will be symmetrical, a
larger amount of material will be deposited on one side of it than the
other, In the diagram this is supposed to take place on the right hand side
and its effect will be to depréss that side of the range more than the other
while the elevated tract will be lightened by the removal of material
from its surface. As a consequence, the centre of gravity will be shifted

 

 

Fig. 26.—Diagram to illustrate Revd. O. Fisher's theory of mountain .formation—

A.—Upward protuberance, or ‘ elevated tract.’
B.— Downward protuberance or ‘ root.’
P. R.—Original limits of depressions on surface,
P.’ R.’— Subsequent limits after the deposition of sediment represented by shading,

a8 removal of the portion of the elevated tract represented by dotted
ines.

Q. Q.’ S. S.’—Original and subsequent limits of downward protuberance.
G. G.’ H. H.’—Original and subsequent positions of centres of gravity of crust, and of
displaced fluid respectively,

towards the side on which there is the greatest accumulation of debris,
while the centre of flotation of the downward protuberance remains
unchanged. The disturbed tract is consequently thrown into a state
of unstable equilibrium, which can only be made stable by such a movement
Ohap. XVIII.] FISHER’s THEORY GF MOUNTAIN FORMATION. 473

of rotation as will diminish the size of the depression on the left hand
side, where there is least sediment, and extend it on the right hand
side, where most has been deposited. As this action goes on, the
depression on that side where least sediment is deposited may become
obliterated, and as the crust is exposed to a tensional strain on this side it
may be that fissures will open and volcanic outburst take place.

Such, briefly stated, is Mr. Fisher’s theory of mountain formation, and
there is, on the most superficial view, a considerable resemblance to what
we know to have taken place in the case of the Himdlayas. The great

_depression in which the Gangetic alluvium has been accumulated corre-
sponds to that in which the greater bulk of debris derived from the denu-
dation of the mountains is deposited, and the obliteration of the eocene
sea of the central Himalayas, accompanied as it was by a great outburst
of volcanic energy, would appear to correspond to what the theory points
out as likely to happen to the depression on the other side of the range.
But the greater part of the elevation nas taken place since the oblitera-
tion of the eocene sea of the central Himdlayas, and we are then met by
the difficulty that we have to assume a rotation of the whole elevated tract
of Central Asia. We have no reason to suppose that so large a mass ol
the earth’s crust would have sufficient rigidity to allow of its rotation as a
whole ; rather, there is good reason to suppose that it would yield infinitely
to a long continued stress. Moreover, Mr, Fisher’s theory takes no notice
of horizontal compression, which in the case of the Himalayas has certainly
gone on up to a very recent period. The fact is this movement of
rotation, with the degree of rigidity it demands, as well as the ignoring of
the effects of lateral compression after the elevated tract is formed, are
mere generalisations necessary to-bring the complex conditions of nature
within the powers of mathematical investigation, and are not intended
to be taken literally, as an exact account of what actually takes place.

 

 

Fic. 27.—Diagram to illustrate the theory of the elevation of the Himalayas, corresponding to
the right-hand half of fig. 26. Horizontal scale about 60 miles, vertical about 30 miies
to 1 inch,

A.—Massif of the Himalayas.

B —Root of the same. :
C.—Earlier marginal deposits, compressed and elevated.
e.— Continuation of the same, depressed and undisturbed.
D.— Subsequent deposits overlapping C.

6.--Sinking of lower surface of crust due to C and D.

Bearing this in mind, we have now to consider what modifications are re-
quired to fit the purely mathematical theory to the more complex conditions
474 GECLOGY OF INDIA~ AGF AND ORIGIN OF HIMALAYAS. [ Chap. XVIIL

of actuality. In the first place the elements of rigidity and rotation being
abandoned, we need not consider the left hand side of the diagram, and may
redraw it in greater accordance with the conditions of the Himdlayan region
(see lig. 27). We have now an elevated region 4 subjected to denudation
and adjoining it an area extending to 2, on which deposition is taking place,
the deposits being contributed by the elevated ground A to the north,
and the waste of the rockareato the south. The tract being supposed to
be in equilibrium, the surplus floating power of B will cause it to rise when
A is lightened by denudation, and the load thrown on D will cause it to
sink, especially in the neighbourhood of A where the load is greatest, till -
the magma displaced by the lower surface of the crust is sufficient to float
the load. The result will be, firstly an extension of the depression ina
direction away from the elevated tract A, and, secondly, a strong tendency
to either fracture or flexure of the crust at the junction of Aand D,

As we may take the crust to be infinitely yielding to long continued
stresses, there is no reason why that produced by the lightening of the
one area and the loading of the other should not be relieved simply by the
sinking of the latter and the rising of the former on either side of a sep-
arating plane. But denudation and deposition are not the only forces at
work, for to bring the case into connection with that of the Himalayas,
we must suppose compression to be continually in progress. This will be
relieved partly by an additional elevation of A, but also by the compres-
sion aud consequent elevation of the marginal deposits D, which would
not offer the same resistance af)the already consolidated beds of A. In
this way the deposits on the edge of the depression would gradually
come to fcrm part ol the tract 4, whose boundary would advance towards
R, but not to the same extent as the shifting of the outer boundary of the
depression towards &,

With this amplification,—for it is no modification of Mr. Fisher’s theory,
but merely a more detailed explanation of part of the process which does
not lend itself to mathematic investigation,— we find it easy to explain the
true nature of the great reversed faults which traverse and bound the Siwd-
lik zone. ‘They mark the successive limits between hill and plain, be-
tween the area of deposition and depression on the one hand and of denu-
dation and upheaval on the other, and the small amount of disturbance
which may generally be noticed in the beds immediately in contact with
the fault plane is explained by the fact that their immense throw is not due
to the effect of horizontal pressure acting-on an inclined plane, but princi-
pally to a vertical pressure, downwards on the one side, ard upwards on
the other, of the fault.

It appears then, that so far as the tract along the foot of the hills is
concerned, the conclusions drawn from observed facts and from theoretical
Chap. XVIILJ EVIDENCE OF THE BORINGS. 475

deduction agree with each other, but we have another check on the theory,
for, if it is true, the Gangetic plain must have originated at the same time
as the great mountain range to the north, and gradually increased in width
by the subsidence of the rock area along its southern margin. Now, when a
stream issues from the rock area on to the alluvial plain, it is the coarsest
debris which is first deposited, while the finer grained material is carried
further and deposited at a greater distance from the margin of the alluvium.
Consequently, if a boring be sunk near the southern limit of the plain, where
it has been encroaching on the rock area, the beds passed through near
the surface should be finer grained on the average than those passed
through lower down, for these last belong to an earlier period, when the
edge of the alluvium was less distant than it is at the present day. Near
the northern margin of the plains the conditions should be reversed, for
there the rock area has been encroaching on the plains and the upper bed
should be composed of coarser debris on the average than those lower down.

The deep borings that have been put down in the Gangetic plain are
four in number, Of these, two, at Umballa and Fort William respectively,
are well situated for testing the hypothesis ; a third, at Agra, is less suitable
for reasons which wiil shortly appear, while the fourth, at Lucknow, being
well out in the middle of the plain, does not appear to have gone deep
enough to give any evidence of importance.

In dealing with the records of these boreholes we are harassed by the
vagaries of nomenclature indulged in by the men, never trained geologists,
to wkom the conduct of the operations was necessarily entrusted. But by
adopting two classes only, sand and clay, and by including in the former
‘sand, ‘coarse sand,’ ‘clayey sand,’ etc., and in the latter ‘clay,’
‘silt,’ ‘sandy silt,’ ‘limey silt,’ etc., a fair idea will be reached of the
relative coarseness of grain of the beds passed through at different depths
of the same boring. This method is perfectly justifiable, as, for the present
purpose, the absolute coarseness or fineness of grain is immaterial, and we
merely want to know whether in any part of the bore hole the beds are on
the whole coarser or finer grained than those above and below.

Adopting this system of classification, we may make an abstract of the
Fort William boring,! thus :—

o ft. to 100 ft... : . ; ‘ : . Sand o Clay 100
100 5, 4, 200 y » OWL » 89
200 4, 4, 300 5, » §=695 es 5
300 5, 59 400 4, . » 98 ” 2
400 4 481 4, , : . A ‘ 2 - oy 81 * 0

The increase in coarseness of grain -of the beds passed through is con-
spicuous enough in this abstract, but the reality is even more striking, for
inthe sand from 180 feet downwards, some beds of gravel and pebbly sand

' For detailed section see Records, XIV, 221, (1881); Calcutta fourn. Nat. Hist., 1, 324, (1841).
476 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [Chap, XVIII

are included, and the boring was finally brought to a standstill in a bed of
gravel which it was not found possible to penetrate.

The second boring of importance is that made at Umballa.1 Adopting the
same broad classification of clay and sand, we get the following result :—

o ft. to 100 ft. 5 é F a .Sand 59 Clay 37 Soil 4
56 » 42 Kankar 2

100 55 3, 200 ,, ‘ . Y e iF Py

200 33 59° 300 55 ‘ f : - 9» 58 s 42
300 99 400 455 . . . . . ory 54 ”» 46
400 45 5, 500 yy . . : : - » 46 » 654
500 55 600 ,, . . . . + » 37 ” 63.
600 55 5 701 3» . . ‘* . eo 6 » 95

Here we have, as the hypothesis requires, a very distinct increase in
coarseness of texture in the upper beds as compared with the lower.

Besides these two borings, one has been put down at Agra, the evi-
dence of which is slightly vitiated by the peculiar local conditions. The
abstract of the section is as follows? :—

o ft. torooft. . ‘ 3 . Sand 65 Clay 35 Kankar o
100 5, 53 200 35, ° . . - » 44 ” 56 ” o
200 ,, 39 300 . . . - 9 0 ” 973 ” 23
300 55 4400 sy, . . . - » 18 ” 705 ” 53
400 ” .By 481 ” * + * * »” 7 ” 74 ” 0

Here there would seem to be an increase in coarseness of texture, both
upwards and downwards, from 200 feet. The explanation of this is to be
found in the fact that the surface deposits round Agra are largely com-
posed of blown sand, and it is probable that the sand beds found in the
uppermost 160 feet of the section are of zolian origin, while below that
the beds are alluvial and exhibit the gradual upward increase in fineness
of texture required by the hypothesis.

The fourth boring at Lucknow has been sunk to a depth of 1,336 feet.
As might be expected from its situation, there is no marked increase or
decrease in the coarseness of the beds passed through, but near the bottom
of the boring some beds of coarse sand were found, and these may indicate
an approach to the base of the alluvium and mark a time when its southern
boundary was not far from Lucknow. ,

To sum up, of the four deep borings which have been made, two are,
completely in accordance with the hypothesis, the third one is in favour of
it, though its evidence is vitiated by peculiar local conditions, while the
fourth is so situated as to give no evidence one way or the other till it is
carried to a greater depth.

We find then that the inductions from observed facts regarding the
southerly advance of the margin of the hills, the nature of the boundary
between hill and plain, and the mode of formation and growth of the

1 Records, XIV, 233, (1881). | 2 Records, XVIII, 121, (1885).
Chap. XVIIL J AGE OF THE HIMALAYAS, 477

Gangetic plain, agree in all essential points with the deductions from Fisher's
theory of mountain formation and using the one to elucidate or amplify
what remains doubtful in the other, we may approach the interesting sub-
ject of the date of the commencement of the elevation of the Himdlayas.

The occurrence of marine nummulitic beds at a height of many thousand
feet on the north face of the main snowy range in Hundes, and at a height of
40,000 feet in Zanskar, shows that the elevation of this part of the Himalayas
must have taken place entirely within the tertiary period, Further east we
have not the same conclusive evidence, but the upper cretaceous fossils that
were brought from north-west of Lhasa show that the elevation of this part
of the Tibetan plateau could not have commenced at a much earlier period.

The limitation of the marine mesozoic and palzeozoic rocks to the northern _
flanks of the main snowy range, and their absence, so far as is known, to
the south of this, may be due to an original limitation of deposit, or it may
well be due to the country over which they (are wanting)having been more
fapidly elevated, and consequently exposed to more active denudation,
bat even if the southern limit of these marine formations represents ap-
proximately the recurrent shore lines of a long series of epochs, it is difli-
cult to believe that a mountain range at all comparable to the Himalayas of
the present day lay immediately to the south of them. The present geo-
graphical and geological connection between the Himdlayan range and the

ibetan highland is too close to make it at all probable that the elevation
af the latter was altogether posterior to, and independent of, that of the
former, and consequently the elevation of the Himalayas as a mountain
range cannot have been long in progress, if it had commenced, when the
spa flowed over Tibet at the close of the secondary period.

' On the southern side of the Himdlayas there is not the same direct
evidence. The close connection between the older rocks of the Assam
range, and the corresponding ones of the Indian Peninsula has already
been noticed as indicating that the present limits between the peninsular
and extra-peninsular areas had not been established at the time that they
were being deposited, and the presence of subaerially formed Gondwdna
rocks in the eastern Himdlayas suggests, though it does not prove, that
they were formed in the same land area as those of the Peninsula and
that no depression, corresponding to that now occupied by the Gangetic
alluvium, was in existence at the commencement of the secondary period.

The complete absence of any known exposure of marine nummulitic

rocks between western Garhwél on the one hand, and the Garo hills on the
other, might only mean that the shore line ran south of the present limit
of the hills, and that the nummulitic beds are hidden’by the Gangetic
alluvium, but there is not so close a relationship between the nummulitic
478 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS, C Chap. XVIII.

faunas, so far as they are known, of the Punjab and of Assam as to neces-
sitate, or even suggest, so direct a communication between the two areas,
There is consequently some degree of probability that the Indo-Gan-
getic depression had not been established at the commencement of the.
tertiary period and we again get the close of the secondary period as the
probable date of the commencement of the elevation of the Himalayas.

The stratigraphical relations, between the nummulitic beds of the north-
west portion of the lower Himdlayas and the subjacent deposits, point to
the same conclusion. There is not only a general parallelism of strati-
fication, which might result from the compression both have been exposed
to, but there is a very close resemblance in the nature and degree of disturb-
ance they have undergone, and the nummulitics lie with perfect parallelism
of bedding on an eroded surface of the former pretertiary deposits, wher-
ever a section showing the original contact between them is found,
Were this merely a local phenomenon observed on one or two isolated
sections no importance need have been attached to it, but when it is seen
wherever the contact between the two rock series has not been modified
by faulting, from the inliers of the Jammu hills on the one side, to the
outliers east of the Ganges on the other, it shows that there had
been no appreciable disturbance of the older rocks, now forming this part
of the Himalayas, when the nummulitics were deposited. In other words,
that if the elevation of the Himalayas had already commenced in eocene
times, it had not extended into the north-western portion, or was confined
to the central portion of the range.

The close connection in structure and distribution of the upper Siwé-
lik conglomerates and the submontane deposits of the present time has
already been appealed to as evidence that the Himdlayan range existed
in pliocene times with very much the same limits and elevation as at the
present day, and with the main features of hydrography already marked
out. But these coarse conglomerates are confined to the upper Siwaliks.
As we descend the section pebbles get smaller in size and less in number
till, in the lower part of the Siwdliks proper and throughout the immense
thickness of the Ndhans, not a pebble is known to occur.!

-It might be held that this was due to the southerly advance of the foot
of the hills, and that we must look for the coarse conglomerates of middle
Siwdlik and Néhan age in the hills north of the main boundary. It has,
however, been shown to be extremely probable that neither the Siwdliks
proper nor the Nahan group ever extended, in anything like their full
thickness, much to the north of the main boundary, and the absence of
any known outlier, though merely negative evidence, cannot be altogether

' There is only one recorded instance of a | with so much hesitancy, that it may well be
conglomerate supposed to be of Ndhan age, and | neglected,—Memoirs, III, pt. ii, p. 135, (1864).
that case is so exceptional and its age given
Chap. XVIII. J EVIDENCE OF THE TIBETAN FAUNA. 479

ignored. A more probable explanation is that during the formation of the
lower portion of the Siwdlik series the hills to the north had not attained
anything like their present elevation, and that the gradients of the river
beds had not become sufficiently steep to enable them to transport any-
thing coarser than fine sand. If this be the true explanation, as seems
extremely probable, the greater part of the elevation of the Himdlayas has
taken place since the miocene epoch, and it is impossible to date its com-
mencement much further back than the commencement of the tertiary or
the close of the secondary period.

Another argument of an entirely different character has been adduced
by Dr. Blanford, which curiously confirms the conclusion regarding the
date of origin of the Himalayas, arrived at on purely geological grounds.

He points out! that the mammalian fauna of Tibet has a proportion of
species, and even genera, peculiar to the region which is not exhibited by
any other continental area of the same size. Omitting. all doubtful
forms, and taking no account of varieties or subgeneric types, the known
Tibetan fauna consists of forty-three species belonging to twenty-six
genera, of which twenty-seven species and four genera are not known
outside Tibet. Moreover, by far the largest proportion of species ranging
outside of Tibet is exhibited by the carnivora, only four out of nine
species of ungulata being known outside Tibet, and two of these are res
presented in Tibet by well marked varieties, while out of sixteen species
of rodents only one is not purely Tibetan.

On the now universally accepted theory of the origin of species by
descent and modification so large a proportion of peculiar species indicates
a long period of isolation. In the case of island faunas, this isolation is
due to the sea barrier which mammals cannot cross or can only cross with
difficulty, but in the case uf Tibet the isolation must be a climatic one, due
to the superior elevation of the region, and after comparing the degree of
specialisation of the fauna with that of various islands Dr. Blanford comes
to the conclusion that this isolation must have commenced in middle tertiary
times. ‘This agrees remarkably with that arrived at on purely geological
grounds, and from a study of the relations of the Siwdlik to the tertiary
faunas of Europe,’ that the elevation of the Himalayas commenced with
the tertiary era and that the range only attained an elevation comparable
to that which it now possesses towards the commencement of the pliocene
period.

Two views have been propounded, regarding the antiquity of the Himd-
layas which are antagonistic to that just put forward. The first of these,

| Geol. Mag., 3rd dec. 1X, 164, (18y2). ' 2 Supra, p. 367.
480 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS, [ Chap. XVIII.

which regards the mountain chain as much older than the commencement
of the tertiary period, requires special notice, as it has been advocated by
Mr. C. S. Middlemiss, the author of the most detailed study of any portion
of the range yet published. It is supported by arguments derived on
the one hand from the special structural features of the southern margin
of the Himdlayas in Kumdun and GarhwéAl, and on the other by the de-
gree of disturbance which the rocks of various ages have undergone in
the same region.

The special structural features are summed up in, and illustrated by, a
section drawn north and south along the Ramgangé and Peldni valleys in
E. long. 78° 49’, which is reproduced on the accompanying plate. In this
it will be seen that, starting from the plainward margin, we have the
Ndhan sandstones conformably covered by a great succession of middle
Siwdlik sand rock, north of which the upper Siwdlik conglomerates are
brought in by a small fault of no structural importance, The upper Siwdlik
conglomerates are brought into contact with the lower portion of the
Ndhan group by a great reversed fault of 11,880 feet thrown along the
fault, or 6,380 feet in a vertical direction. From this fault there is again,
after some undulation of the strata, an ascending section through the
greater part of the sand rock, but before the conglomerates are reached
the beds low down in the Ndhan group are again brought up, and after a
series of anticlinal and synclinal folds the topmost beds of this group are
brought into contact with the pretertiary beds of the Himdlayas along the
main boundary,

The rock in contact with the Ndhans at the main boundary is a massive
unfossiliferous limestone of unknown age. It is overlaid by the T4l beds,
presumably mesozoic, and these again by marine nummulitics of the Subdthu
group. North of the nummulitic band there is again a reversed fault and
purple slates, and volcanic brecias come in, beyond which a great reversed
fault brings in crystalline schists.

It will be seen that there are here five zones, each bounded on the north
by a great reversed fault, and each successive one showing an older group
as its newest member, In the outermost zone the section ranges from
Ndhans to the upper Siwdlik conglomerates, in the next the sand rock in
the newest group seen, in the third this is wanting and the section only
ranges up to the upper Ndhans. In the succeeding zone an entirely new
set of rocks comes in, the newest of which is eocene, while in the last the
rocks are of unknown, but certainly at least palaozoic age.

It has already been pointed out that the great reversed faults of the
sub-Himalayan zone probably mark the successive positions of the outer
margin of the hills, that is the limit between the area of elevation on the
one hand and subsidence on the other, Mr. Middlemiss opines, with a

» Memoirs, XXIV, 124, (1840).
Chap. XVIII. J MIDDLEMISS’ THEORY. 481

very good show of reason, that in each case the youngest rock seen south
of the fault marks approximately the period of its completion, and of the
commencement of the one next to the south, so that the first fault counting
from the south may be ascribed to the upper Siwdlik, the second to the
middle Siwdlik period, the third to the close of the Ndhan period, and the
fourth to the eocene.

' The argument on which this conclusion is based may be epitomised as
follows.! It isa common character of these long narrow zones bounded
on the north by a reversed fault, that they carry along their northern
border a still narrow zone of the newest rock they contain, a zone which
has been preserved “ because the fold involving that zone, and the re-
versed fault to the north of it, were the companions of the upheaval of
that zone from a condition of deposition; that is to say, the uppermost
stratum had only just been deposited when it was folded and faulted, and
so wrapped up with the older zone to the north that it was preserved from
subaerial denudation.”’ It is argued that if this was not so. if for instance
the nummulitics had been covered by the Ndhans and Siwédliks, and ex-
posed to denudation before they were folded and faulted, the upper members
would have been removed in places and left in others, and that when the
faulting subsequently took place the irregular patchwork of strata resulting
could not have been formed into the regular zones now observable. This
argument, however, assumes that the deposits must have been elevated
without disturbance and exposed to great denudation before the faulting
took place. This is by no means necessary. The whole thickness
might have been deposited over the nummulitics, and if the faulting and
folding had gone on fart passu with the clevation an arrangement, ana-
logous to that which now obtains, might have resulted from the different
degree of elevation, and consequent different intensity of denudation, the
different zones had undergone. It has been shown, on quite independent
grounds, that this supposition is an improbable one, but there is no reason
why the nummulitics of Garhw4l may. not have once been covered by a
great thickness of deposits corresponding to the upper members of the
Sirmur series further west, if not by part of the Ndhan group. The regu-
larity of width of the outcrop, its narrowness, and the absence of these
upper members can be sufficiently explained by the high dip of the strata,?
the narrow patches remaining having been preserved by their having
been elevated to a lesser degree, and consequently less exposed to
denudation, than the higher beds.

But even if Mr. Middlemiss’ argiment be admitted to its fullest extent,
it does not throw the elevation of the Himalayas further back than the

1 For the full statement see Memoirs, XXIV, , drawn cros-es the strike at an angle of about
174 ff, 40°. The dip shown is consequently con-
2 The line along which the section is | siderably less than the true dip.
482 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. (Chap. XVIIL

commencement of the tertiary era, for to the north of the fault, which
bounds the nummulitics, we come to conditions so different, both strati-
graphically and structurally, from those wkich obtain to the south that we
can no longer apply the same arguments, and as will be shown presently
the question of whether the existence of a land area, immediately north
of the nummulitic outcrop at the time when these beds were being deposit-
ed, can be admitted as evidence regarding the date at which the elevation
of the Himdlayas commenced, depends entirely on the exact meaning we
attach to these words.

The second argument depends on the different degree of disturbance
exhibited by the successive zones. As can be seen’from the section, the
beds in the outermost zone have undergone least compression, those of the
next more, and so on. Mr. Middlemiss argues that this increase in the dis-
turbance the beds of each successive zone have undergone, is the result of the
successively greater periods of time during which they have been exposed to
the disturbing forces, and that the far more intense compression, to which
the rocks within the innermost tertiary zone have been exposed, indicates
that they have been exposed to pressure, during a period of time, which
would carry back the origin of the Himdlayas far beyond the tertiary era.
To this it might be answered that, even if the compression of every zone
had been contemporaneous, it is natural to expect that its intensity would
not be everywhere uniform, but would die out laterally, graduating from the
zone in which it was greatest to that in which there had been none.

It is, however, probable that, in the tertiary zones, the different degrees
of compression exhibited by the successive bands is, to a large extent, the
result of the different periods during which they have been exposed to com-

pression. And when we come to the far more intense compression ex-:

hibited by the older brocks, which, as descried by Mr. Middlemiss, have
been cleaved and foliated by the intensity of the compression they have
undergone, he himself affords an explanation, in the observation that the
strike is often transverse to that of the tertiaries, indicating that the com-
pression had not all been in a direction transverse to the course of the
range! This diversity of strike is appealed to as showing that the com-
pression of the Himdlayas was in part due to other, and older, directions of
thrust than those: which produced the folding of the sub-Himdlayan and
many of the Himalayan rocks.

Here we are at once brought face to face with the question of what is
meant by the commencement of the‘elevation of the Himdlayas. Seeing
that the present state of the range is the result of a long chain of physical
causation, each step of which was the inevitable result of all that went
before, it is impossible to say what was the first origin of the Himdlayas.

! Memoirs, XXIV, 183, (1890).
 

= J

TURKISH
\NOIAN WATER SHED
WATER SHED TIBETAN TABLE-LAND >

NORTHERN SLOPE
LITTLE TTT TOT ee TTT erect ees

HIMALAYAN SLOPE

mmm MOOT mM a

rome
T t

 
 
 

PLAINS OF INDIA
PLAINS OF CENTRAL
ASIA

         
 

 

 

SWITZERLAND ALPS ITALY
BERNESE PENNINE

Comparative sections of Alps & Humnalayas, after Strachey. Scale lurch - 35 miles.

 

 

Section across the Dehra Dun, after R. D. Oldham. Scale lich -2miles.

   

     

 

 

 

 

S.8° W. : ONL BPE.
JIRINJALA pte mo i
“SUKIRAU d : RAMGANGA R. pandas mu arais « f NN :
NEAR LALDHANG Ae a Eero, do | DT MAULIA ee mN MANDAL R.
i i oes SES Se . oe ns =< Met ia eee ete SS : : < ee i = ~ — wa . fj <\ j ; oF ‘ Th
See, ee me 2 we i ee eae ee mst LoS ~ eS ae ‘ X \ YQ £ ~ ; m . A N yj al) ih} m1
b

 

 
 
    
  
  
 

 
  
    

  
 

        
 

   

 

S. | N.
m Es ig ae Mea - & Ga ree Re Pees Rd
a oe N piteten, ttt, Ff a es . settee, “t ~.SIDHRUR DHAR
See OB * HaLousyaLa GIRAAD! SOT HAR GADI
RAMGANGA R. 7" . Soe : Bae
NEAR KALAGAAH RAMGANGA R.: GUTUAGADH |g \PELANI R. Pod eSBELANI RSF te “
Tr i 3 ON R

   

  
   

   

   
 
 

Us

+e, i Sn
OI | WS - | NWS SSS
NS

m fr nm

  

6 | #6

}

  
 

“i ox) cx ey
SS PON aN XD © Tax
SOSA SSNS S DOIN WN WS
6 b c c c doc b b c c
Section across the Sub-Himalayan xone exhibiting its relations to the older zones, after C.S. Middlemiss. Scale linch=-/tmiles.

SECTIONS ACROSS THE HIMALAYAS & SUB-HIMALAYAN ZONE.
r=recent, d= upper stwalik conglomerate, c= middle siwalik sandstones, b=lower (Nahan) siwaliks, a= nmumemulitic, n=older rocks of Himalayas.

   
Chap. XVIII. J DEFINITION OF TERM. 483

0

We may throw it back to the period when the earth first acquired a solid crust,
or still further back to that primeval chaos from which, according to one
hypothesis, the universe was evolved. But in this place the words are
used in a much more limited meaning. If the limitation of the palzozoic
and mesozoic deposits along the northern flanks of the snowy range repre-
sents at all approximately the general limit of land and sea during their
deposition, it might be maintained that the general course of the Himalayan
range had been determined in paleozoic times, and yet the elevation of
the Himalayas in the sense in which the words are here used, might not
have commenced till the dawn of the tertiary era. The further back in
time we go the more difficult does it become to follow the sequence of
cause and effect, and in speaking of the elevation Himdlayas only that
final compression is meant, which caused it to rise as a conspicuous moun-
tain range with the same limits and extent as at present, and the antece-
dents which may or may not have been the direct cause of this result are
excluded.

Taking this restricted definition, the transverse sttikes mentioned by
Mr. Middlemiss, and the systems of compression they indicate would not
be connected with the elevation of the Himalayas or belong to the Himé-
layan system of disturbance. :

It would be unnatural to suppose that the great area now occupied by
the Himdlayas bad in no part been exposed to compression, previous to
the end of the mesoZoic era, and it is noteworthy that the most conspicuous
instance of transverse strike quoted by Mr. Middlemiss, where a north and
south strike extends for sixty miles, lies on the continuation of the Ardvalli
range. Now, without assuming, what there is no possibility of proving,
that the Ardvalli range ever extended so far north, the supposition would
explain how there might be an intense crushing of the older rocks,
accompanied by a strike transverse to the general direction of the range,
which was due to a totally distinct system of disturbance from that which
produced the Himalayas.

In this way we see how the crushing of the older rocks of the Himd-
layas and the divers strikes they exhibit, which Mr. Middlemiss rightly in-
terpreted to indicate successive systems of compression ranging over along
period of time, may have been largely anterior to that final compression to
which the elevation of the Himdlayas is here restricted.

Another opinion regarding the antiquity of the Himdlayas, which re-
quires notice, is that recently propounded by Sir H. H. Howorth,! who has
gone to the opposite extreme and regards the elevation of these mountains

1 Geol. Mag., 3rd dec., VIII, 97, 156, 294, (1891) ; IX, 54, (1892).
21
484 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS. [Chap, XVIIL

as having taken place almost, if not quite, entirely within the pleistocene
period. This opinion appears to have been adopted in the first place
to satisfy the requirements of his theories regarding the entombment
of the mammoth in northern Siberia, but is based principally on the
supposed absence of any traces of glaciation in the Himdlayas, as also in
the Altai, Caucasus and Ural mountains, all of which are supposed to have
been elevated within the pleistocene period.

We have already shown that there is evidence of a former great exten-
sion of the Him4layan glaciers, evidence which might be amplified to almost
any extent, but the only instances, in addition to those already quoted,
which need be mentioned here, are the glaciers on the Babeh pass, now
barely more than a mile in length, which at one time extended at least
fifteen miles and probably more, and the morraine recorded by Colonel
McMahon on the southern slopes of the Dhaoladhar, at an elevation of
only 4,700 feet.

The Himalayan glaciers, it is true, never spread over the low ground in
great ice sheets like those of Europe, but there is reason enough for this
in the thirty degrees of latitude by which the Himdlayas are nearer the
equator than the Alps, and in the much greater distance which separates ©
the watersheds from the lowlands. In the Kangra valley, where alone the
high mountains rise steeply from-the low ground at their foot, there is good
reason to suppose that the glaciers once reached to below 2,000 feet above
the sea.? And the erratics of the Potwdr® show that ice in large quanti-
ties was not unknown there at one time. As it is out of the question to
suppose that even in the glacial period these glaciers could have origi-
nated at low altitudes, the only possible conclusion is that the mountains
must then have had very much their present elevation.

Another argument for a greater age of the Himalayas is the time re-
quired for the excavation of the great river valleys. Sir H. H. Howorth
avoids this difficulty by denying that the valleys are the work of the rivers
that flow through them, but it is impossible for any one who has studied the
action of subaerial denudation not to see that the forms of the hills and
their intervening valleys are due to the action of rain and rivers, aided by
frost. A glance at the photograph so admirably reproduced in the frontis-
piece of this volume will show to the initiated eye that the shape of the
mountain is due to the disintegration of the rock by frost aud the removal of
the debris from the hollows by streams and glaciers, and not to any disrup-
tive force. Even if we could acknowledge that the courses of the drain-
age were in the first instance determined by fissures, a long period of time
would be required for the opening out of the valleys and the removal of that
vast mass which the beds of the Siwdlik series tell us was brought down
from the Himdlayas.

' Records, XV, 49, (1882). ® Records, X, 140, (1877) ; Supra, p. 42.
* Records, 1X, 56, (1876).
Chap. XVIII. J RECENT INCREASE OF ELEVATION. 485

These arguments would be sufficient to show that the Himalayas must
have existed as a mountain range previous to the glacial period at any
rate, but it is not necessary to appeal to them, for the evidence of the
pliocene sub-Himdalayan deposits shows that the range must then have bad

very much its present elevation, with the main features of the existing
drainage system already marked out,

The close agreement in the results attained by the several distinct
lines of purely geological reasoning, and that derived from the pecu-
liarities of the living Tibetan fauna, gives a very strong presumption in
favour of the correctness of the conclusion arrived at, and discredits at once
the hypotheses of an older or a later date for the origin of the Himdlayas
than here maintained. There seems, however, to be this much truth
in Sir H. H. Howorth’s supposition of the recent rise of the Himdlayas, that
their elevation, if not still in progress, has only recently ceased, and that
they are probably now somewhat higher than they were during the glacial
period.

The evidence pointing to this is of various kinds. There is, firstly, the
natural presumption that the mountains which now form the most elevated
peaks of the world cannot be in a state of decadence, and as there is no
such thing as rest in nature, that they must be still growing. Then
there are two recorded cases! where a differential movement of the oppo-
site sides of a fault has taken place at so recent a period as to cause
interruption of the minor drainage courses, and to exhibit itself as a distinct
rise in the surface of the ground, which has hardly been modified at all by
denudation. These earth movements show that the Himdiayas are still
in a state of strain, and we may naturally conclude that this s:rain is due
to the compression which has caused their elevation.

Better evidence is yielded by the sub-recent fossil fauna of the Hundes
plain. It was formerly believed that this fauna was tertiary, the presence
of a rhinoceros was supposed to indicate that the deposits must have been
formed at a-very. much lower level than that at which they are now found,
and that they had subsequently been elevated several thousand feet without
any discernible disturbance. The incorrectness of the first supposition has
already been shown.” The last is one that cannot be granted, and as regards
the second, the presence of the peculiar Tibetan genus Pantholops out»
weighs the evidence of the rhinoceros. It is true that a rhinoceros could
not exist on the present plains of Tibet, not on account of the cold, for the
Tibetan species may well have been protected from that by a thick coat of
fur, but on account of the impossibility of its picking up a living from the
scanty vegetation of these arid plains. It must not be forgotten, however,
that there would almost certainly be shallow lakes and swamps, when these

1 Records, XXI, 158, (1888). |] 2 Supra, p. 425.
212
486 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS [Chap, XVIL

i ‘
deposits were being formed, and at’ the present day the river valleys of
Tibet, even at a height of over 13,000 feet, can under such circumstances
support a growth of grass and shrubs which could easily have given sus-
tenance to the rhinoceros of Hundes. This animal in any case shews that
the climate of Hundes must have been somewhat milder than it now is, and
as there is little difficulty in supposing that these deposits may have been
raised 1,000 to 2,000 feet without any appreciable disturbance, though it is
impossible to grant an elevation of 10,000 to 15,000 feet, we may well sup-
pose that this increased inclemency of climate is partly due to the desicca-
tion resulting from the change of condition of the rivers, from deposition to
erosion, and partly to an increased altitude of the plains and of the moun-
tains south of them.

The gradual desiccation of the Tibetan lakes points to the same con-
clusion. There are no data available regarding the rate at which this is
taking place, but the fact that some have dried completely up, while others
show but little reduction on their original size, indicates that the process is
still in progress and that the climate of Tibet was once moister than it
now is. There appears to be but one explanation possible of this increased
dryness of climate, and that is a rise of the mountains to the south, which
has resulted in a more complete cutting off of the moisture from the mon-
soon winds.

The cause of the origin of these lakes in Tibet is not thoroughly
established. Ever since the publication of Mr. Drew’s book on the Jammu
and Kashmir territories it has been customary to attribute their origin to
the damming up of the main valleys by the fans of tributaries, which
attained a great development during the glacial periods, when the disinte-
gration of the rocks was more rapid than it now is, while the transporting
power of the streams was no greater if so great. The present writer is
unable-to accept this view in its entirety. In the case of the Pangong lake
he believes that its formation is entirely due to differential movements of
the surface, which raised a portion of the original river bed at a more
rapid rate than the stream was able to erode.and dammed back the drainage
to produce the present lake. Even in the case of the Tsomoriri in Rupshu,
which is accepted as the typical instance of a lake formed by a tributary
fan, he has shown! that there is reason to believe that this fan could not
have caused an interruption of the drainage had there not been an eleva-
tion of a portion of the river valley further down its course, and a conse-
quent diminution of the gradient. Whatever may be the cause of origin of
these lakes there seems no reason to doubt that the broad shingle plains,
which so frequently occur just above where the rivers enter a gorge, are
produced by a check in the gradient consequent on a recent elevation of
the river bed in the gorge, and consequent checking of the gradient im-

1 Records, XXI, 156, (188),
Chap. XVIIT.1 GROWTH NOT COMPLETED. 487

mediately above it, A similar action might well, under favourable circum-
stances, give rise to the formation of an actual lake, while the existence of
an exit would depend on the rapidity of the movement, the supply of
water, and the nature of the climate. Whether such has actually been the
case or not, there have certainly been irregular movements of the beds of
the streams and rivers within what is, geologically speaking, a very recent
period, and these irregular movements can only be regarded as evidence
that the disturbance which caused the elevation of the Himdlayas is still in
progress,

Thus, from whatever point of view we look at the subject, we see that
the decadence of the Himalayas has not yet begun, but whether they have
yet reached their maximum development is not so clear, There are no
data from which we can decide whether the rate of elevation in the imme-
diate past has been greater than at present or no, but looking to the
general indications, throughout the world, that the great earth movements,
which caused such profound changes in the form and distribution of
land and sea during the tertiary period, have reached their close, and that
the present is a period of comparative quiescence in the history of the
earth, we may suppose that the chapter devoted to the elevation of the
Himdlayas is reaching its close and that they soon will enter on their
decay.

There remains one more point to be referred to before finally dismissing
the subject of the origin of the Himélayas, and that is the supposed con-
nection between mountain ranges and sedimentation. The enormous
thickness of sedimentary deposits seen in the sections exposed in moun-
tain ranges. has been frequently noticed, and by many observers their ac-
cumulation has been regarded as the immediate precedent, and proximate
cause, of the mountain ranges. It is, however, doubtful whether in this.
case cause and effect have not been confused. Sedimentary accumulations
of great thickness are known elsewhere than- in mountain ranges, but itis
only where the beds have been turned up at steep angles and extensively
denuded that their thickness becomes conspicuous, and it is only where a
great thickness of sediment has been previously accumulated that moun-
tains can be formed of stratified deposits. Otherwise the underlying
crystalline and metamorphic rocks will soon be exposed by the denudation
which is always much mere active in mountain ranges than in more level
ground. The subject is, however, of sufficient importance and interest to
make it necessary to inquire whether there is any indication of a con-
nection between the present position of the Himdlayas and the distribution
of the sedimentary deposits which preceded its elevation.

In the north-west Himdlayas there is a great series of sedimentary

  
488 GEOLOGY OF INDIA—AGE AND ORIGIN OF HIMALAYAS [Chap. XVIII.

deposits, ranging through the palzozoic and mesozoic eras, which represent
long periods of accumulation of sediment in enormous thicknesses of strata,
Here there is a distinct temptation to regard the mountain range as the
result of this vast accumulation of stratified deposits, but 4s we trace the
range to the eastwards difficulties come in.

The zone of marine deposits found north of the line of highest peaks
in that part of the Himdlayas which has been accessible to exploration
has been referred to, as well as the non-recognition of these beds south of
the main range. How far the same distribution holds good further east
it is impossible to say with certainty, but we know that jurassic and
cretaceous fossils have been obtained from the region north of the hills of
Nepal and Sikkim and the discovery of sedimentary strata of unknowa,
but probably tertiary, age near the Cholamo lakes seems to indicate that
the same parallelism between the boundary of the sedimentary deposits
and the line of highest peaks prevails at least as far east as Sikkim,
Whether the present limit is in the main due to an original limit of de-
position or to the effects of disturbance and denudation is for the moment
unimportant. ‘The absence from the main range and the hills to the south
of them, so far as is known, of any extensive series of sedimentary strata
later than older palzozoic or even older, precludes the idea that the eleva-
tion of the range was immediately consequent on a great accumulation of
strata.

In the eastern Himdlayas our difficulties are still very great owing to
the scanty observations available. The only sedimentary deposits that
could possibly be marine, or that have any great thickness, are certainly
long anterior to the carboniferous in age, and these occupy a very
small area in comparison with the great expanse of crystalline schists,
gneisses and granites. But there are some small patches of coal bearing
Damuda rocks, which have been recognised at several spots along the
outer edge of the Himélayas, and are important as showing that this re-
gion was dry land, at the close of the palzeozoic era, when marine forma-
tions many thousands of feet in thickness were being deposited in the
north-west. It is not pussible to say that no marine strata of later date
than permian exist in the eastern Himalayas, but it may be taken as
tolerably certain that, if present, they cannot be of a very great extent
or thickness, and this portion of the Him4layas appears to have been a
land area continuous with that of the Peninsula throughout the secondary
era, such interruptions of continuity as there may have been, if there were
any at all, being of minor importance and only temporary, But though this
portion of the Himalayas was a land area, there is no reason for supposing
it was a mountain range at these early periods; the great height of the
snowy peaks suggests that their upheaval must have been comparatively
recent, and the palpable unity of the range as a whole prevents us from
Chap. XVIII NOT DUE TO SEDIMENTATION, 489

ascribing a much earlier date to this portion than to the rest, which it has
been shown could not have existed in its present form in the secondary
era.

In view of this divergence between the eastern and western portions
of the range, it is impossible to attribute the rise of the Himalayas to the
~ sedimentation in what is now its north-western portion, and we must look
to some more wide reaching and deep seated cause for its present position
and course—a cause which was independent of and able to obliterate long
standing structural features and to introduce new lines of separation
between areas of elevation and subsidence.
CHAPTER XIX.

2

GEOLOGICAL HISTORY OF THE INDIAN PENINSULA,

Earliest periods—Origin of Ardvallis and East Coast— Mesozoic Indo-African continent—Origin
of the West Coast and Western Ghats.

The previous chapters of this book have principally been devoted to the
stratigraphical description of the various rock systems of India, and though
reference has been made in the course of this description to changes in
the distribution of land and sea, and to the earth movements which have
marked out the salient features of Indian geography, such references have
necessarily been somewhat swamped by other matter. This chapter will
consequently be devoted to a brief resumé of the geological history of
India, of those changes of land and sea through which it has reached its
present form.

The earliest stages of the geological history of India, as of all other
history, are wrapped in obscurity. Dimly we can discern an old land
surface composed partly of a very ancient granitoid rock, which had even
then solidified, been penetrated with quartz veins and trap dykes, and
exposed to extensive denudation, and partly of later rocks, themselves the
product of the denudation of the granitoid gneiss. From the waste of this
Jand surface the rocks of the NDhdrwdr system were formed, in a sea where
volcanoes poured forth their lavas and ashes, much as at the present time.
But whether any living thing was to be found in this sea, or whether the
earth was still unfit for the support of either animal or vegetable life, it is
impossible to say.

These Dharwdr deposits were in their turn compressed, contorted and
exposed to great denudation before the commencement of the Cuddapah
epoch, but it is impossible to trace even approximately the changes of dis-
tribution of land and sea during this earliest period of the geological
history of India.

With the commencement of the Cuddapah epoch, some definite indica-
tion of the distribution of land and seaappear. All Southern India, south and
west of the Cuddapah and Kalddgi basins appears to have been dry land,
while the sea spread out to the east over part of the present Bay of Bengal,and,
to the north over what are now the Nizam's dominions and the Central Prov-
Chap. XTX.] ELEVATION OF THE ARAVALLIS. 491

inces. The exact limits of this sea cannot be defined with accuracy, western
Bengal and Chutié Nagpur were probably dry land, and this rock area
probably stretched to the north-east over the Gangetic delta, to Assam and
the eastern Himalayas, In Bundelkhand there was dry land, to the south
of which the Bijd4war sea spread to the valleys of the Narbad4 and the
Son but had probably been obliterated by the time the Cuddapahs werc
deposited, and at a later period a fresh depression admitted the sea.to the
north-west of Bundelkhand, in which the beds of the Gwalior system were
deposited.

Nothing is known of the early geological history of the great area
covered by the Deccan trap, nor of what was going on where the Himé-
layas now stand, or where the Indus and Ganges rivers have spread their
alluvial plains. In fact, what with complete want of information regarding
the greater portion of the area, and the incompleteness of that available
regarding the rest, the conclusions that can be drawn regarding these
- earliest periods of the geological history of India are of the most meagre
description. This much, however, seems certain that none of the leading
features of Indian geography of the present day had been marked out, none
of the mountain ranges had arisea, none of the great river valleys had com-
menced, and the distribution of land and sea was very different to what we
now see.

The close of the Cuddapah epoch appears to have witnessed the com-
mencement of the earliest of those earth movements whose effects on the
surface contours and geography of India are still prominently noticeable.
It was then that the great mountain range, of which the present Ardvallis
are but the wreck, was raised, and extending far beyond its present
limits, stretched across what is now the Gangetic plain, possibly even to
the Himalayas. Atthe same time another zone of contortion was formed
running along the south side of the Son and Narbadé valleys, which was
probably marked by a range of mountains or hills, not rising to the same
height or importance. as the Ardvallis, and bearing much the same relation
to them as the hills west of the Indus alluvial plain do to the Himdlayas
of the present day.

To the same date must probably be ascribed the zone of contortion
which runs along the eastern margin of the Cuddapah basin and can be
traced northwards tc the Godavari valley.

These three zones of contortion, whose disturbance took place during
the Vindhyan epoch, and must once have been marked by mountain ranges
much more important in size and. elevation than their remnants at the
present day, seem to be due to the last great movement cf compression
which has affected the rocks of the Peninsula, Since then the disturb-
ances have principally taken the form of movements of elevation and sub-
492 GEOLOGY OF INDIA—-RESUME, [Chap. XIX,

sidence, only to a comparatively minor extent accompanied by compres-
sion of the rocks, and it is interesting to note that the earth movements of
this period have still their influence on the limitation of the Peninsula.

On the north-west the Advallis have remained the boundary of the
peninsular land area. West of them the great desert of western Raj-
putdna was alternately land and sea through long ages, but the sea never
spread east of the barrier of the Ardvallis, On the south-east the bend
ot the east coast north of Madras follows too closely the general course
of the Nallamalai range for the connection to be accidental and as we
know that from the jurassic period to the present day the position of the
coast has been practically where it now lies, we may naturally conclude
that its course had been laid down at an even earlier period, contempo-
raneously with the great Vindhyan epoch of disturbance. In the course of
ages there have no doubt been alternate elevations and depressions of the
land, at times it has encroached on the sea, at times the sea has flowed
over what is now dry land. But the fact that the only marine deposits -
in this part of India are confined to the neighbourhood of the coast, their
small thickness, the manner in which they thin out away from the sea,
and the character of the rocks, indicate that when they were formed the
shore line could not have been far off, and point to a persistence of

the general run and position of this the oldest feature of the geography
of India.

The Vindhyan epoch is the age to which the rise of the Ardvallisand
the demarcation of the east coast has been ascribed, but what this age is
in terms of the European sequence there is no means of determining. The
upper Vindtyans have been looked upon as devonian, on the strength of
their resemblance to the Table Mountain sandstone of South Africa, and
though the evidence is insufficient, it is certain that they cannot be much
newer than the date indicated, and it seems difficult to make them much
older. They may consequently be ascribed to some portion of the middle
or latter end of the lower palzozoic, and itis to be regretted that no
more exact correlation can be made, for we find that in silurian times the
sea flowed over the north-west Punjab and the north-western portion
central Himdlayas, and over the hills east of the Irawadi valley. No
silurian deposits have been found in that small portion of the eastern
Himdlayas which has been examined, nor in Assam, and if is probable
that the land area strctched north eastwards from the Peninsula over

these regions in silurian times, as it seems certainly to have done at a later
period.

Towards the close of the palzozoic era, at a period corresponding
to the upper carboniferous of European chronology, we have some definite
information regarding not only the distribution of land and sea, but also
Chap, XIX, ] MESOZOIC INDO-AFRICAN CONTINENT. 493

the climate. The great Gondwana era opened with a period of exception-
al cold. The Peninsula was a land area over which many large lakes were
probably scattered, while on land there were glaciers flowing down
into these lakes, and into the sea which covered part of the great [ndian
desert, the north-west Punjab, and a very any large portion, if not the
whole, of the area occupied by the Himdlayas west of the Ganges valley,
The same sea appears to have stretched westwards to the furthest boun-
dary of Afghdnistén, and it was continuous in some way with that
which flowed over eastern Australia. It is not clear whether this com-
munication was round the south and west of the Peninsula or round the
east and north. We know from the evidence of the Salt range fossils
that after the glacial period there was an irruption of European forms and
a complete change of type of the fauna; this period was one of extensive
changes of land and sea when vast areas in South Africa and Australia
were converted into dry land, there is consequently a possibility that sea
stretched south of the Indian peninsula and the close affinity of the
two faunas is more in favour of this direct communication, than of one
round by the more circuitous route round the north of the peninsular
area, which seems at that time to have extended much further to the north-
east than it now does. At the close of the jurassic period the Indian
peninsula was still dry land, the east coast was not very far removed
from its present position, and on the west the sea flowed over Cutch, the
Indian desert and the north-west Punjab and central Himdlayas. It is
not possible to say whether the north-easterly extension of the peninsular
area over Assam was still dry Jand, but the land connection with Africa was
still maintained. Still the presence of some eastern species in the western
sea shows that there must have been tither a temporary and direct,
or more permanent and circuitous, connection between the two. If
the first of these explanations is the correct one there may have been a
temporary subsidence, by which the land communication between Africa
and India may have been severed for a time, and certain forms of life
enabled to cross from one marine province to the other. The alter-
native explanation would be that the form which is common to the two
areas, being an abundant and wide ranging one, was endowed with great
powers of spreading, and reached the western sea round the north-easterly
prolongation of the Indian peninsular land area.

In the cretaceous period the land connection with Africa was still
maintained, the eastern coast line of the continent ran not very far from
the present east coast of India, across the Ganges delta,’ and along the
south side of the Assam hills. On the west of India a different sea flowed
over Arabia and the Arabian Sea and extended inland at least as far as
Bdrwai on the Narbada. Sea also flowed over the hilly country west of

the Indus alluvium and over Tibet.
494 GOLOGY OF INDIA—RESUME. (Chap. XIX,

The close of the cretaceous period witnessed that grea’ outburst of
volcanic activity which buried the whole of western India deep in lavas
and ashes, and extended from Sind on the one hand to RdéjAmahendri on
the other. It isnot improbable that this great outburst may have been con-
nected, as it was probably contemporaneous, with the commencement of
that great series of earth movements which resulted in the elevation of the
Himalayas and the extra peninsular mountain ranges generally. But how-
ever this may be, the lava flows must have obliterated all the pre-existing
surface features and the origin of the main features of the drainage system,
of the northern part of the Peninsula at least, cannot be ascribed to an
earlier date than the close of the Deccan trap period.

In the tertiary era we find no further evidence of a land connection
with Africa; at an early period the west coast was approximately in its
present position, and it is probable that at the clese of the cretaceous or
commencement of the eocene period the great Indo-African continent was
finally broken up, and all but the remnants in India and South Africa sunk
linally beneath the sea.

The eocene sea flowed over western Rajputdna and the Indus valley
to the west, over a large part of Baldchistén and Afghdnistan, and over
the whole of the north-west Punjab and the outer Himélayas as far east
as the Ganges river. We do not know if this sea stretched eastwards to
the north of the Peninsula till it joined that in which the nummulitics of
Assam and Burma were deposited, but on the whole it more probably did
not. Sea also flowed over the central Himdlayas and was probably con-
tinuous with that just referred to, across the north-western termination
portion of the range,

One of the first effects of the great series of earth movements, which
resulted in the formation of the mountain ranges of extra peninsular India,
was an encroachment of land on sea, and the driving back of the sea first
from the Himalayan and Punjab areas, and finally from Sind and Burma,
The same period as witnessed the gradual growth of the Himalayas also
saw the rise of the Arakan Yoma, and Manipur and Naga hills, on the one
hand, and the greater part at any rate of the Afghdnistan and Baldchistén
hills on the other, and as the most important part of this. history has been
told in the last chapter it will not be necessary to repeat it here.

It would have been in the last degree extraordinary, if such extensive
and violent earth: movements all around it had been accompanied by
absolute quiescence in the Indian peninsula, but such disturbance as may
have taken place in no way took the form of compression, and the only
change which can be attributed to this period is the origin of the Western
Ghats. Reference was made in the first chapter to the difficulty of ac-
counting for this feature, and its resemblance to a line of sea cliffs modi-:
Chap. XTX.] ORIGIN OF THE WESTERN GHATS, 495

fied by subaerial denudation was noticed, as well as the occurrence of a
land shell closely allied to a marine form. But though the sea may once
have washed the foot of the ghéts, it is impossible to grant that they owe
their origin entirely, or even largely, to marine denudation. Marine
denudation works slowly on hard rocks, and during the ages that would have
been required for the sea to carve the low lands of the Konkan out of the
Deccan trap, it is inconceivable that the rivers would not have cut their
valleys much further back into the scarp than they have done. It is far
more probable that the main features are due to late tertiary earth move-
ments, and the great rock basins of the Narbad4 and T4pti valleys show
most conclusively that there has been a movement of elevation to the west,
which certainly checked and may even for a time have interrupted the
flow of those rivers, while the ground along the foot of the ghdts has not
been closely enough examined either to prove or disprove the hypothesis.

Whatever may have been the cause of the origin of the Western Ghats,
the present easterly tread of the peninsular drainage must be an ancient
one, for had there been any considerable rivers flowing to the west they
would have preserved their channels, or if the movement had been suffi-
ciently rapid to reverse the course of the drainage, deep gaps would have
been left to mark their former course. . There is only one such gap, the
Palghat, north of the Travancore hills, and it is possible that a river may
once have flowed westwards through this, whose drainage was reversed by
the earth movements which raised the Western Ghats, leaving the lower
part of its course to be occupied by a much smaller stream, while the bulk
of the drainage was diverted to the east. With this possible exception it
is probable that the main features of the peninsular drainage, the two great
westerly flowing rivers to the north, and the series of easterly flowing ones
further south, were marked out at the close of the Deccan trap period.

We see then that the origin of the west coast of India dates from the
middle of the tertiary epoch or a little earlier, when the dry land which
stretched westwards into the Arabian Sea was depressed, and at the same
time that to the east was elevated to form the Western Ghats, the most
recent and also, perhaps more correctly therefore, the most conspicuous
feature in the geography of the Indian Peninsula.
 

a 78

 

  
 
     
      

re 8
PO AMIN Thi Ml, fn
so os NY

 
 
  
 
 
 

 

es .
+ rf Virling P sl = |
: Darigige Mant bg OPT \ 3
T22PTOS

 

 

E
&
*

 

Malian FP,

 

= re

 

eg

pang h

        

tion gee :
i 7 ase | 16650
: Zon, 4 ais 2 , fuga ged
i / Al a Cae My aregik? SEEN ee Z
. v5 { | —<ig8 aN 2 e é "ang YNC000% oAidunalds — (Chahidsela
PGasings Wisin [Pgh yet) * Se Pe a ee
«J in Ohi pe None Q.
ay oS y [epee Re
I x

  
 
 

         

ha | Yn ORM, Soe
oo tf ee Feat

     
      
 

     

 

 

 
 
   

 
  
  

  
        
  
  
  
   

Z
‘ eee te hi ve an shlasheLanyary,
; bedi tte b.] EMA aE TO) The *
DirTegets Ft OO TNE Pn a Lee ae ; “phe :
Lg at Lean Portarrith Pm emer | Ww Karatogh taco
| soa e My
Pe nf } ¥ 1 We9 Ban P
| Bela) | Sp ‘hi
| |o¢ 9 fe OMA :
late \ @ Sa Le 2 hy
|. & ~
/ ga Tatdat
| ef Care iso0o00" —
An
:
ge

— =
|W heakeoser

Pakuie

   
  

~
~ inesoPLings: Thang

 
 
 
 
 

     

 

sChuruk Lakrriv

 

       
  

PTH Hy

     
  
 

  
   
    
      

e800
° rf tO oma

°SRINAGAL

Whhasliruee

  

*Klurrial

 

. <
er
: , * Ze £
No LOG yar ee Shae Prag, << O

cto Ia

eae ‘ off
Audok = Dp TN Myr
4 UC OT My

eleanin  Saniang Co a

The name Dha'ola-
Ahayr has beer mise
placed on this map.

Lhe Dhwoladhur

   

Dome Road
hoowval

  
  
   
  
 

Yh.) == Fao

7) Churkangy

   
 
 
 
 

    

JAMOO_ =F

  
   
    

 
   
      
 
  

is the range running ee a :
south-eastward From ‘ eee Hn
Dalhousie. as gn sap

 

 

fp Rang
— ae irripoor

   

 
 

alla
Moolaria “,

 
              
  
 
 
  
    
 
 
 
 

se

/
6
Bahadrun

le vaqaon AdF,
; i PE TB
$ ' Gheuo Bawamkhent ef ] "
pie 8 reat ,
Mahajan. Sain fe
2 gous Stour ‘a (
: Kumara Shimla gran )
Kajasar Shad / oRaigurh
ays ay ads haikhsar / Chala j
‘ar’ 3
darshahre thansar
y eet ees
Pholerao MotSrcn =a och

 
 
  

Bitaodi B
Sona

  

 

 

   
 
 

   
       
 
 
 
   
     

  
     
 

 

ie ‘\ 3
a BICKANEER Ratangarho pat, g- Pubvul .
F Badno, Plo? Pra s ‘ yp BES »lyara ae f
/ Deshnek 9 : : 3 /FeRoxpoor 4
aac | Serge, 5 ataenety raat} EEimann) Spore SNe
Jasrasar: ea fj. re Mana Swick
a ° Qo BY
i peat (Sipe
Diddwapa rp Mu 78 Wee
AUP S YS BP Akachinangiih Bhaing AW
Naé ° Chapha ine ets phen
"| 3 eT khanth
o8ANTT raat Auchamai, oR POMEL or a
Mundwas | °Parod Mariteg Sere *
Aacherul Manjio ° ie ER
ur P Ber amar Baebes <ilpgph BAR
7 = J, I ere S
iSoia “es. lien g
é

  
 
   
  
 
     

Kas

Nerte oud Jo
°

», Merta

lag ne Lait
"8 KIMEREE
vl’ XIMERES
fisangan

Jean

 

 

i yh <_ PUbundty

 

 

 

 

 

 

Reg. No. 198, Geol. Sur.—July 92.—1,500.
 

 
     
    
   

 

|
i ¥ \
| Yosu hun vat
ae . lg hun \ | \
VN fSorghak (Cold rield' '
7050 7
|
|

ee f |

‘
ate

       
  

  

Sulyteet ¥

\
Chubolikec 18260 |

—_——/

   

  
 
   
  
 
  
 
     
      
  
       
   
    

 

 

   

k /
7
‘ | |
|
; | Barkhalu® | |
eas | \
& | \
Neston | |
_ Aa 8)ep | \
°° “Dak Nak * | .
mh pres |
aQhobake 9 70D
umik Lakmo . | H I M A A Y A Me 0 U ]
D0 |
\
| |
| |
|
$ |
: | = 1 eer
7 Seale 1 Inch = 64 Miles or szochoa5
ye : | 100 90 80 70 €0 50 40 30 20 10 O 100
Sue | to Es
ah deg
0 TN
the = : \
ney |
|
Taal Herd |
Clea. 7 |
p'salt min ofhok Rakung ensze 7 Chana : rong Crake,
iiiold ratd ox ‘Spx Lake
4 Nawal Chaka .

okt Lake
hy BLO Aas Sramaons Onaks
2 y % ‘ Bonaye naka =" Baeod
A ron ‘€ Gale Laks : ae
Ca oom TNE Bruup Cho | -
: aeMliok Daurakpa | hE SE Sian,

 

     

Inara Cho Lake

  

eee
Shubak the 0

350 / ‘* =
FE ujang a j [Damen
S ¥

    
 
 

 

 

>
A

7 kiachan
x = Ruidapehyy! Monaster
ie ly ey -" Gon istand?
Kingloy ~~ Ghaiarmg Cho

 

oKhala S.P* =

  

Wamcho (Lake)

Teng Nur 7}
Lake ST ‘

nn

   
   
     
   
    
   
   
 
  

  

Khalamba E+

(oe Cuyja

Cc H Naming gy

12220/

 

  
    
    
   
 
   
        
   

ok -Cro 11380
or Fiahte or Palte Lake} /*3

 
 
  

 

 

Nee
mn ;
Miekyn Z
TLhochunuhyai
i100

  

     

aa

err
TAME MMs
Cho Lake

aX

     
  
 

iy
jkhamba Jong

ong Lake — ov

    

  

Z\ \Romeho Vatke |

  

    
 
    
   
 
     

Lo Neg ag, uy)
Ne if ~y Peale
\ . By Bs Mowy Peale
= j 2 Niall? ie
‘ 0,

 
   

 
 

Ye, Sutesar

 
 
 
 

 silligors She
Nicaapack ‘ Miuriagetee
\

Julpigthes) ° ap

 
    

 

 

 

 

   
 

  
  

 

—
cal 9\B°
: fe ge

—
oN + --

 

 

(Second Edition)
1891

Seale 1 Inch -96 Miles

6 SETUNSCNEAT:
100

  
   

   

80 60

  

40

  

 

y = =]
B00

 

y

  
 

Chien Chuan Chon
©

     
    
  
 
  
  
 
  
     

Atvncaig oy Kanne ' > /
a \/

Picmpanh aif
The age of the rocks in the area’
soloured t yn is uncertain. a re
c fl 4 $ ( oChuntan Sr ang Chivnepltaen
e e yy ) e
ie gtchow
(firme 4s a4 4 y longtchow Luke Erhte
ae e os) ieMines x, ) ‘tite
7 Kyayuhsenadwite (ta » Ee \ e ‘api a es =
\ § Yung Bingo
°
sd

Papok J £ eyahs \ . Sy
tf e l } ‘a onytog >. :
A fi rapa \ {

‘ r Tes hook

[ioubary J e !

fomnhin Montes a ge | |
: Sieed ele \ Rogcaoe

) Slaaunng to? o Kini" matoo

\

  

Hahn of OF B51

Ethene S y= ; 4
y

“, down ‘
i oKaxanks gon ey, od l
/onah :
Roe ss re aa - ) ee Ye ine

      
  

Kinng me

~~ \. Kiang, %
k yy
Sy. ¥

Chiengr
oMane-that ny?

\Mane ©

+ Makdane’@

PanyPoor,

‘ Buy Point
ous fossils have been Y
ured brown. The age . "
ris unknown. | Gwa Khyoq

Maingiyey, Woig ent long
si ) Tony Rewer \

ae mae

 

  
  
 
  
   
  
 
 

lenieng Rang

  
 
  

   

 

 

 

 

—
 

 

 

 

 

 

 

 

2

 

Venginerda)

U

Me a

The Malabar Coast from
chiefly cornposed of laterit
South of Malwan are belie
Laterite is omitted on this

Ratnagiri Southwards is
e. The underlying rocks

ved to be ioetaroorphic.
map.

 

pip
J
an
et Bassas de Pedro
Maaryal pu
al
“
- _ TBS OIC AS ay Ap ae
aq  V—_——_
co “ tattore'
ANTOIPY
I

 

* Nine Degree Channel

_ A sinall exposure of
tower tertiary rocks
near Quilon,

Minicov 1. G

      
    

 

 

 

re | s 2
| | Le :
| \ — a 2
| SS,
\ | ’
Tropic of Cancer | %.
| 4,
} Ce
!
|
j |
|
| |
| |
|
|
|
|
if
| |
fi20"
| es
| Sete
SS |
a !
es
|
|
|
|
|
|
|
|
| S
o |
16 He |
aoa
|
|
|
|
|
!
! |
|
12"
Ae Fn:
De eee ee
{
|
!
INDEX.
|
EXTRA PENINSULAR. PENINSULAR. |
Recent and Subrecent Recent and Subrecent
Upper Tertiary (Siwalik) ser ee Upper Tertiary
i Oy. Lower Tertiary een Lower Tertiary
Eocene voicanic of Heeoan tra
Himalayas P
i
Cretaceous Cretaceous
Jurassic [eae] Upper Gondwana
Carboniferous to Trias Eee Lower Gondwana
Cambro-Silurian Upper
pyndayant Older Paleozoic
(Cuddapah)
Transition systems ea Transition systems
Crystalline (Gneiss, Crystalline (Gneiss,
Tanite, &c.) Tanite, &c.)
£ I N
6\6° 70°

 

Reg. No. 10, Dir., Geol. Survey.—Mar, 93 ~1,520.

Eight Degree Channe]

 

 

 

 

 
  
  
  
   
  
   

ol) MADRAS
Bi Hort SU George

alleen

 

 

Be p False Pou
~~ S
Yah annddy

B

oo

OF

 

‘

$1 pfartinn ISN \
wy

aN

Combat

fo

 

a Cretaceous fossils hav
a coloured brown. 7
erally is unknowr.

Triassic aa

 

 

Cor

j

North Awvawat (
Ez

Interview If} a

ne

Port trakanun 4

ANDAMAN &

  

 

  
   

Trengiiebaar

Wardha

Niypapatiun

Cc

°

e
v7
a
1s LAN ps
a :
| . 4 - 7 South Andsanios =
| Lort Mout, |
North Saniinel~

Trinkmnlig

 

ae =<

    

 

 
 
 
 

COLOMBO

e

1G.
¥ e oo

A N

EY

4

7379 Y, .

aU

a
‘5 PE
ae /
He -
po r

be

fa a

sweat Kasses

 

The Island of Ceylon
is principally composed
of metamorphic gneiss.
Subrecent limestone
occurs in the low
ground of the North of
the Island.

C2

  
 

Rutland 1. 3
J

Dune an Pa

a ed ect cee :
Intrusive serpentine 18 largely
developed in Rutland and Little @,

Andaman Islands.

Teu Degr

Kr: Nicobut

 

 

xr1CcOoOBAR

1SLAND S

 

 

 

 

 

 

 

 

 

 

 

 
morphic TOCsS:*
ferous sedime
stone of caToor

  
 

mein.
Coal of Teraary and
ae an the Tenas

Thereis muca
the islandas ©

 

gree

|
|
Channel |
|
|
'

’ Little Nicola a

&

DPalc War
Po Bras eS

“Ae Ackan Me

serim Walley

<oh lpmestone of UBEROWT
¢ the Mere g
=

Gweas

 
    
   
  
 
 
  

   

Keds Sect

t

 

 

 

 

 

 

 

 

 

 

 

Titho., 8. 1. 0., Calcutta.

 

 
GEOGRAPHICAL INDEX

OF

INDIAN LOCALITIES.

Latitudes are all north ; Longitudes east of Greenwich.

Abbreviations used: —H. Hill or mountain; R, River or stream; I. Island; L. Lake.

In the case of rivers whose names are printed on the accompanying map, the coordinates
indicate the position of the name. Where no name is inserted on the map, the coordinates are
those of a point on its course which will indicate the locality referred to in the text.

The coordinates of States or Districts are usually those of the capital or head-quarters;
where this course could not be followed and in the case of hill ranges, a central point has
been selected.

Casual references or the use of the name of one place to indicate the position of another
have not been indexed, except where such are the only refreences in the text.

Presidencies and Provinces have not been indexed.

 

 

 

Latitude, Longitude, Page.

A o t oO t
Abbottabad . . . . ‘| 34 0 73 16 | 116, 138, 286, 352.
Abi H.. 3 i : . -| 24 36 72 45 | 14, 40.
Abur. ; ; . . -| 27° «5 7o 37 | 228,
Adjai R. ‘i ‘ A ‘i «| 23 35 87 25 | 444.
Afridi H. : . : . *| 33 45 71 35 | 352.
Agani R. . ‘ . * -| 16 33 76 35 | 88.
Agori . ‘ ; ; : -| 24 33 83 2 | 56.

Agra. s 5 . . -| 27 Io 78 5 | 432,435,448, 476.
Ahmadabad . , ‘ , -| 23 72 30 | 407, 408, 412.
Ahmadnagar - . ; . -| I9 5 74 55 | 402.

Aka H. . é : i ‘. -| 27° «0 92 45 | 149.
Akauktaung . 2 : ; -| 18 30 95 11 | 337.

Akbarpur_ . ‘ 7 -| 25 13 80 51 | 97.

Akola’. ‘ . 5 : «| 20 42 77 2 =| 401.

Aksapur ‘ ; : . | IQ 21 79 28 | go.

t

 

 

 

 

 
498

GEOLOGY OF INDIA.

 

 

 

Albaka .

Aleppi .

Ali : i : ‘
Allahabdd =, ‘ ‘
Almod

Alwar

Amarkantak .

Amb. 7 ‘
Amerumbode

Amraoti

Amri

Atnru H. :

Anamalai ,
Anantapur. .
Andaman I. 2 ‘
Angami Naga H. .
Anthi .

Antri. - Fs ji
Arakan . : ‘
Arakan Yoma

Arakere . :

Aravalli. °

Arcot . . . ,
Ariyalir :

Api «= a w %
Asfrgarh r ; .
Athgarh ; . ;
AttaranR. . ‘
Attock .

Auk

Ava ‘ .

Babeh . ‘é é
Bachao. j ¥ ;

 

Latitude, Longitude, Page,
° I °o !
18 13 80 44 | 89, 402.
9 30 76 23 | 405.
F 22 11 74 24 | 253.
25 26 | 81 55 | 427, 446, 450.
22 2% 78 26 | 170.
‘ 27 34 76 38 | 70.
7 22 40 81 48 ‘| 5, 263, 374, 383.
34 19 72, 55 | 419.
12 24 80 gs} 389.
20 56 | 77 47 | 396, 4ot.
26 II 68 4 | 286,
27 «18 67 21 | 311.
. “10 20 77. «0 «| 4414.
i Sa 14 41 77° 39 | 33, 36, 40.
12 30 93 O | 12, 343.
25 40 94 10 | 335-
27 32 74 35
r 26 3 78 16 | 105.
20 7 92 50 | 11, 20, 143, 338340.
18 30 | 95 © | 8, 143, 297, 494.
16 17 | 75 32 | 84
26 30 | 75 © | 6, 40, 68, 73, 94, To4, 106,
, 483, 491.
12 55 | 79 24 | 33,39. 40, 78.
. 11 8 | 79 7 | 239, 243.
27. 39 | 68 59 | 451.
é 21 28 76 20 | 5, 308.
. 20 31 85 41 | 178.
. 16 8 98 2 142.
33 53 72 17 = | 419.
15 12 78 10 | 85.
2t 52 | 96 1 | 45, 341.
. 3143 | 78 4 | 42, 484.
: -| 23° 17 7o 23 | 223.

 

 

 

 
GEUGRAPHICAL INDEX.

 

 

 

 

 

 

 

499
Latitude, Longitude. Paye,
. f ° ’
Bachimali . . e : .| 16 29 76 42 | 88.
Badalgarh «2s . : .| 20 53 77, «18 | 70.
Badhano : ‘ ‘ ‘ .| 26 7 78 24 } 65.
Badnera : j : . .| 20 52 77. 46 | 4o1.
Badnir . . . 5 .| 21 54 77, «57—~«| «167.
Bagh . < ‘ « : .| 22 24 74 52 | 54, 211, 248.
Bagpura . . ‘ i .| 24 32 79 14 | 28.
Bagra . ‘ _ ; ‘ .| 22 37 7 3.>«|:«a173»
Bahadur Khel ‘ ‘i , .| 33 10 jo §9 | 326.
Bakdwalpur . ; é ‘ .| 29 24 q1 47 | 428, 452.
Bahir . : : ‘i - .| it 56 79 52 | 404.
Bairenkonda . . i : .| 15 38 79 3 | 81.
Balkh . F F 5 , .| 36 45 66 56 | 204, 325.
Edlmer . : ‘ ‘ j »| 25° 45 125, | 74, 227, 455-
Baltistan s i A : .| 35 12 75 35 | 149.
Banaganpalhi 7 . ‘ | 5 15 78 20 | 85; 87, 107.
Band Vero. : 5 ‘ .| 25 34 68 8 | 303.
BangalR. .. i : | 30 45 77. «45~=«|:«LW‘YT7.
Bangalore. : é : .| 12 58 77. 37:«| «48.
Bannu . 7 ‘i e . -| 33 0 qo 39 | 326, 418.
Bap. ss i ‘ : .| 27 22 | 72 23 | 106, 160.
Baplaimali H. ‘ ‘ : Ig 213 83. 2 | 370
Baa. ¢ 6 wh a wf) GO 6] 98 ao]
Barak R. 4 ‘ : : .| 24 25 gt 2) | 2.
Barakhar R. . , ‘ ‘ -| 24° 5 86 20 ; 164.
Bardhi. . . «  «  ~| 24 33 | 82 27 | 98
Bardwan 5 : ‘ 23 «14 87 54 | 166, 378, 427.
Barhi .. 3 : ‘ ‘ .| 24 18 85 29 | 58.
Bari H. . 5 ‘ ‘ : .| 78 28 22 40 | 55
Baroda . 2 3 .| 22 217 73 16 | 73, 249, 259; 262, 278, 407.
Barmuri : : ‘ ‘ “1 23° 45 86 52 | 437-
Basil. « .» « «| #@ § | G7 56 ae,
Barren I. : ; ‘ : .{ 12 12 93 50 | 16.
Bérwai. . se | 22 5 | 76 6 | 53, 152) 1745 249s 396.
Barwani : e é ‘ .| 22 2 74 57 | Att.
BawiR. .. . ef 1g 13 | 73 20 | 273.

 

 

 
500

 

 

 

 

 

GEOLOGY OF INDIA.

Latitude, Longitude, Page.

6 ' o t
Basi : 26. 50 | 76 6 | 68, 69.
Bassein . 16 46 94 48 | 18, 337, 424.
Bastar . 19 82 4 | 4, 91, 375.
Baudwengyee 23 97 20 | 45.
BaxA . . «| 26 50 89 36 | 118.
Bedésar 27 3 70 49 | 227.
Behar . ‘ «| 25 11 5 34 | 20; 32) 57, 427.
Bela. ‘ : 26 7 “78 10 | 66.
Bela I. . ‘i «| 23° 52 7O 45 |} 218, 220, 32.
Belgaum, 15 52 | 74 34 | 48, 263, 381.
Bellary . 1§ 9 | 76 57 | 24, 35, 40, 48, 375
Benares 25 19 | 83 3 | 427.
Betdl . . 21 54 77° («57—=«| «92.
Betwa R. 24 20 | 78 10 | 436.
Beypur . Ir 10 75 51 | 4.
Beyt I. . 22 27 | 69 10 | 323.
Bezwada . ‘ 16 31 80 39 | 34.
Bhagalpur 25 15 | 87 2 | gr.
Bhagirathi R.. ‘ 24 0 88 15 | 440, 444.
Bhagothoro . . -| 26 21 67 54 | 309.
Bhagwa : 24 35 | 79 14 | 27.
Bhander H. -| 24 10 80 40° | roo.
Bhartpur . 27 13 77 32 100.
Bhiaura . «| 24 40 85 45 | 31,57,5%
Bhima R. C 17 0 77: +59 =| 84, 87.
Bhit H. . 26 10 67 40 | 305.
Bhita . . 25 19 82 21 | 94.
Bhopal . 23 16 77 26 | 3, 264, 275.
Bhor Ghat 13 47 73 23 | 261, 271.
Bhoj 23° «13 69 49 | 216, 220.
Bhusdwal . 21 ot 75 47 | gor.
Bhutan . 27. «0 89 30 | 45, 140, 348.
Bhutra . . 23. «2 78 50 | 309.
Bidna . . . 26 54 77, 21 «| 70, 72.
Bids R, . ‘ 3 45 75 30 | 450.
Bibinani . 29 42 | 67 26 | 293, 310.

 
GEOGRAPHICAL INDEX,

501

 

 

Bidar . ‘3
Bidoung H.
Bijawar .
Bijori . .
Bikaner :
Bilaspur
Bilheri . 7
Billa Surgam .
Birbhim
Bisramganj .«
Blaini R.
Bodhan ‘
Bogapani
Boghin R.
Boileauganj .
Bokaro R.
Bolan Pass.
Bombay .

Bor H. .
Borobhum
Brdhmandbad
Brahmani R.

Brahmaputra R.

Brahuik H.
Broach .
Budavada
Bugdoti R.
Bugti H.
Bukk or .
Bulsdr
Bundelkhand

Bundi . ;
Burhdnpuc .
Burikhel

Byangyi ;

 

 

 

 

Latitude, Longitude, Page,
9 ' o !
17 33 77° 34 «| 374
: 9 23. | O4 57 | 147
24 37 79 3t | ST-
22 22 78 29 | 167.
> 28 Oo 73 22 | 226, 308, 455.
+} 220° 5 82 12 | 5, 165.
23 47 80 20 99.
: 15 25 | 78 15 | 395:
: 23 54 | 87 34 | 152,176, 378, 392.
24 50 80 19 | 97.
go 55 | 77 & | 133.
2. 17 73 8 | 301.
. 25 20 | gt 46 | 61, 330, 331.
24 38 80 32 | 97.
. 31 6 77 10 | 134.
‘ 23 «47 85 42 | 166,171.
‘ 29 52 67. 16 | 290, 293, 304,307, 310, 319.
18 55 q2 54 | 11, 255, 259, 271, 278, 406,
| 409.
26 «68 67 56 | 287.
23 86 25 | 63.
»| 25 53 | 68 49 | 452.
22 Oo 84 55 | 162.
, 26 20 | 92 0 | 45s 3321427, 431, 439-
20 Oo 66 4o | 8.
21 43 73 2 | 282, 300, 407, 412, 416.
15 51 80 12 | 181.
4 25° «5 82 48 | 436.
‘ 29 «0 69 «OO 8, 318.
27 43 | 68 56 | 452.
20 36 72 59 | 409.
: I, 96, 104, 279, 3725
‘ -| 25 27 75 41 | 103.
~| 21 19 76 16 | 396, 411.
«| 32 43 qx 48 | Lit.
; -| 16 18 94 46 | 18.

 

 

 

 
502

GEOLOGY OF INDIA.

 

Cc
Cachar . ; ‘ A
Calcutta ‘ . 3
Caligudi é °

Cambay 7 .
Candahar, see Kandahar.
Cauvery R. . ‘ 7

Cawnpore . ‘ ;
Chaibasd 7 j ‘
Chakrata 3 5 :
Chamba . .

Champaner_ , . ’
Chandar . 7 . a

Chanda . . . :

Chandernagore . $
Changchengmo R. . ‘i
Chari. ‘ ‘ ;
Charli. ‘ ' ;
Charwar H. . .

Chebu . .
Chedubal. .
Ghela_ . . .
ChendbR. . ‘
Chengalpat . j

Chenpura ; : :
Cherra Punji . ‘
CheyairR. . 5
Chhatisgarh . . .
Chhattarkot .  . .
Chhindwéra . .
Chhota Udaipur. :
Chichali H., see Maidani.
Chichali pass. . :
Chidru . . : .
Chikidla .

 

 

 

 

Latitude. Longitude. Page,
° I o ,
‘ 24 50 92 52 | 331.
. 22 34 88 24 | 415s 432, 441, 475.
-| 10 59 7® 59 | 234.
22 18 42 40 | 13, 301, 408, 412.
‘ 14 40 77 50 | 38, 40, 48, 237, 396, 403,
413, 416.
. -| 26 28 80 24 | 427.
, -| 22 33 85 51 | 32, 63.
. »{ 30 43 77, «54 «| 517.
. «| 32 29 76 to | 137, 461.
22 31 | 73 36 | 73.
“| 25° 7 82 55 | 94.
. -| 19 56 79 20 | 159, 164, 165, 265,279, 402,
413.
. 2z2 51 88 26 | 234.
; “| 34 15 78 30 | 140,348.
23 34 69 39 | 218, 219.
19 51 79 20 | 158.
. 23 «10 69 40 | 216, 220, 222, 223.
. 25 18 | 81 65 | 94.
18 50 93 35. | 12, 20, 338.
25 12 gt 41 | 330.
g2. 15 73 30 | 72; 461.
12 42 80 1 40.
27 53 | 76 10 | 69.
25 17 | 98 47 | 247: 205, 329.
Iq 15 79 15 | 80.
2115 81 41 | gt, 150, 163, 174.
25 12 80 53 | 97.
: 22 3 | 78 59 | 92, 167.
A 22 20 74 +1 | 206, 278.
3 33 «8 Ji 25 | 229, 2°6.
: 32 33 71 50 | 123,
19 3 | 79 59 | 185.

 
GEOGRAPHICAL INDEX, 503

 

 

 

 

 

 

Latitude, Longitude, Page,
°o 7 ° '
Chikkim H. . 5 - : »| 32 22 78 2 | 204.
Chiknayakanhalli_ . ; : | 130 25 76 40 | 48.
Chilkd L, - 7 ; . -| 19 40 85 25 | 404.
Chilpightt . . . 22 10 | 81 7 | 64.
Chimur . . : . : 20 31 79 25 | 92.
Chindwin R. . . “ 5 -| 22 40 95 20 | 18, 336, 343, 423.
Chiniot . : ‘ - ; -| 3l 44 73 «+t ‘| 72.
Chinna Tirupati. : : -| 16 57 81 19 | 180,
Chirdkhan s > . Ze -| 22 22 75 249.
Chitaldrug i : : . -4 14 14 76 26 | 48
Chitarkeot , : ‘ j | Ig 12 82 46 | I.
Chitor . . g : ‘ -| 24 52 74 41 | 68, 93, 98.
Chittagong. ‘ . . «| 22 21 gt 6, 12, 335.
Chitapahdr HH. . eee] 833 go. «|g | 352.
Cholamo L. . 5 . ‘ .| 28 2 88 48 | 488.
Chopra 5 ‘ 3 ‘ .| 24 27 79 28 | 96.
Chor H. . : 7 : -| 30 52 77 32 «| 43, 117.
Chorar I, 7 23 52 qi 14 | 215, 218, 321.
Chota Nagpore, see Chutia ican
Chotiali 3 3 ‘ 4 -| 30 2 68 56 | 291, 305, 418.
Chunar, see Chandar,
Churna I. é . 7 . 24 54 66 38 | 352.
Chutid Nagpur 3 é 5 -| 23 0 84 o | 5, 32,63, 154, 375
Cocandda . : ‘ : -| 16 57 82 13 | 180, 182.
Cochin . ‘i 2 . ‘ : 9 58 | 76 17 | 299, 405.
Coleroon R. . ‘ g ° il o 79 20 | 232.
Colgong j i : r -| 25 16 87 17 | 176.
Coimbatore . ‘ . -| iro 76 0 | 38, 40, 412.
Comorin, Cape ‘ ; . 8 4 77°39 | 3: 377:
Cossyah H., see Khasi H.
Cuddalore. : 7 : -{ Il 43 79 49 | 392, 404.
Cuddapah ‘i ‘ 5 -{ I4 29 78 52 | 33,35, 40, 78, &6, 375-
Cumbum «wet 34 «| 79 | 8
2 207, 215. 279,
Cutch , , , : .| 23 20 69 30 | ee ce ee 79
Cuttack : . - é .| 20 29 | 85 54 | 178, 378.

 

 

 

 

 
 

 

 

504 GEOLOGY OF INDIA.
Latitude, Longitude. Page.
° ° i
Cutch, Runn of 24 «0 7O O | 11, 13, 408, 430, 453
D
Dabrai . 31 8 66 18 | 294.
Dadu , 28 34 76 21 | 72.
Dagshai ‘ go 53 77 6 | 350.
Dahihanda 20 53 77, «I =| gol
Dalhousie 32 32 76 o | 137.
Daling . 27 #1 88 46 | 76.
Dallipur é 24 27 79 13 | 28.
Dalma H. -| 22 53 86 17 | 63.
Daltonganj 24 2 | 84 7 | 161.
Daman 22 25 72, 53 (| 407, 409-
Dambal 15 18 75 50 | 48, 375.
Damdama , 26 54 77. 19 | 70.
DAmodar R. . 2 23 40 85 20 | 149,152 162, 170, 173,
174, 176, 177) 444.
Daphla H. ‘ 27° +0 93 40 | 150.
Darjiling , 27 3 88 19 | 42, 45, 76, 468.
Dehri . 24 32 79 3 | 27.
Dehra Dun 30 19 78 5 | 356.
Delhi. 28 39 77. 16 | 69, 70, 427, 428.
Denodhar H. 23 27 69 23 | 279.
DenwaR. . af 221 33 78 15 | 167, 173.
Deoban H. ‘ 30 44 97. «+50 «| 117.
Deokalli 25 2 79 58 | 28.
Deola 22 27 74 36 | 240.
Dera Bugti 29 69 12 | 310,319.
Dera Ghazi Khan 30 3 Jo 50 | 292.
Dera Ismail Khan . 31 50 7o 59 | 428.
Devala . II 29 76 27 «| 37.
Dewalmari 1g 18 80 2 | g2.
Dhamni ; 20 16 79 12 | 265.
Dhaneswari R. 26 30 94. «0 | 44.
Dhankua H. . | 24° 47 78 47 | 27.
Dhansiri, s7e Dhaneswari R.

 

 

 

 

 
GEOGRAPHICAL INDEX. 505

 

 

 

 

Latitude, Longitude, Page,
° ! oO !
Dhdola Dhar ~ : : »| 32 15 76 30 | 44, 140, 461, 484.
Dhar. : : : ; -| 23 36 75 4 | 248.
Dhar Forest . ‘ : ‘ .| 23 20 76 10 | 53, 103, 253.
Dharampur . ‘ : : -| 30 51 77, 8 =| 138.
Dhararah * : Z * -| 25° 15 86 27 | 59.
Dhariawad_ . 24 «1 74 30 | 68.
Dharwar 5 . : . .| 15 26 75 5 | 48,375.
DhasanR. . 24 10 78 50 | 94, 96.
Dhaulapani . . ‘ ‘ «| 24 15 74 44 | 79, 72.
Dhawara ‘ - 5 : -| 25 13 78 41 | 27.
Dhosa . : : . -| 23° 19 69 41 | 220.
Dibrugarh. ‘ 5 ‘i .| 27 28 94 57 | 331, 427-
Dihing R. : ; : .| 27 30 96 30 | 334 463.
Dinajpur . : ‘ 3 .| 25 38 88 41 | 431.
Disang R.  . ; ; ; .| 27° 9 95 25 | 148.
Dohad . : ‘ $ -| 22 53 74 19 | 267.
Doigrung R. . ‘ ‘: js .| 26 20 93 50 | 296.
Dongargaon. «is : .| 20 13 | 79 9 | 265.
Drés_. ‘ : : 5 .| 34 23 | 76 45 | 348.
Dubrdjpur_. ‘ : .| 24 25 87 32 | 174.
DudatoliH. . ; : . -| 39 5 79 15 | 43
Dédkér s a ee eh ope 81 37 | 269.
Dulchipur. . ‘ ; «| 24° «15 79 5 | 96.
Dunghan H. ‘ ‘ .| 29 52 | 68 22 | 290.
Dwarka ‘ : : 5 .| 22 14 | 69 S| 324
E
Edlabad , i ‘ ; .| 19 41 78 35 | 90.
Elephant point ‘ , ‘ .| 16 28 g6 23 | 425.
Ellichpur ‘ : : : at 15 77 29 «(| 265.
Ellore . . ; ~  ,.| 16 43 | Bt 9 | 1522979, 269.
Enchapalli .- i ' ‘ .| 19 2 79 57 | 186, 402.
Enchardm ‘ P 5 .| 18 28 79 47 | 92.
Eshwarakuparu- . . -| 15 50 79 40 | 82.

 

 

 

ae
506 GEOLOGY OF INDIA.

 

 

 

 

 

 

| Latitode, Longitude, Pag2,
F o , oO ’
Fatehjang . ‘ . ‘, 33 35 q2z 38 | 420.
_Ferozpur . . . : +| 30 57 74 38 | 428.
Foul I. . i ‘ ‘ . -| 18 6 0447 | 1:
False I. . . . . : -| 18 41 93 50 | 21-
G

Gadani . ‘ ; 3 ‘ “| 25°67 66 45 | 315-

Gddawaéra . . . - 22 55 78 50 | 3972 398, 399.

Gaira H. i . ‘ : . 23 37 68 4o { 321, 322.

G4j RK. . 7 s : z -| 26 52 67. 20 | 304; 305, 306, 311.

Gundahari H. . : “ -| 27 6 69 46 | 310.

Gandak : . : +; 30 22 67 12 | 318.

GangesR. . : : ; . | 25 45 84 o | 426, 450.

Géngta . 5 : “ +' 23° 45 JO 32 | 220.

Garhwal . ‘i : : -| 30 8 78 48 | 43,117; 134, 23%, 349, 465,
406, 48.

Garo H. E : ‘i - -| 25 32 go 15 | 296, 323, 332.

Garudamangalam . : 3 -| oa 5 73 58 | 237.

Gatparba R. . . : : - 16 15 75 45 | 84, 403.

Gauhati ‘ , 3 ‘ 26 11 91 48 | 427; 440.

Gaull. : : : i -| 15 34 74 24 {| 38.

Gaya. 3 3 : : -| 24 49 85 3 | 575 58.

GhaggarR. . ; ; ‘ «| 24 37 83 10 | 99.

Ghansura : ; ‘ , -| 24 59 85 20 | 58

Ghazipur . : ; ; -| 23° 35 83 38 | 437-

Gidhaur a 3 : , .| 24 51 86 14 | 57, 59-

Gilgit. . . . F -| 35 55 74 22 | 419.

Girnaér . a ‘ 3 : 21 30 qo 42 | 279.

Giumal . . . . é 32 10 78 14 | 229.

Goa : . . : . .| 15 30 973 57 | 377-

Godlpard ‘ . : i -| 26 41 go 41 | 427.

Godavari R. . : : : -| 17° 30 81 © | 33, 89, 91, 151, 162, 168,
179, 184, 264, 268, 402,
413, 470.

Gogi . Fi 7 ; j -| 16 44 76 49 | 88

 

 

 
GEOGRAPHICAL INDEX, 507

 

 

 

 

 

 

Latitude, Longitude, Page,
° t o t
GokAk . 7 ; ‘ Ss .| 16 to 74 53 | 83, 403.
Goldghat : : , «| 26 30 94 © | 206.
Golapilli ‘ 7 : ‘ 17 43 so 58 | 179.
Goona . ‘ ‘ ‘i ‘ .| 24 40 77, 20 «| 255.
Gooty - . ‘ d ‘ : 5 7 77, «42 78, 80.
Gondicotla. : , F .| 14 49 78 21 | 81.
Gopat R. . : : ; .| 24 25 82 15 | 50.
Gujérat ‘ ‘ : . 25 #0 72 0 | 300, 375, 337, 407, 414.
Gujrh . * - J i .| 22 20 75 34 «(| 258.
Gulcheru ‘ 5 “i : -| 14 18 78 48 | 79.
Guntir . : ; ‘ 16 18 80 29 | 78, 81.
Gwadar . s ‘ é 3 .| 25 © 62 yo | 316.
Gwalior - ‘ 5 3 : «| 26 13 78 12 | 65, 256, 279, 275
H .
Hab R. ‘ 4 : : .| 24 §2 66 42 | 306, 309.
Haidardbad . ‘ ; ; .| 47° 22 78 30 | 33, 261, 265, 411.
Haidar4bad . . : .| 25 23 68 25 | 453.
Hakra R. é s i .| 28 0 Jo 0 | 450.
Hdla_. ; ‘ “ : .| 25 48 68 27 | 312.
Halaman H. . “ : ‘ 23 20 69 51 | 218, 220.
Hamadun ; i . 30 29 67 24 | 293.
Handid - « @ & gb 82 28 7 FF 2 | Sang08
Hanle . ; : ; é .| 32 47 |. 79 4° | 460.
Harangaon . : : . -{| 22 45 77 2 =| 256.
Hardd . : i ‘ : .| 22 24 77. 8 | 54, 398.
Han Pal «© « « = 22 2 | 74 45 | 396.
Haripur j “ é . «| gas 72 59 | 116,
Hasan Abdal . ‘ 3 i -| 33 49 72 45 | 138.
Hatni R, a . ‘ 5 .| 72 42 74 33 | 213.
Haveliyan , F . : -| 34 3 73 14 | 116,
Hawshuenshan ‘ ‘ ; .| 24 28 98 46 | 15.
Hazara ‘ ‘ : - .| 34 : 9 73 15 | 43, 138, 220, 352.
Hazdribdgh «ee 23 59 | 85 25 | 3s 300312 62, 154s 166, 177«
Henzada ‘ ‘ ‘ ‘ .| 17° 38 95 32 | 147s 337s 425

 
 

 

 

 

 

5c8 GEOLOGY OF INDIA.
Latitude, Longitude. Page,
Le] I QO /
Herat +| 34 20 62 I1 140.
Hindaun 26 44 | 77° 5 | 67,71.
Hindubagh . go 51 67 47 | 143.
Hindu Kush . 36 Oo 71 0 | 7,41, 140.
Hingir 21 57 83 46 | 168.
Hingl4j 25 34 | 65 47 | 315-
Hingoli 19 43 77 11 | 402.
‘Hiran R. 22 12 74 to | 266.
Hlwa R. Ig 26 94 12 | 144, 366.
Hoshang ab 4d 20 45 77. 46 | 103, 159, 258, 275.
Hothian Pass 25 45 67 57. | 303:
Hughli R. 22 55 82 26 | 444.
Hukong 26 45 96 39 | 423.
Hundes a Ba 348, 422,464, 477,
495.
Hurnai go. OS 68 o | 290, 395, 37-
Hutar is 23 50 83 53 | 161.
Indargarh . 25 44 76 14 | 93, 103.
Indore . . 22 42 75 54 | 3
Indrawati R. +| Ig 10 81 Oo | QI.
Indus R. 29 5 | 70 30 44, 305, 345» 419, 428, 449,
460.
Inikurti . 14 21 79 46 | 34.
Innapdérazpdlayam 17 15 82 28 | 180.
Irawadi R. 23 30 96 o | 302, 324, 330, 378, 423-
Irlakonda 16 2 78 41 | 82,
Iskardo 35 12 75 35 | 140.
J
Jabalpur ‘ 23 «11 79 59 | SS 187, 262, 264, 372, 396
39°:
Jabi . . «| 31 54 72 10 | 123.
Jacobébad  . . . .| 28 17 68 29 | 457-
Jaggayyapet . ° «| 16 52 80 9g | 78.

 

 
GEOGRAPHICAL INDEX.

 

 

Jaintia HH. .
jaipur. :
Jaisalmer ‘i
Jakhmari .
Jalna. .
Jamalamadigu
jJambughora
Jamird Pat.
Jamkhandi
Jammu

Japvo . :
Jarra’. .
Jashpur
Jaunsar
Jehlam
Jerruck .
Jessor

Jhalra Patan
Jhand

Jhansi

Jharid .
Jhilmilli
Jind.
Jobat

Jodhpur

Jowai .
Jumna R,

Jura H.
Jutogh .

Kabaung R. .
Kabul

Kach .
KachaoH .

Latitude. | Longitude,

25 30 92 15
18 55 82 38
26 55 | 70 57
26 g 67 56
T9 50 75 £6
14 51 78 26
22 19 73° «47
23 5 | 83 45
16 30 75 22

25 36 | 94 6
23 Al 69 5
22 53 84 12
39 43 77 «54
32 35 73° «47
25 3 68 18
23. 10 &9 15
24 32 76 12
33 26 7205
25 27 78 37
23 44 | 86 29
23° 24 83 55

26 17 | 73 «4
29 15 77 10

23 23: «| 69 36
3. 8 | 33 30

18 52 96 18
-| 34 30 | 69 18

25 8 | 94 46

 

 

 

5°9

Page,

eee

296, 334.

4s 375:

227, 308, 455.
287, 289.

19, 402.

85.

73:

255s 374.

82.

359 354s 357, 467.
335+

218,

375+

445 117.

140, 419, 466.
303, 306, 309, 312, 315.
434.

93: 94, 98, 102.
419.

27.

165.

157.

72.

40, 51, 54-

74, 106, 430, 455.
296.

436, 446, 450.
218, 220.

134.

342.
7s 419.
293.
148.

 

 
 

 

 

510 GEOLOGY OF INDIA.

Latitude, Longitude, Page,

° fe] !

Kachaoda . : 22 29 75 11 | 253.
KAfiristan ‘ 35 30 71 o | 41.
Kagdn ss . | 34 40 73 33 | 138.
Kdimur H. ; 24 30 81 45 | 3.
Kakindiya : 23 47 Jo 26 | 220.
Kdlabagh 32 58-| 71 36 |. 228, 326.
Kalddgi ‘ 16 12 95 32 | 50, 82, 265,
Kalahand: .| 20 0 83 20 | 375.
Kalapani R. 25 24 gt 48 | 61.
Kaliana ; 28 «it 76 6 | 72,159.
Kaliani . f 97 «+21 17 52 | 374.
Kalka . ‘ 30 50 7® 59 | 351.
Kalu R. ‘i 25 35 go Oo | 332.
Kalyan , ‘ 19 14 73 10 | 271, 278.
Ka-ma . 3 19 O 95 II | 339.
Kamamet . -| 17 16 | 80 11 | 33, 150.
Kamatki ghat . 18 1 74 5 | 259. S
Kamthi . H 21 13 79 «14 | 168,
Kd4nara . 12 52 74 53 | 48.
Kandahar . -| 31 37 65 30 | 293.
Kangra a . +] 31 20 33 357, 484.
Kanhan R. 21 30 79 167, 168.
Kanhar R, . 24 20 83 15 | 56.
Kami. , 22 27 | 94 53 | 18.
Kentkot : 23. 27 Jo 28 | 222.
Kapili R. 7 rar) 92 40 | 296 ,331.
Kappatgod  . 15 36 | 75 54 | 375.
Kdpra . . . 18 30 79 48 | 92.
Karachi . . . 24 51 67 4 | 311, 314, 315.
Karakoram range . . -| 35 29 77. 6 | 140, 229, 460, 461.
KaraliH. . . «| 22 12 73 53 | 278.
Karanpura. . ate 28059 85 5 | 166,177. ‘
Karauli . . . «| 26 30 77, «4 «| «94, 103.
Karchat . . 25 45 67 46 | 306, 312.
Karen-ni . . 19 30 97. Oo | 142,
Kargil . 3 34 «34 76 6 | 346,

 

 

 

 

 

 

 

 

 
GEOGRAPHICAL INDFX. git

 

 

 

 

Latitude. Longitude. Page.
° ! ° ,

Karharbarf . : : . .| 24 10 86 20 | 160, 164.
Karnil . ‘ . : ’ «| 15 50 76 § | 35, 40, 86, 89, 395.
Karo R. . . ‘ : -| 22 42 74 10 | 266.
Kartse . F : é . -| 34 16 706 o | 132.
Karwi . i ‘ ; -| 25 12 80 57 | 04,97.
Kasakanahal . . . . -| 16 31 76 39 | 88.
Kasauli . . . , -| 30 53 77, ~««1~=«|:«350.
Kashmir ‘ ‘ ‘ . -| 34 6 74 51 | 41, 44, 116, 134, 420.
Kashmor . 28 26 69 36 | 15, 345, 428.
Kasom H.  . . ; A -| 25 0 04 43 | 334-
Kéteru . . . A ‘: -| 17 3 81 48 | 268, 270.
Kathidwar. : ‘ : .| 22 20 qo 55 | 189, 253, 259; 279, 323, 295,

408, 409, 412, 410, 454.
Katikela ‘ 3 : ; .| 20 47 84 9 | 33
Katrol . ‘i c , . .| 23 12 69 50 | 221.
Katta. : . . : -| 32 31 72 30 | 122,
Kau-ran-gyi_. : 2 F -| 16 31 93 48 | 340.
Kavhad . : : . .| 32 27 q2 11 | Il.
Kawant . . . ‘ ‘ {| 22 6 74 5 ~+| 253-
Kayal . ‘ ‘ : ‘ : 8 38 78 10 | 13.
Keantali : ‘ ‘ i .| 18 0 94 34 | 297) 338.
Ken R. . 2 2 , ; .| 24 22 79 20 | $2, 59; 94) 97-
Khdaibar Pass és 3 : | 34°55 gt 8 | 141.
Khairemirut H. . s 3 33 28 q2 50 | 353-
Khairgaon . j : ; -| 19 49 79 15 | 158.
Khairpur . ‘ ; . .| 28 3 69 44 | 451-
Khalsi_ . igs ois - 34 20 g6 52 | 345» 347-
Khandesh_. , ‘5 ‘ .| 20 54 74 46 | 261, 278, 396, 398.
Kidneue: se <a. “, (2) BGreg [oe BRL eeeage.
Kharakpur . ws | 25 7 [86 35. | 59 375:
Kharir I. : : : : .| 23 52 qo 22 | 218, 220, 321.
Khési H. . ; : ; .| 25 34 91 55 | 60, 150, 246, 296, 329, 332.
Khattan i é i 5 .| 29 34 68 29 | 291, 203) 374, 397-
Khelak« « «. = # «| 28 gg | 66 28 | 143-207
Kherly. 2. 6 os ap 1a | 97-8. |. OO
Kheura . 7 ‘ : . -| 32 39 73 3 || V0:

 

 

 

 
512 GEOLOGY OF INDIA.

 

 

 

Latitude. Longitude, Page,
oO t ° 1
KhisorH.  . , : ‘ «| 32 20 qi 10 | 109.
Khuln4& . 5 : ‘ ; .| 22 49 Xo 37 «| 434.
Khundghat_ . . ‘ ‘ .| 32 28 72 16 | 122,
Khwaja Amrdn H. : ‘ -| 30 39 66 30 | 142, 243.
Khyber, see Khaibar.
Kim R. . ; . . .| 2% 25 72 40 | 300.
Kira H. : ‘ ‘ ‘ «| 23 37 69 17 | 218, 219.

Kirdna H. . : : . -| 31 57 72 44 | 72.
Kirta. 7 ; - . -| 29 32 67 32 | 293.
Kirthar H.  . ‘ ‘ . -| 27° 0 67 12 | 8, 305, 308, 311, 315) 451

Kishengangd R. . i ‘ -| 34 45 74 0 | 138.
Kistna R. F ‘ 5 5 -| 16 30 79 20 | 33, 40, 82, 402, 412, 416.
Koari Bet. : , ‘ -| 23 58 69 47 | 218.
Kohat . . . ’ -| 33 36 qi 29 | 325.
Kohima : ‘ . 7 +} 25 40 94 9 | 148.
Koilath . ; ; ‘ : -| 27 50 73 31 | 308.
Koil Kuntla_. 7 : . +} 15 34 73 23 «| 86.
Kolamnala. ; : é -| 16 0 79 52 | 82.
Kolaér. : : : . -| 13° 8 78 10 | 48.
KopilasH. . ‘ : . «| 20 41 85 59 | 375,376
Korba. . : . : +] 22 20 82 46 | 165.
Kori R. . - » P 3 :+ 23° 45 68 40 | 452,453.
Korkai . 7 : . ‘ of ar 2 79 49 | 13.

Kosi R. . : ‘ ; ‘ -| 29 30 79 1 | 469.
Kota ; . . -| 18 55 80 2 | 184.
Kotasir . E . ‘ ‘ -| 23 41 68 35 | 453.
Kotri. - : : ‘5 .| 25 22 68 22 | 253, 306, 312, 315, 428.
Krol H. : . : : -| 30 57 76 10 | 133,13}
Kubo. : . ‘ . ») 24 15 Ot 3° | 9,423.
Kuch Behar . . . -; 26 20 89 29 } 332.
Kuchri . ’ . : c | 27° 4 qO 37 | 228

Kuling . . . . . oft 88h) 38 78 = Qg | 139.

Kulu. : . . é -| 3r 58 | 77° 7 =| 117.

Kumaun ‘ . : ‘i -| 29 35 79 41 | 43 17, 349, 464, 466, 469,
480.

KundairR. . : : ‘ -]| 14 50 78 40 | 78, 84, 86.

 

 

 

 

 
GEOGRAPHICAL INDEX. 513

 

 

Latitude, Longitude, Page.
° ! ° /
Kuram RR... ‘ 3 ‘ | 33 37 70 34. | 328.
KyaukpG 5 se ee} tg 25 | 93 gt | 20.
L
Laccadive I. . ‘ . ' -| 12 0 77 0 | 12,
Ladakh . : i i j -| 34 10 77. 40 | 44, 4€0, 461.
Ladera . . . ' ‘ 26 3 78 24 | 105.
Lahore . ‘ ; ‘ . -| 31 34 74 21 | 428.
Lairangao. , ‘ ‘ .| 25 20 Q1 47 | 330.
Laisophiang . . ' é -| 25 13 gr 46 | 246.
Lakhi H. : 5 : i -| 26 0 67 50 | 286, 393, 305, 317.
Lakhpat 3 : . : -| 23 50 68 49 | 321, 322.
Lametd ghat . . : ' -| 23 6 79 53 | 262.
Larkhana . . : ‘ +| 27 33 68 15. | 452-
Leh 7 ‘ j a * -| 34 10 77. 40 | 44, 345-
Lenya R. : ' : : -| 11 30 99 © | 297.
Lhasa . ‘ . ‘ ‘ -| 29 41 ot 6 | 205,477.
Lilang . . ‘ ; : -| 32 9 78 17 | 130.
Lintzithang . ‘i é . +| 34 50 79 15 | 204.
Lodai . . . ‘ : -| 23° 24 69 57 | 218.
Lokapur ‘ . ‘ : -| 16 10 75 26 | 84.
Lokhzung H.. i ‘ ‘ -| 35 10 Jy 40 | 204. ;
Lonér . : i , ‘ ‘| 19 59 76 33 | 19.
Long Island . : : ; -| 16 15 94 40 | 337-
Luckeeserai . 7 : -| 25 11 86 9 | 59
Lucknow ° F ‘ . -| 26 52 80 §8 | 432, 475.
Ludhiana ‘ ' ‘ ‘ -| 30 §5 75 53 | 428.
LiniR.. : - ' . 26 35 72 35 | 430.457.
Luni Pathan H. . j ‘i «| 30 10 69 40 | 305.
Lus. ‘ é ‘ ; .| 25 20 66 45 | 22.
Lynyan . ; : ‘ : 25 39 | 68 12 | 303.
M
Mach . . ‘ ‘ , .| 29 22 | 67 23 | 304.
Madanpur_ . ‘ ‘ ; -| 24 15 78 46 | 29.
Maderapaucum .. : ; 13 27 80 4 | 389.
Madhupur_ . 7 ‘ . +| 24 34 go o | 44I.

 

 

 

 

 

 

 

 

 
 

 

514 GEOLOGY OF INDIA.

Latitude, Longitude, Page,

o t 3° tf

Madras ‘ 13. 4 80 17 | 4, 12, 378, 388, 4o4.
Madura 9 55 78 10 | 39, 184, 412.
Mahdabaleshwar 17 53 73 42 | 4, 278, 374.
Mahdabalipur . 12 37 80 14 | 12,
Mahdabar H. . 24 35 85 55 | 57,53
Mahadayi R. 15 27 74 25 | 381.
Mahadeo 25 12 or 47 | 246.
Mahadeva H. 22 20 78 30 | 5, 167, 279.
Mahdnadi R. «| 20 45 84 30 | 39,91, 107, 151.
Mahdnadi R. 24. «0 80 53 | 96.
Maharajpur . 26 54 78 18 | 105.
Mahendraganj 25 18 go 54 | 332.
Maher . A 24 43 85 13 | 57:
Maheswar 22 It 75 37. | 258.
Mahuagarhi H. 24 29 87 26 | 376.
Mai-f 19 20 94 13 | 297, 338.
Maidanf H. 32 51 qt iu | 228, 326.
Main Pat ‘ 22 48 83 20 | 263, 267, 374, 393.
Maiwand m 31 43 65 16 | 294.
Makim 27 18 95 41 | 33t.
Malani . . 25 45 7l 25 | 74.
Maldivel. . 6 oOo 73 lo | 12.
Maléri , Ig It 79 40 | 184.
Malkdpur.. 20 53 76 23 «| 4or.
Malparba R. 16 o 75 57 | 403.
Malwan . . 16 73 30 | 377; 409.
Madmand 29 39 68 45 | 307.
Man R. , 22 20 75 %0 | 253.
Manbhum. 23 20 | 86 25 | 62.
Manchhar Lake 26 25 67 42 | 3°5, 313, 451.
Mandalay 21 59 96 168, 336.
Mandar H. 24 50 87 31.
Mandld 22 35 | 80 24 | 92, 279, 383.
Mandogarh 22 21 75 26 | 249.
Mandsaur 24 3 75 8 | 70.
Maner R. ‘ -| 18 30 79 45 89.

 

 

 

 

 

 
GEOGRAPHICAL INDEX. 515

 

 

 

 

a
Latitude, Longitude, Page,
° ! Q t
Manga Pir see Pir Mangho.
Mangli . ‘ . ‘ : «| 20 22 79 «4 «| 169.

Manimalai_. 2 . ‘ -| 1 36 76 11 | 37.

Manipur s+ ee | 24 48 | 93° 59 | O 147, 207, 334, 3355
423, 494.
Mansurah : ; , . -| 25 53 68 49 | 452.
Maobehiarkar . i | 25 24 gt 48 | 295.

Maosmai - . . . -| 25 15 | Ot 47 | 246.

Marai . 6 ‘ a 2 -| 24 9 81 16 | 56.

Maraura ‘ F . : -| 24 23 78 52 | 26.
Maravatur. ; cn 2 aif aR ag 79 oO | 236°

Mari H. ; . - 7 -| 29 20 68 50 | 291, 304, 307, 318.
MarkhaR. .  . «sed 33. «55—«|«77~=—«20—«*|- 346, 347.

Maret . . . 3 . si) 23; £6 75 8 | 70

Marpanmadi ‘ : 7 +] IL 31 76 28 | 37

Martaban . - ri . -| 16 32 97 38 | 142, 378, 425.
Mashalak H. ‘ ‘ ‘ -| 30 20 66 50 | 417.

Maski - 2 + tt -| 15 57 | 76 43 | 48.

 

 

 

Masulipatam . . . . -| 16 9 81 12 | 150.
Matabhanga R. a‘ . . -| 23, 40 88 45 | 440.
Matapenai H. . 2 3 -| 22 12 73° 53 «| 278.

Mathar. - + + + | 33 33 | 67 50 | 325"
Miihertn « « «© « ‘#1 98 $0 | 7 08: | apa, gp, 38%
Ma-tun . 3 - 5 - -| I9 12 95 Oo | 337.
Maulmain.. . . : »| 16 30 97 38 | 142.
‘Mauphlong . : : 25 27 9! 48 | 61, 295.
Meerut . : : . . -| 29 #1 77 45 | 427.

Meghna R. . ‘i : -| 23° 45 or oO | 44r.
Makalgandi ghat . . : 19 35 78 48 | 267, 276.
Memkal = . ; 4 15 1 77, 9 | 50.

Mergui . ‘ ‘ ‘ 5 12 11 98 38 | 141.
Midnapur. : i i 22 25 87 21 | 3745 392) 405.
Milam . ‘ ‘i i 3 .| 30 26 80 13. 129, 130.
Minbu . ‘> é i : | 20 13 94 57 | 20.
Minet-toung . : . : -| Ig 10 95 35 | 339

Min-gyi. ‘ : 7 . .| 18 8 95 30 | 425.

 

 

 

2L
516 GEOLOGY OF INDIA.

ee moe en ot ae ay ela cnenee cae en ieee tists

Latitude, Longitude, Page.

 

° Be he '
Mirzdpur a ‘ é ‘i .| 25 10 , 82 38 | 29, 56, Go, 430
Mogouk Pe ee ee ee
MohanR. . «© +» + +f 25 54 | 67 56 | 303.

Mehar , , 7 : é -| 25 38 78 13 | 105.
Mohpéni - ett 22 45 «| 78 54 | 161, 397.
MoirSH. «ef 25 10 | 86 GI | 370, 375.
Molim, see Myllim.
Momien. j : : : “| 250 5 98 46 | 18.
Monghyr . «© = +] 235 23 | 86 30 | 177, 440.
Monze,Cape- + + + <‘f 24 50 | 66 43 | 312, 315.
MorR.. «© «© + «+  <f 24 © | 87 30 | 444.
MoranR.  . . . 5 ‘| 22 20 | 77 32 | 54.
Morar . B 2 z . -| 26 14 78 16 | 65.
Mortaka . © + «  +{ 22 16 | 76 6 | 53 93.
Motir . . ° . ‘ 22 17 78 37 167.
Mugger Pir, see Fir Mangho.
Muhammadpur «ee | 27, «9 | 77, 3. | 6,
Mio. eee tf 2 | 79 44 | 2
Mulakhel : . . . *[ 32 55 71 13° «| 228.
Miltan. .  - . - +f 30 12 | Zr 3r | 428.
Miungi . : . . . “} 19 25 73 309 | 402.

Muree. . + + + =| 33 54 | 73 26 | 352, 355
Murshidabad . . . ‘ -| 24 4 88 19 | 440.
Murtazapur . ‘ . . -| 20 44 77 25 | 401.

Musakhel . a Fa . -| 32 38 73 49 | 216.
Mussooree . : . . *| 39 27 78 6 | 133.

huth ; ; a fs -| 31 58 78 6 | 114
Mutla R. . . . . +} 22 19 88 43 | 434.
Myanaung . ‘: . j «ft 18 17 95 22 | 337, 378, 425-
Myitmakha Khyoung ‘ -} 18 15 95 30 | 425.
Myllim . r é ‘ . 4 25 30 gr 52 | 61, 62.
Mysore . F - z 5 -} 12 18 q6 42 | 51, 37,48, 412.

N4gAH. 2 ee tf 260 | 0g 15 | 8, 335, 494-
Nagar Péarkar os ‘ -{ 24 21 Jo 47 |.454,

 

 

 

 
GEOGRAPHICAL INDEX.

 

 

Nagari Nose.
Nagode. . ‘.

Nagpur. ‘
Ndhan . . . . e
Naira R. ‘I . . :

Nal Lake i ‘ «
Nallamalai H. .
NambarR. . ‘
Ndmdang R. . ‘

Nancowry . -
Nandgaon . é :
Nandial. : > : ‘
Nandidlampett

Naoshera ‘

Naosir . ‘ 3 ‘
Narakal ‘

Nardoli . , i : ‘
Narbada& R. .

Narcondam I. j
Narha . ‘ 3
Nari R.

Narji . el 7 : is
Narra R. .

Narsinghpur .

Narwar -

Naushahra . ‘ .
Neilgherry H., see Nilgiri.
Nellore . ‘ .

Nerdl . ‘ . .
Nga-pu-tau . : . :
Nga-tha-mu .

Nicobar, I. . 6 ;
Nilang , ,
Nilgiri . :
Nimach :

 

 

517
Latitude, Longitude, Page.
° ! ° !
13 23 ; 79 39 | 78, 80.
24 34 80 38 101,
2169 79 7 | 33, 151, 168, 262, 264, 268,
280, 374, 402, 413.
30 32 | 77 «21 | 356, 358, 466°
3° 39 77 45 | 133:
22 48 | 72 5 | 408,454.
15 0 | 79 © | 481,493
26 17 93 50 | 296.
27 16 | 95 45 | 331,
8 o | 92 34 | 344
Ig 50°) 79 12 | 157-
15 29 78 32 | 86.
14 43 78 52 | 8.
33 10 | 74 18 | 467.
25 46 ql 52 | 227.
10 2 56 17 | 405 ,
26 20 76 4f | 103.
22 30 77. 10 | g, 248, 249. 396, 431, 495.
12 36 g94 15 | 17:
23 39 69 10 | 189, 222.
26 40 67 20 | 308.
14 39 | 78 35 | 86.
26 25 69 oO | 451,452
22 57 79 14 | 5+ 54, 398-
25 38 | 77 58 | 105.
32 34 72 13 | 228.
14 27 | 80 1 | 33, 40) 50 78, 373, 416.
27 42 | 85 12 | 75, 421,438.
16 30 94 46 | 424.
16 30 | 93 49 ; 34%
8 o | 93 35 | 1% 343
3 6 | 799 4
In 25 | 76 45 | 4s 14s 37.4% 375» 409) 415:
24 28 74 54 | 70: 103-

 

we et ee

 

 
GEOLOGY OF INDIA.

 

 

 

Nimaér ...
Nimawar i
Niniyur .
Nirmal .
Nithahar ‘

Niti . .
Nongkulang H.

Nullamullay H., see Nallamalsi.

Nurla_ . %
Odiam . °
Olapadi .
Ongole .
Opalpad .
Owk, see Auk.
Pédbar R. -

Pachamalai .
Pachmarhi .
Padwani .
Pachpadra .

Pagan :
Paitan ‘
Pakhal . 7
Palamau F
Palar R. :
Palghat .
Palkonda H. .
Palndéd ji
Palni H. :
Pamir 5
Panchet .

Pangi

 

 

Latitude. Longitude. Page.
o , ° !

21 50 76 23 | 253, 398.

a2 30 | ,77 3 | 256.

11 16 79 13 | 240, 242, 243.
19 6 78 25 | 150.

26 58 17 4 68, 70.

30 46 | 79 52 | 130, 464.

25 17 gt 61 | 331.

34 19 | 76 so | 346.

1 13 1 79° «2 | 236.

11 20 79 «8 | 235

15 30 80 8 181.

15 10 | 78 6 | 80.

gr 8 77 54 | 117; ¥34-
Woo15 78 40 | 4-

22 27 78 29 | 5, 167,170, 172.
2% 45 7317 «| 278.

25 56 72 13 | 430.

21 9 | 94 52 | 17-

1g 28 75 22 | 402.

17.57 | 79 59 | 89.

24 21 84 71 | 279.

12 40 80 oO | 413.

10 46 76 42 | 11,495.

14 0 79 10 | 81.

16 40 80 o | 78, 82, 84, 86.
10 10 77° +40 41.

37° 30 73. 30 «| 4

23 37 86 49 | 170.

76 go | 78 42 | 14.

 

 

 
GEOGRAPHICAL INDEX. 519

 

 

 

Latitude. longitude, Page.
= Q ’ ° t
PangongL. .. . . -| 33 45 78 45 | 486.
Paniam . . . 7 -| 15 31 78 25 | 86.
Panna . - . 3 -| 24 43 80 14 | 97, 101, 107.

Pan-ta-naw  . . . . .| 16 55 95 28 | 426.
Panwari . . : é .| 25 26 79 32 | 97- 102,
Pépaghni, R.. ‘ ‘ . -| 14 20 78 30 | 79.

Par . F 7 . ‘ .| 26 3 78 6 | 65, 105.
ParaR. |. . . : -| 32 35 78 12 | 130.
Pdrasnéth . ‘ ‘ .| 23 58 86 10 | 14.
PariharH.  . < . 7 27 WW 7O 42 | 227
Patcham, J. .- : , - «| 23° §2 69 50 | 215, 218, 220, 321.
Patko, Ho =. se | 25 40 | 94 | 8, 335.

Pand 2 wwe | 20 2 | 83 TB | 375

Patthargatta . . . ‘i .| 25 41 87 52 | 428.

Pavulur . oh, x - .| 5 5L 8) 4 | 181,

Pawagarh. . . . .| 22 31 73 36 | 73) 259-

Pegu  . ts = aan | 17 20 | 96 30 | 145, 336, 378, 413. 424

PeléniR. 2. «6 ey 29 10 |: 78 49 | 480°

Pengangé R. ‘ ‘ : -| t9 45 78 30 | 158, 160, 213, 402.

Penner R., North . ‘ : ~| 14 40 77 20 | 48, 78, 80, 403, 413-
» South .  .  « «| 12 20 | 78 20 | 232,413.

Perambalir . ‘ 7 ‘i | ar 14 738 54 | 236.

Perim, [. ‘ é ‘ ‘i .{ 21 36 72 23° «| 302, 323-

Peshdéwar i si ; ‘ .| 71 37 34 2 | 418.

Phondaghat. . +. + 16 20 | 73 56 | 82.

PhuleliR, . 3. «ef 25 30 | 68-29 | 453-

” Pichor 2 oe ew wt 25 57 | 78 27 | 105.

Pid : a 1 oe 32 4% | 73 2 | ¥2t.

Piram I. see Perim I..
PirMangho. . . ~ «| 24 59 | 9% “4 J 312

PirPanjffl .  . «= 33° 40 | 74 40 | 43, 44,461, 462
Pisdura F ‘ é ; 20 21 | 79 © | 265.
Pishin . z i ‘ : 30 37 67 5 | 142, 318.

Places garden . : . | 13 2 79 53 | 182.
.| 26 55 | 71 58 | 106, 160, 455-

icherri y 6 | 79 53 | 10, 18, 231, 235,239, 247s
Pondicherri . . : ‘ m § ee

Pokaran : 3 s

 

 

 

 

 
 

 

 

§20 GEOLOGY OF INDIA.
Latitude, Longitude, Page.
or eo 7

Poona. é 5 18 31 73 «55 259, 271, 278. :
Porebandar . : 21 37 69 48 | 395.

Port Canning - 22 1g 88 43 | 434.

Porto Novo . 4 Ir 29 79 48 | 404.

Potwar 33 «8 72 O | 14, 418, 484.
Prénhita R. , ‘ 1g 30 80 oO |} 90, 184, 279,
Prome é . 18 43 95 15 | 339 340.
Pédikattéi ‘i 10 23 78 52 | 40.

Pugha 33. 13 | 78 20 | 346
Pilivendala . 14 25 78 17 | 80.

Pulkoa ws 24 35 | 79 53 | 96
Pullampet . 4 15 | 79 5 | 8h

Pulney H., see Palnf.

Puppa H. 7 . 20 57 95 10 17.

Puran R, 24 40 | 69 15 | 452.

Punch R. : 33 25 | 73 47 | 351, 467.
Pungadi 17 #1 81 41 | 269.

Puriam Point 15 .19 93 58 | 337» 425.
Purna R. : 21 Oo 76 25 | 396, 401, 449.

Q
Quetta . . go 12 67 4 a a 305, 307, 318)
Sule sey 8 5g | 76 37 | 299, 304.
R

‘Ragavapuram i 17 2 81 23 | 180.

Raialo . ‘ z 27 6 76 17 | 69.

RaidakR. . 26 43 89 50 | 45.

Raichur Doab 160 78 0 | 36, 86.

Raipur . 21 15 St 41 | 64.

Rajagriha H. . 25 0 85 30 | 58, 60.
Rajah’s choultry , 15 52 79 50 | 244.
RAjdmahendri ; 17 0 | 8% 48 | 179, 263, 268, 279, 30%

 

 

 

401.
GEOGRAPHICAL INDEX.

521

 

Rajanpur
Rajapur
Rajmahai
Rajmahal H.

Rajpipla

Ramganga R.

Rdmgarh
Ramgarh
Ramnad
Ramnagar
RAmpuré
Ramri I.
Rangoon
Raniganj

Ranikot
Ranj R.
Ranjit R.

Ranthambhor .

Ratanpur
Ratnagiri
Ravi R.
Rawalpindi
Rebni .
Red Hiils
Rer R.
Rew4

Ridsi
Robdar R.
Rohri
Rohtdsgarh
Rongreng
Rotés .
Ripbds
Rupshu

Latitude.

Longitude,

Page,

 

 

29

. . 24
’ . 25
eo] 24

21

° . 29
. -| 23
. >| 22
: 9
« . 23
. * 25
. . 19
16

. 23
25

. 24
27

.| 26

‘ | 24
17

. . 30
«| 33

. «| 19
‘. «| 13
. 24
° 24
e}) 33

29

27

. 24
27

. 32
27

33

5)

30

54
33
38
53
22
45
58

47
36

54
48
15

44

37
16

- 30

31

35
42
37
16

58

 

70
80
87
87

73
79
85
86
78
86
76

99
87

67

. 89

88
76
73
73
72
73
79
80
83
81

74
67
68
83
95
73
7

78

22
25
53
30

50
20
56
56
46
39
39

 

 

432.
28.
427, 428.

150, 159, 174, 372, 376,
383, 413.

5» 261, 278, 200.

469, 480.

165,

63.

"412

437.

98.

20, 297, 338.

340; 342, 424.

159, 161, 164, 166 179,
177 437, 392.

303.

97-

76.

103.

301.

36, 300, 377; 389:
355) 401.
352, 418.
186.
378.

56, 60,

29, 56,153, 161,167, 173»
186, 253, 279, 383»

347; 351.

293.

305; 430, 451+
95:

331.

AIO.

100,

42, 486.
522 GEOLOGY OF INDIA.

 

 

 

Ss
Sabzal-kot . ‘ “
Sadiya . - -
Sadri . .
Safed Kch
Sagar . . . 3

Sahdranpur . : : ‘
Sahyddri H., see Western Ghats.
St. Thomé . é
Sainthoray . . . é

Salem . ; ‘5 ° s
SAletekri a 3 . -
Salsette ; : ji ‘
Salt Range. ; : .
Salwin R. : : é ‘

Samaguting .

Sambalpur . . . :
Sambhar Lake é ‘ ji
Sandoway . . . .
Sandur . ‘i ‘
Sanjd. . .
Sankaridrig . :

Sanktoria s ‘ ‘ a
Sara R. ‘i é ;
‘Saraswati R. . : ; .
Sargiga ee ae
Sariab F f a
Sdsser4m - A : ;

Sétpura H. . ‘ : .

Sattavédu

SAawantwari .

Sehwan

Semri R. . ‘

Sesni . ;

 

 

 

 

Latitude. Longitude, Page,
oO i fe] I
/

29 9 | 70 2 | 432,435.

27 50 | 95 42 | 427, 459.

24 23 | 74 46 | 70.

34 oO qo 0 75 41, 140, 229.

23. 50 | 78 49 | 93: 383

29 58 | 77 35 | 427:437-

13 2 80 19 12,

mw 15 79 «13° (| 240.

Ir 39 78 12 | 25 38, 412.

21 47 80 52 | 64.

Ig 10 92 57 | 259» 271,

32 40 73. © | 7,8, 109, 119. 228, 286,
352, 353, 355) 418.

16 45 | 97 40 | 142, 424,425.

25 45 | 93 46 | 334

21 27 8, 1 32, 70, 107, 151, 169.

27 55 75 5 | 7%

18 28 94 25 | 297-

15 5 | 76 35 | 48,49.

33 4 | 78 30 | 204

11 29 | 77 56 | 38

23 «42 86 52 | 437.

29 9 | 79 58 | 468.

29 658 76 52 | 450.

23 20 °83°«I5 153» 157, 255, 263, 267,
279s 374, 383-

30 «6 67. © | 290.

24 57 84 31 | 93.95:

22 20 79 0 | 3,5, 54,150, 154 167, 170,
172, 2U1, 393.

13 26 | 80 1 } 182, 389.

15 54 | 73 52 | 36.

26 2° | 67 54 | 309, 428, 451, 452.

24 35 | 79 35 | 96.

22. «5 79 35 | 92,411.
s

GEOGRAPHICAL INDEX,

 

 

 

 

 

 

 

 

 

 

 

 

523
Latitude. | Longitude, Page.
°o ’ ° ,
Seringapatam ‘ ‘ 12 26 76 43 | 48.
Shahgarh A 24 19 79 «U1 26, 96.
Shaikhawati . 3 : 28 oO 75 0 | 69.
Shaikh Budin 32 18 7O 51 | 228,
-Shaikhptra H. a +| 25 8 85 53 | 57,59.
Shali H. . . . +} 31 1 77, 20 «| «117,
Shayak R. >; 34 45 | 77. © | 420, 460,
Shevaroy H. . II 52 78 13 | 4, 14, 387, 415.
Shikaérpur 27 57 68 4o | 315, 428.
Shillong +! 25 33 91 56 | 44, 6, 295.
Shimoga -| 15 55 75 36 | 48.
Sholdpur . ° -| 37 40 75 57 | 261.
Shordpur i -| 16 31 76 48 | 48.
Shrishalam : . -| 16 5 78 56 | 82.
Shyok R, sez Shayak.
Sibi F 39 4 67 50 | 318, 418.
Sibsdgar 26 59 94 38 | 331, 427.
Sichel H. . . . *| 19 35 78 50 | 267.
Sikkim . ‘ e : a7 | 88 19 | 149, 348, 488.
Simla . si - 31. 6 77, ‘11 | 107, 117, 132, 136, 340,
459, 465.

Simra. : 25 2 87 26 | 176.
Sind R. ; : . 25 45 78 15 | 65, 105.
Sindri . 24 16 69 11 | 454.
Singareni a é 17 31 80 20 | 91.
Singarh, 18 22 73° 49 | 259.
Singhbhiém 22 33 | 85 5t | 24, 32, 62,
Singhe LA. , ie 33 58 | 76 54 | 347.
Singpho H. . 27 30 96 30 | 335.
Singrduli ‘ : 24 6 82 55 | 30.
Siran R. . -| 34 7 72 57 =| 419.
Sirban . -| 34 6 | 73 16 | 116, 138, 139, 229, 286.
Sirmur . . «| 39 33 77 42 «| 117, 133, 349.
Sironcha : . : ; -| 18 31 80 1 | 151, 185, 268.,
Sirst . $ : . : «| 29 32 75 7 =| 449.
Sitsyahn i.  <& 18 54 | 95 14 | 339.

 
524 GEOLOGY OF INDIA.

Latitude, Longitude.

Page.

 

Sittaung R. . : ; ‘ -| 14 15 | 96 30
Skardo, see Iskardo.
Sohdgpur. . . . .| 27 52 78
SohanR, « ». «© » «| 33 5 | 72
Solan. : 7 : . »| 30 55 77
Son R. . a ‘ : : -| 24 15 81 30

Sonat, < « «© «© «| 83 93 | 2 %
Sonmifni wg weet 2h 27 | 88
Spinutangi . . ‘ . -| 29 55 68 8
Spita Raga . : ’ . «| 30 33 67 46
Spiti. a fa ‘ : J} 92g 78 15
Spiti . 7 ‘ . : .| 28 21 77. 36
Sripermatir . ; , ‘ -| 12 58 80 1
Subathu . . . . -| 30 58 17 «2
Sukakheri . ‘ : .| 22 49 98 52
Sukkur . . ‘ i ‘ .| 27 42 68 54
Suldiman H. . . . . .| 31 40 jo oO

Sullavai : : . é .| 18 12 80 10
SumesariR. . i ; ‘ -| 25 20 go 45
Supur . é . : ‘i afi 23> OE 86 56
Surarim Fi 5 i .| 25 «18 Ql 47
Surat. . * a ‘ -| 21° «9 72 54

Suri ‘ ‘ , F ‘ .| 23° «54 87 34
Sutlej R. 30 15 73 20

Swarnamukhi R. . - ‘ -| 13 45 79-47
Sylhet . : 7 . . -| 24 53 gl $5

T
Tadputri ‘ : i ; -| 14 55 73 4
Tagling Pass . : . «| 32 32 77 «+58
Takht-i-Suld4iman . ; . »| 32 36 jo 2

 

 

 

142, 342, 378; 423, 426.

555 153 267, 398,
418.
133, 138.

51, 52, 55. 94, 98, 9 103,
154, 162, 279, 383.

105.

315

290, 307.

142.

75s 129, 130, 137, 229, 294.

132, 244,
350.

397; 293.
308, 457.

7, 141, 229, 292. 304, 305,
310, 318, 425.

92.

333

63.

61, 295, 330

259, 262, 278, 282, 300,
3745 407, 408, 412, 414.

175» 392.

42, 351, 449) 452, 461, 464,
469.

50.
44s

“80.

131.
292.

 

 
GEOGRAPHICAL INDEX. 525

 

 

 

Latitude. Longitude, Page.
fs) ° t

TAR. . . . ’ . -| 29 47 98 42 | 230.
TAlcher. * < ‘ é »| 20 57 85 16 | 32, 149, 153, 157, 1695 174.
Talra. , ‘ . : -| 27 12 76 47 | 70.
Tandwal . . ; : -| 34 20 72 55 | 130.
Tandra Rahim Khan. ‘ -| 26 32 €7 25 | 312, 315.
Tanjore ‘ ‘ F E .| 10 47 79 10 | 304, 404, 412.
Tarkesar ee : F ,| 21 22 73. 6 | 301.

Tapti R. . . . . -| 2% 30 75 40 | 9; 300, 396, 398, 400, 407.
411, 431, 495-

Tarnot . . : : ‘ .| 20 45 82 31 | 64,91.

Tatta. ‘i . . ey ee ee 68 o | 306, 312.

Taung-gup_ . 7 . ‘ .| 18 50 94 20 | 297,338.
Taung-ngu. . e ‘ .| 18 55 96 31 | 342, 424.

Tawa R. : ‘ ‘ : «| 22 45 78 =§ | 167,173-

Tenasserim . . ‘ ‘ 12 5 99 3 | 45,141, 2907, 343, 378.
Tezpur , é : : é -| 26 37 | 92 53 | 256.

Thal. é ‘ . x «| 33 37 7o 34 | 328.

Thal (Chotiali) ‘: 3 ‘ -| 30 1 68 46 | 291, 305, 418.

Thal Ghdt . ‘ a j -| 19 43 73 30 | 261.

Thalapidi . 7 ; ‘ -| 17°79 81 44 | 179.

Thana . - ‘ : 7 »[ Ig 4 73 «4 | 271.

Thar . ‘ ‘ ‘ .| 25 20 | 69 45 | 455

Tharia. . «© «©  « «| 25 af | of 48 | 247, 296, 329, 331, 332.
Thayetmyo . . «. + «| 19 19 | 95 16 | 144, 336, 340.

Thondoung . ‘ : ‘ .| 19 16 | 95 14 | 338:
Tinnevelli a é c 4 . 8 44° 77. «44 sig? — 39, 392, 495; ne
Tipam H. é ‘ , i -| 27° 15 95 3° | 334.
Tipperah Hills - : a -| 23 50 gt 23 | 443.
Tirhowan, see Tirohdn.

Tirohan ; i ; i «| 25 12 80 58 | 96, toz.
Tirupati ‘ i. ; , «| 16 57 81 19 | 180.
TirupatiH. . : ; .| 13°38 | 79 28 | 78.
Tirupatir «=. ssf kg | 8 SQ | 238:
Tiruvakarai . F ‘ ‘ .| 12 1 79 43 | 235) 393+
Tisté R. . oe ee] 25 50 | 89 46 | 76.

Todapurti, see TAdpatri.

 

 

 

—= SSS
[ree

 

 
 

 

 

526 GEOLOGY OF INDIA.
|
Latitude, Longitude, Page,
° ' ° ;
Todihal . ; ‘ : 16 25 75. 37 | 265.
Tow 6 es 25 46 | 67 35 | 312.
Tongiup, see Taung-gup.
Tons R. . 7. ‘i ‘ -| 30 59 7 o | 134.
Toshim H. . 28 54 | 75 56 | 74.
Tranquebar . Ir 2 79 54 | 12.
Trap. : > 25 33. 3 7i 58 | 419.
Travancore 8 29 | 76 59 | 4,49, 200; 377, 385, 304s
405, 416.
Trichinopoli . : Io 50 78 44 | 38, 39, 40, 152, 182, 211,
231, 378; 414.
Tripetty, see Tirupati.
Trivicary, see Tiruvakarai.
Tso Moriri L. A “ 32 55 78 23 | 486.
Tungabhadra R. . ‘ r 15 45 77, 0 | 48, 79.
Turd. - . 25 20 go 16 | 296, 329.
U
Ubra . x - . ‘ 24 25 83. 3 O«|s«#50.
Uchar . i A i - 25 53 78 30 | 105.
Udaipur ‘ fi 24 35 73, 43 «| 68, 69.
Ukra H. , . 23° «45; 68 55 | 216, 223, 286.
Umarkot . ‘ ‘i A .| 25 21 69 46 | 429, 451, 452s 455-
Umballa . : . a »| 30 21 76 52 | 428, 432, 435, 476.
UmblaiR. . : . , -| 25 20 or 5 | 331-
Umia . ‘ ‘3 ‘ »| 23° 41 69 1 188, 232.
Umrali . . : 3 . -| 22 10 74 21 | 249.
Undutla 3 F 5 3 -| I5 36 78 23 86..
Upalpdd . ‘ 15 10 738 6 | 86.
Utatir . ‘ , ‘ : -{| Ir 4 78 54 | 152, 184, 233, 236, 237-
Vv
Valimikam . ; - A ‘ 9 10 78 42 | 11, 405.
Valudayur. . . : Ir 59 79 45 | 231, 232, 233, 241.
Vellar R. z . 7 II 29 79 48 | 232,404.

 

 

 

 

 
GEOGRAPHICAL INDEX, 527

 

Latitude, Longitude, Page,
° t ° '
Vemavaram . . : ; «| 15 41 80 13 | 181.
Vempalli . 14 21. 78 30 | 79.
Vengurla . : . . -| 15 54 73 30 «| 83, 279.
Venkatagiri . . . . -| 13° 57 79 37 | 33
Venkatpur. . . : -| 18 15 80 3 «+I 92.
Vigori . . . . : +} 23° 31 69 8 | 223.
Vindhya H. . . . 3 -| 23 °0 78 oO | 3,92.
Virdwah 2 ‘ -| 24 30 Jo 48 | 454.
Viruddhdchalam. 7 . .| If 31 79 24 | 231, 232, 233, 255, 240,
241.
Virgal . . . ° 2 .| 32 27 72 07 | 122.
Vizagapatam . 3 i «| 17 42 83 20 | 34, 403.
WwW
Wadhwan . . ; © ef 22 42 71 44 | 254.
Wagad ° é : ‘ -| 23° 35 JO 40 | 215, 220, 222, 224, 319.
Wagalkhor . . . . «| 21 45 73, 16 | 301.

Waindd . . is . -| Ir 50 76 3 | 37540.
Wainganga R. ‘ ; . «| 20 39 80 o | 184, 402.

WajhiriH. . . . < /} 15 52 73 46 | 36.

Wajra Karur . . . «| 15 4 77. 27 | 40, 107.

Wandan R. . _ ” . .| 28 o | go o | 450.

Warangal . 7 . 5 «| 17 58 79 40 | 150.

Wardha R. . ae Se - ,| 20 30 78 30 | 90, 92, 169, 186.

Warkalli ~ + «» « «| 8 4h | 7 46 | 299) 304

Warord s . . . | 20 14 79 «2 «(| 265.

Wer. : : ‘i , | 27° «21 77. 14 | 70.

Western ghats ‘ 5 5 -| 19 0 73 30 | 3s 4, 10, 257, 279) 415, 494+
Y

Vedakalmolai H. . : . ¢| Yt 37 76 18 | 37.
Yeddihali. . . . -| 16 32 76 36 | 88.
YVellakonda H. . é ‘ -| 1§ 0 | 79 10 4, 78.
Yellamala H. . . ‘i | 14 45 78 20 | 86.
Yenangyoung - «* ~ a) aay | oa me [te

 

 

 

 

 
 

 

528 GEOLOGY OF INDIA.
Latitude, Longitude, Page.
oO ‘ oO ,
Z
Zami R. j ‘ -| 16 Oo 98 10 | 142.
Zanskar ‘ . . 33 30 77. +O | 421, 132, 347, 460, 477.
Zhob . . . . . -| 3100 68 o | 142.

 

 

 
INDEX OF SUBJECTS.

—+—-

PAGE : PAGE
Aravalli Range, antiquity of— 6, 71,
104, 106, 491

A + mA former extension
; of— ‘ » 483

Abur group 2 . i; 226, 228 4 a relation to upper
Africa, cretaceous e . 247s 252 Vindhyans 103
» Gondwana reptilesin — , 195 Aravalli system - 6 40, 67, 74
» Karoo RELIES Chae, ony RAOd 5 relation to Delhi system 70
» land connection with nee 209, Archipelago series : : . 344
; : 233, 493 Arenaceous group, 320, 322
» Uitenhage series. -182, 205 Argillaceous group 320, 322

African affinities of Billa Surgam

cave animais 7 : » 395
Agate . : ‘ : 176, 260
Agra boring ‘ i 432, 435, 476
Albaka division . 2 Fi 89
Aleppi, smooth water anchorage . 405

Almod group . : 156, 170
Alum shales . z 303, 352
Alwar quartzites . : : 69

+» ,relationto Ardvalli system 68
American fossils in upper Pro-

ductus limestone. i » 125
Amethystine quartz. ‘ . 260
Ammonite zones in Cutch Jura-

ssic . . i 225
Ammonitide, permian, in Salt-

Range . . . 123, 127
Amygdules 7 . 176, 257, 289
Analcime . . ‘ : » 176
Anamesite A . é 175, 257
Anchorages, smooth water . « 405
Andugu Kyauk . ‘ . - 337
Andaman Islands, evidence of

changes of level. 7 ; 12
Anhydrite, inclusions in quartz. ‘112

3 Salt range 7 . 12
‘4 origin of — ‘ - 112

Apatite. ‘ . ‘ 5 30
Apophyllite . . - 260, 261
Arabian cretaceous deposits . 252

 

Ariyalir group . 233, 236, 239, 244, 250
»  fossilsin Assam, + 247
35 » 9» Africa. » 249
» gneiss peblesin— . 236, 239

ss Telation to Utatir and
Trichinopoli groups . 235,

238, 241

Arkose in Alwar group ; . jo
»» resemblance gneiss 24, 50, 53,
59> 71, 134

Artesian wells, Lucknow. « 435
3 Pondicherri . + 404

Assam range, comparative absence

of disturbance. 44
analogy to Salt range 8
a detached fragment of

the Peninsula 44, 60, 439
Athgarh group . ‘ 156, 178
Athleta beds. ri 220, 221, 226
Attock slates - e * - 116
Augite ‘i ‘ . . » 30

39 Ed

gf #9

Auk shales . . . . 83
Australia, carboniferous fossils in

India. 121, 132, 135

coal measures. - 198

relationship of flora

to Gondwana flora 200

cretaceous ammonite in

Cutch ‘ e . 286
. . 7 » 144

a ”

”

Axial group
PAGE

B

Babeh series : 114, 117
Bacchus marsh beds . ‘ » 198
Bagh beds 7 152, 248

connection with cret-

aceous of Europe and
Arabia . : 211, 452
relation to Deccan trap 249,
266, 275, 281

2?

»

Pe 3 Lameta . 264

. 3 upper Gond-
wanas 173s 249
Bdgra group» . - 173
Bairenkonda quartzites . . 81
Ball coal. . . ‘ « 165
Balmer sandstones . : 226, 227

Banaganpallf group . . 85, 86, 107
Bap, Talchir boulder beds at . 160

Barakar group é . 16, 162, 166
»  ageofthe . 197, 206, 208
» coal. * . 161, 164

supposed representation
near Athgarh >» 178

”

Barren Island. A A : 16
Basalt . . . . 175, 257
Baxa series 5 . . - 118
Bedesar group . * 226, 227
Behar transition series + 539573 75
Belemnite beds . ‘ . » 290

Bellerophon, in Ceratite (trias) beds 129
Bengal gneiss, contrasted with

Bundelkhand é . i‘: 30
Bengal transition series fs 62, 75
Bezwada gueiss . . : ° 34

Beryl s 3 ‘ i 30
Bhibay . . 437, 446, 469
Bhingay and Khddar, distinction

of . 438, 444

” ” cause 445; 451
Bhander group < ‘ i . 101
Bhiaura quartzites . . ‘ 58
Bhima series : ; ‘ 84, 87
Bhur . a 438, 449
Bijawar system . , 575 59; 64, 75
Bijori group 7 : 167, 195
Billa Surgam caves. 5 - 395
Biotite 7 , . 5 30

Blaini group + 129, 133, 143, 148
- » age of the ‘i e 137
Bloodstone ©. ‘ F . 269

 

530 GEOLOGY OF INDIA.

PAGE

Blown sand + 415, 438, 455, 458
Boilleauganj quartzites ‘ 134, 138
Bole, inthe Deccan trap, » 259
Bombay, intertrappeans .. « 272
» submerged forest , Ul, 407

Boulder beds (conglomeratic slates)
»  Bijdwar ‘ - 54, 55
»  Champdner ,. a ‘ 73
» Dharwar ; . 7 49
», Jaunsar system . + 117
see also glacia! boulder beds.
Brahmaputra alluvium + 45, 431, 439
Breccia formed by fracture zz situ
28, 51, 83, 88, 96

» pseudo. 291, 292, 304
Brine wells, Gangetic alluvium 429, 448
9 Purna alluvium « 401
Bronzite . . ‘ ‘ . 4
Budavada group i + £56, 181
Bundelkhand or older gneiss + 24
a Bellary 245 355 49

"6 Bengal area . . 32

ss Bundelkhand . 2s, 30, 46
Schistose beds

in Marauraregion 26

an old land

” ”

” ”

surface 29, 491
” Dhar forest area. 53

Cc
Cail, an abandoned seaport . . 13
Calcite ; - é ‘ 260, 289
Calcutta boring . * 432, 475
Cambrian fossils, Salt range » 113

Capra, occurrence of on Nilgirihills 15
» in pliocene of Perim Island 324
Carbonaceous system + 132, 137, 139
Carboniferous and trias, absence
of break . « FQ
contrast between
peninsular and
extra-peninsular =. 119

se beds included in
Panjal system * «116
i glacial epoch 120,206, 493
» 3» boulder beds 129, 135,
157s 159°
» oo Africa . « 203
» 9 Australia. 121, 198:

+ 9 doubtful examples 160,345
” period, changes of
land and seaatcloseof 125, 493
SUBJECT INDEX. 531i

PAGE
Cardita beaumounti beds, Salt range 532
” 7 Sind . 288
Cassiterite : . 305 45
Cave deposits, Billa Surgam » 395
Central gneiss of Himdlayas » AI, 42
Ceratite beds. 7 . » 128
Cerussite . er) » 30
Chabasite . ‘ . ‘i » 260

Chalcedony . ‘ 176, 289, 412
Champaner beds 3 ‘ 73; 73
Chari group- " . 217, 219, 222
Charli, flexible sandstone. » 158

Cherra sandstone ‘ + 295, 330
Cheyair group . < 79; 80
Chidru group : 123, 128, 207
Chikidla group . - 186
Chikkim series ~ 7 j 294
Chilpi ghat beds . : . 64
China, Productus series fossilsin . 125
Chiniot hills ‘ . . + 92
Chlorite : oo + 25,26
Chloriticshist . - : . 26
Chrysotile . : + 30, 146, 292
Cleavage . . + 55: 73, 86
Closed drainage areas + 9, 418, 461
Coal Gondwana, origin of . - 150

Talchir group . 157
Karharbari group 160,
161

Damuda series 160
161, 164, 166, 202

” P Jabalpur group. 187
» Jurassic, Salt range . - 228

” ”

” ”

a ”

» Cretaceous, Assam 265, 296
i a3 ? Tenasserim . 297
» Tertiary, Assam + 329, 330, 331

Balichistan . «305
Burma 334, 336, 338
Punjab 350, 351, 352, 353
” » Sind . . + 303
»  Tenasserim . 298, 343
Coal, jonwer Vindhyan black shales
taken for ‘ 95
» pebbles in Gangetic "delta 434
Coalfields, Gondwana, cause of
limitation of =. . » 153
Columnar structure in Basalt 175, 258
Conglomerate, deformation of peb-

” 2”
39 2

37 ‘os

bles in . 49, 54, 59
” imitated by concre-
tionary action 291,

292, 304, 307

 

: PAGE
Conglomerate, in gneiss 335 595 73
Contemporaneous erosion, in Gond-

wanas 5 . . « 150
Copper. . . 7 63
Coral reefs, etieeaiay South India 234.

‘ permian in Salt range . 126
Coralline limestone group . 249
Corundum, in gneiss of Singrauli . 30
Cotton soil 5 . . « 410
Cremnoconchus . . . II

Cretaceous, continuity with tertiary

deposits 255, 285, 303, 305
‘3 distribution of land
and sea * 211, 493
es fossils in Attock. slate
area. z - 146
a Nummulites . + 291
55 South India
3 affinities in
South
Africa 247, 251
PP 3 affinities in
Khasi hills 246,
251

contrast to
West India 211,
240, 251

original li-
Mitationof 231,
235, 237, 241, 244

relation to
Gondwanas_ 152,

” ”

182

35 Narbada valley - 248
35 ‘5 affinitiés in
Europe and

Arabia. 21,

252

contrast to
South India 211,

246, 251
es » divergence
from South
Africa . 252
5 es Relation to
Deccan
trap 249, 266,
275, 281
relation to
Mahadeva
series « 1733
249

2M
532 GEOLOGY OF INDIA.

PAGE
Cretaceous, Narbada relation to
Lamela group . . - 264
Cuddalore sandstone 3 « 232, 392
Cuddapah system 51, 75, 78, 89, 108
»y disturbance of the . 47, 81, 87
» Original extent of 79. 490
» possible transition age 75, 108
x fepresentative in Cham-
paner : 73
»»  Telation to Dhaewdea 48, 50, 75
ss 35 Karnils . 8c, 81, 84
Current marking, see Ripple marks

Cumbum slates ‘ i F 81

Cutch, great earthquake of . II, 453

35 Rann of . s A 215, 453

Cypris, supposed, in Karnil series 86
D

Dagshdai group . . 3472 350 351
» equivalent to Murree beds 355
» resemblance to Ndahan

group . ‘ + 357

Daling series. q . : 76
Damuda series 3 ‘ 162,192
33 age . . . . 206
58 coal. 160, 161, 164, 166, 202

i flora contrasted with
Rajmahdl +1559 177

i foreign relationships of,
flora . 192, 194, 200, 203

- outliers in eastern
Himalayas . 76, 118, 488

re Pre-Rajmahdal disturb-
ance of . . - 177

9 supposed _representa-
tives on east coast 178, 182
Deccan trap . + 255, 379, 413

5 age of the . - 280, 289
1», compared with Rajmahdl
traps. . + 176, 279

7 connected with upheaval
of the Himdlayas . 404
» foci of eruption a + 297

» newer than Vindhyan
scarps in Bundelkhand . 29

» elation to Bagh beds 249, 254,

275, 281

Pe x  nummulitic 282, 320
» fepresented in Sind 255, 287,
289

 

PAGE
Deccan trap, represented in eastern

Balichistan é + 201

»  sedimentarybeds . 152, 262,

280

»  subaerial origin . - 274

» outliers at RAj4mahendri 152, 268
Delhi system. . - 68, 69, 73
» age of the. : ‘ «95

» elation to Gwaliors at
Hindaun 7 : - 67, 7%

Denwa group ue 186, 195,
Deoban limestone

; - I
Deola and Chirdkhdn marl . . is
Depression, East coast ‘ 10, 404
Ganges delta . *10, 433
West coast II, 407
Desert of W. Rajputana . - 455
» rocks and structure of . 2

Devonian, ?Central Himdlayas 114,115
= fossils in Eastern Thibet 118

Dhéman . . F ‘ + 417

Dharampur, reputed trias fossils

from. 3 - 138
Dharwar system ; » 48, 75» 490
»» supposed outlier — in
Nellore -. : + 34
Dhosa, oolite  . é 217, 220, 221,
222, 225
Diamond . s 85, 86, 88, 97, 100
», supposed original source of 101, 107
- » occurrence near Simla. 107
Diamond gtavels. . - = =403
Diaspore , . . . . 30
Dihing gtoup  . + 324
Diorite .28, 40, 52, 61, 62, 63, 74, 84, 106,
348

Disang group. . » 148, 335
Disturbance of bedding caused
by the solution of fimestone 102, 330
Disturbance of lower Vindhyans,
possibly newer than river
valleys . . . . 99
Dolerite . . . » 175, 257
Dolomite 30, 76, 116, 113, 125, 127, 133s
138, 139) 299
Dolphins, in Indus and Ganges

Rivers ‘ ‘ ‘i « 443
Dome gneiss. : ‘ - 31, 58
Dubrdjpur group . + 174,176
Dulchipur sandstones . » + 96
Dunghan group ° - 290
Dwarkd group . A 2 . 324
SUBJECT INDEX 533

PAGE
Dykes, Deccan trap . » 277, 2785
279, 280

» in Ddmodar valley coal-
fields . ‘ - 177
» in gneiss of Souttiern. India 35, 372
38, 40, 56
» in Bundelkhand gneiss  . 28, 32,
40, 63
» in Shillong series . 7 61
» intransitions of Bengal. 63
9% Rer valley 36

_ ney in Rajmahdl hills . 176

E

Earthquake, changes of level ac-
companying 11, 12, 454
East Coast, antiquity of . 2, 39, 244, 492
iy changes of level 10, 404, 405

a plain of marine de-
nudation 4 10, 388
Be recent changes . . 12
Ecca beds é . ‘ + 203
Echinospharites : . - 118

Elevation, evidence of . 10, 404, 407, 408
Ellore, Gondwanas_, 152, 179, 394

Enchardm quartzites . : . 92
Endogenous volcano of Narcon-

dam . . . 17
Encroachment af, sea on ni land ‘ 12

Engdain tract . . : - 424
Epidote . ‘ ‘ . 26, 27, 30
EpistilLite . . 5 . - 260

Epsom salts . . ‘ » 110
Erosion of sea coasts . ‘ é 12
Eruption of mud volcanoes . . 21
x» » submarine volcanic . 18
Euphyllite . c . . »- 30

Extra-peninsular area, contrast
with peninsula ; 7 2

» transverse drainage of the
hillranges . , ‘ 8

F

Facetted pebbles . : + 120

False bedding . 123, 150, 233, 234, 237

Faults, movement apparent at sur-
face. a : « 485

 

PAGE
Felspar, puphyritic in Deccan trap
2575259
sy undecomposed in gla-
cialbeds . . » 201

» Gangetic delta F + 434
» in sandstones 134,157, 170, 201

Felsite, Malani . é 94
Fisher’s theory of mountain forma.
tion . . » 471
5 modification to suit Hima-
layas . 473
Fjords in Andaman and Nicobar
Islands 6 ‘ . 12
Flexible sandstone, Kalidna , 72

3 Charli. - 158

Forest, submerged, at Bombay 11, 405
>» on Tinevelli coast . TI, 407
Fort William boring . - 432, 475

Fossil wood group . - 18, 340, 424
Fossils, doubtful, Bhimd series. 88
Cuddapah series 81
Deoban limestone 118
Gwalior series . 66
Karnul series. 86

” ”
” ”
” ”

” ”»

» reputed, triassic near
Dharampur._—138
6 upper Vindhyan rot

Fossil plants, uncertainty of rela-
tion in . ° . . 193
Frog beds, Bombay . ‘ . 272

Fuller’s earth. a . + 308

G
Gabbro. 7 . - 146
Gaj group . 309, 311, 318, 339, 340
33 fossils in Salt range

and Himalayas » 354
»  tepresentative in Cutch
and Kathidwar . « 322

Galena . . 30
Ganges, former conciniilty with

Indus . . 428, 443

» delta . 3 » 10, 433, 440

» age of the . - 443
» course through the Gane.

tic plain . . - 446

Gangetic alluvium. + 427
» analogy with upper Vin-

dhyans . : - To4

» deep boringsin . 432,475

»  fluviatile origin . 15) 428
534 GEOLOGY OF INDIA.

PAGE

Gangetic fossils , ‘i . » 430
» nature of boundary with

sub-Himdlayas . . 470

» salt wells . : 429, 448

Garhwal, supposed jurrassic » 230

Garnetiferous gneiss . 30, 32; 34s 36, 37

Ghazij group yj . . + 304

Giumal sandstones. . 229, 204
Glacial action, in Hawkesbury sand-
stones. . + 200

Glacial boulder beds carboniferous,
Indian 120, 129, 133, 135+
157, 2060
Africa and Australia 121,
198, 203, 206

confined to low lati-

” ”

” ”

tudes . 120, 212
x»  ?nummulitic, Ladakh « 346
»  Pokaran Z ‘ - 106

Glacial period ; carboniferous 120, 206, 493
, pleistocene, evidence
of - 14, 368, 418, 486
Glaciers, carboniferous, in Peninsula 160,
213, 493
Glauberite . ‘ - Ilo
Glauconite. x . 257, 260, 261
Gneiss, limestone in . 30, 38, 39, 40, 45
» confused with gneissose gra- 23,

” a”

nite, < ; 40, 41

» Of different ages 24, 33, 36, 40,
41, 50

» sedimentary. origin of 7 42
» simulated byfoliated arkose 50,
59, 71, 134

»  Himdalayas ‘central’ gneiss 41, 42

o 3 in sub-Himdlayan

zone. » 45
a 5 relation to slates 115,
116, 117

» pebbles, in Cuddapah con-
glomerates . . 79

2 »» in cretaceous of S,
India + '237, 239

» probable transition beds
classed with ; . 53
» newerorbanded ., « 24, 29
»» accessory minerals in ‘ 30
» apparent interstratification
with sedimentary beds 24, 40, 41,
I
» banding not due to stratifica- 3
tion, . . » 46, 53

 

PAGE
Gneiss, relations to transition sys-
tems . . . 24, 31, 32
» -Older or Bundelkhand, ab-
sence of limestone in . 26, 27
» antiquity of . : - 46
» Bellary . : . - 35
Bengal ar - 32
ss Bundelkhand . f » 24, 25
» Dhar Forest . ‘i » 53
Gneissose’ granite, confused with

gneiss . 7 «41, 71
of Ardvallis ‘ . 40
5 of'Himalayas  . ~ 41, 42
“ 3 date of intrue

sion 2. 44
Godavari, alluvial deposits 396, 401, 402
gorge of the. ‘ + 401

”

Golapilli group. . 179, 186
Gold. -. . . + 37s 40, 63
Golden oolite . ‘ 219, 228
Gondwana rocks, Son-Mahanadi
watershed. : 7 i 92
Gondwana system . 6 119, 148

contrast of upper and
lower floras. 155, 178
distinction between upper

and lower. . » 155
us distribution : - 152
53 disturbance F 153, 177

Pe doubtful representative
in W. Rajputéana . 106
floras, -foreign affinities

of . ‘ . - 192
3 fluviatile origin . » 150
Es homotaxis . : - I91
‘0 range in time ‘ - 209°
” relation to cretaceous 152, 173,
182

existing river
valleys 52153
5 Vindhyan pebbles in . 151
Granite, Afghanistan . : - ‘41
»  Ardvallis . . 40, 69, 71
as Himalayas . ‘ - 42, 44
Peninsula 32, 359 37+ 39» 55 58; 63

” ”

”

rr gneissose structure in . 23, 42
. intruded by solution » 43, 62
Graphitic schist . . . » 34
Graphic granite . . . 32
Gravel slopes at foot of hills 417, 424

Green earth or Glauconite 257, 260, 261
SUBJECT INDEX, 535

PAGE

Gryphea, reputed occurrence in
upper Vindhyans  . : + 107
Gulcheru quartzite , . 79
Gwalior system 47, 50, fi, 69, 100, 108

a at Hindaun. . 67,71
Gypseous shales, . + 320, 321
Gypsum i TI0, 288, 303, 307, 311,

321, 326

»  Oiginof , . 112, 327

H

Hematite 430, 525 55, 3060, 372
Hzematite schists : +36, 49, 50, 67
Haimanta system ‘ 114, 117, 137
Hakra R. é - 449
Hawkesbury sandstones 108, 200, 207
Hawshuenshan, extinct volcano 18
Heulandite ‘ ‘ e 260, 261
High level laterite _ 3 373+ 374

a originof , - 379, 385
Hill nummulitic limestone 352, 354

Himalayas, absence of fossils south

of snowy range . - 4797
‘ ageofthe . » 367,477
$2 antiquity of river val-
leys ‘ ‘ 463, 484
7 cutting rock of trans-
verse valleys . - 463
5 drainage system . - 461

encroachment on Gan-
getic plain . - 471

9 former extension of
glaciers . 3 14, 484

7 high peaks due to special
upheaval 2 41, 462

7 orography . . + 460
post-tertiary increase of
height . . » 485
present position inde-
pendent of previous
sedimentation - 489
theory of elevation » 473
transverse strike in . 482
parallelism of bedding
between eocene and
pre-tertiary . 350,478
relation to Indo-gange-

a

7

tic plain . < 104, 466
Hindaun, outlier of Gwaliors 67, 71
Hingir group. . 156, 168, 169

Hippuritic limestone . 287, 290; 294

 

 

3

PAGE
Hislopite . ' 5 . + 260
Hornblende a c 25, 26, 30, 266
9 schists. ‘ ‘ 29

* »» replacing diorite
in Dhdrwdrs . 49
Hornstone 66, 81, 95
Hornstone breccia s 4 51
Hundes, subrecent gravels aa, 422, 4604,
485

I

Hmenite  . ‘ . + 27, 30
Implements, iron, east-coast + 390
‘i stone 378, 388, 399, 402, 403
Indo-African land connection —_209, 253,
493
Indo-Gangetic plain . : - 1, 427
Indus river, alluvium . ‘ + 451
59 changes of course » 452
$5 delta : : » = 452
ii floods ; 7 - 419

es former continuity with
Ganges - » 428
raised river eral . 41g

tidus valley, former extension of
sea . : + 429
» marine shell in salt jakeed » 430
Infra-Krol group ‘ 133

Infra-trappean, Chutia Nagpur,
Gondwana pebbles in - 154
» Rajdmahendri 152, 265, 268, 269,
279, 274; 275
» see Lameta.

Inikurti, Mica mines . a «. 34
Intertrappean beds. - 276, 265
PP Bombay . 262, 272

+ R4jdmahendri —152, 270,

271, 274, 275

Intrusion of granite by solution . 43, 62
Irawadi delta. . 7 - 424
Irlakonda quartzites . . ‘ 82
Iron clay, or laterite . ‘ 373

Ironstone shales . 162, 163, 165, 169
Iron ore 38, 186, 341, 371, 37 2s 431, 442
Itacolumite 7 . 72

Jabalpur group . 184, 186, 187
age . ° - 207, 208
alliances of the flora. 104

”

”
530 GEOLOGY OF INDIA.

PaGB
Jabalpur group, relation to Lametd 264
Jabi group . ‘ 122, 123) 127
Jade . . . . A + 30
Jamalamadigt group . F 85
Jaisalmer group 7 . + 227
Jammu, pre-tertiary inliers in sub-
Himdlayas . 140, 247, 351
Jasper . . - 260
1 beds in Gialiee system - 64
» » g Ardvalli system : 69

Jaunsar system . . » ity
Jumna river, fossils in alluvium - 436
»» Oldcourseof . : - 449

Jurassic, correspondence of Cutch
and European A - 225

»  velationships of the
Gondwana floras 192, 193, 195

K
Kdimur group. ‘ 98, 100, 102
<i outliers on Arévallis - 103
Kalabagh, bipyramidal quartz
crystals ‘ ‘ - . 4110
Kalabagh group . . - 122
Kalddgi series. ‘ é « 50, 82
Kalar, or impure salt . ‘ + Ig

Kamthi group. . . 168, 17
» relation to Kota-Maléri . 185
33 9 Rajmahdls of Ellore 179,181

Kankay . + 3972431, 436
Kantkot sandstone +217, 222, 225, 228
Kapra quartzites . . 7 92
Karakoram stones, . . 140
Karambar rings < 7 . 389
Karewa deposits ‘ . + 420
Kares . e . - 418
Karharbéri ecu . 160, 165, 198
Karnil series. : 80, 81, 84, 108
Karoo beds 7 7 - 195
Kasauli group + 3494 351, 355) 356
Katrol group. 189, 217, 221, 224, 226
i age of the . 189, 217

» , plant remains , 189, 222
Katta group . . 122, 123
Kayal, abandoned seaport . - 13
Khddar and Bhangar, distinction of 448
444, 451

Khasi trap . ‘i ‘ - 61
Kheura, salt mines. . « 19

 

PAGE
Khund ghat group. 122
Khwaja Amrdn slates, doubtful
age of . . . : - 142
Kieserite . . . . *« 10
Kirdna hills . . , s 42
Kirthar group. . 301, 305, 309
i fossils in Himalayas
and Salt range 347, 354
vs relation to Manchhar
seties. 309, 314

Nari group 309, 30
Ranikot group 303,

” ”

” bed

306. 307

3 tepresentative in
Cutch . » 321
Kistna alluvium , + 396, 402
Kistna group. : F - 82, go

Kohat tertiary . ‘ ‘ «325

» rock salt and gypsum 112, 326
Koil Kuntla limestones . . 86
Kolkot emporium . : . 13
Kolumnala shales. . . 82
Konkan . . 7 + 377
Korkai, abandoned port of . .

Kota Maleri group. 184, 195, 208
» correlated with Denwa 173, 186
» relation to Chikidla . - 186

5 Kamthi - 185
Krol limestone. . . . 134, 138
Krol] quartzites . . . . 134
Kuchri ammonite bed ‘ + 228
Kuling series. ‘ . 130, 138
Kundair group . : é ; 86
Kyanite . e . » 27,42
Kyauk-pyu, mud volcanoes + 20.

L

La Croix, on gneisses of Salem . 25, 39
Lake basins of Himalayas . 421, 486
Lacustrine deposits, simulation of 418
Lametd group. 29, 262, 263
» Mode of formation . - 276
»»  Réjdmahendri outlier 265, 268
» felationto Deccantrap . 264
Bagh beds . 264
» ss Gondwanas 264, I51
Laramie fossils compared with
Deccan trap . . - 281
Latitude, secular variationof 212, 213
Laterite . A A « 55» 256, 369

eo

a ”
SUBJECT INDEX 537

PAGE

Laterite, low level of east coast —_10, 388,

393, 404/
300, 305, 308, 321,

»  nummulitic

—

350s 353, 374

Laumonite ‘ . . 7 260
Lead . 45, 63, 81
Lemuria . . . 209
Lenya river coal beds : . + 297
Lepidolite . . . . + 30
Leucopyrite . * 3 » 30
Level, changes of i: : . 12

Lias, probable erroneous identifica-
tion in Central Himalayas 131

»» fYelationships of Gondwana

floras . . 192, 195
Lignite . 299, 300, 377; 404
Lilang limestone 130, 138

Limonite . 370, 372
Limestone, absence in older gneiss 27,.36

i disturbance caused by

solution of . + 102, 330

Pr occurrence in newer
or banded gneiss 30, 38, 39,
49, 45
Lithomarge. * 372, 377
Littoral concrete 3 . 407, 408
Loess, of Baluchistan . . 9, 418
Lonédr Jake . . . 7 19
Low level laterite + «373, 376
- of east coast . 10, 388, 393, 404
Lucknow boring y 432, 434, 476
Lydianstone . m ‘ . 266

M

Macrocephalus beds . 219, 221, 227
Madanpur gorge, Deccan trap in 29

Madhupur Jungle. . - 442
Magnesia, carbonate of i . 146
Magnesian sandstone group . 109, 113
” fragments in carbo-

niferous boulder clay 120
Magnesite . 7 . ‘i - 38
Magnetite 7 : « 30, 38, 259
Mahabar schists : . - 58

Mahddeva series : - 1972

doubifully represented
by Nimér sandstone 253
relation to cretaceous . 152,
1745 249

a

”

 

 

PAGE

Mahdnadi alluvial gravels + 403
Ma-f group . ‘ + 297, 338
i” “in Manipur . « .148
Main boundary fault . 349: 357» 467
Malabar laterite ‘ 3 + 383
<i gneiss . . . 36

Maldni series + 73, 106, 227
» fragments in earboniferous

of Salt range . + 120

Maléri group, 184, 185

Manchhar series 312, 313, 316, 364, 366

8 estuarine beds in = 34
a »» fYelation to Gaj group 314
35 »  Kirthar group . 309,

344

» »  Mekran group . 315
Manganese, oxide of, inlaterite . 371
Mangli group 155, 170, 195

Manipur, alluvial plain . « 423
Marble rocks, Jabbalpur_. : 55
Marine denudation, plain.of, East
» coast . ‘ ‘ 10, 388
» Westcoast . : . IL

Marsh gas, from mud volcanoes . 20
Massive limestone, of Jammu inliers 140

% of W. Garhwal 117
Maulmain series . “° 141, 298
Mayo salt mine . . + 110

Mekrdan group a 314, 430

»  velationto Manchhar . 315
Mica mines ‘ c . 32, 34
Miliolite . . : 340, 395
Momié‘en, extinct voleano near. 18

Moraines, at low levels 14, 484
Mordr group. 7 . - 65, 66
Morung . : ; . - 438
Motur group. . : » 167
Mud volcanoes . ‘ . . 20
Multani mitti . : is - 308
Murchisonite . . 1 34
Murree beds. . 349, 355» 467
Muscovite ‘ ‘ 30, 44
Muth series 4 115, 1295 137

N
Nagari quartzites : ; . 80
Ndahan group. . - 356, 364
» absence of pebbles in 465, 478
» fossils. ‘ 3506, 358
Nallamalai group . : $1, 90
538 GEOLOGY OF INDIA,

PAGE
Nandidl shales . ; . 86
Narakal, smooth water anchorage 405

Narbada river, alluvial plains . 396,

397> 431, 436

” » formed in a_ rock

basin =. —-398, 495

» Supposed former course
of . - 398
Narcondam Island. ; 17
Nari group 308, 312
» relation to Gaj 309, 312

» elation to Kirthar 309, 310
a 86

o 6 © © @ © ©

Narji limestone :

Natrolite . : . 1976

Needle shales. . ‘ + 157°

Negrais rocks . j 145, 336. 344

Neobolus beds. ; + 109, 113

Neocomian, Afghanistan . 196, 294
3 Ammonites in Sriper-

matir group 182, 207, 208

$3 Cutch 152, 207, 216, 286

si Hazdra .- « 286

3 Salt-Range . -. 286

Nepal valley, subrecent deposits 421

Newcastle beds ‘ 198, 200, 206

Nimdar sandstone . . - 253

Nodular limestone group. + 249
Nummulites, cretaceous in Bald- .

chistan . . . + «29%

ey at high altitudes in
Himalayas . 347, 477
Nummulitic group of Cutch 320, 321

ss limestone, hill type 352, 354

od sea, extent of . - 404
O
Oblique lamination; see false
bedding.

Obolus beds . . : 109, 113
Olenellus . : ‘ . . 113
Qligoclase . . . + 30, 44
Olive shales, Sind : é - 288
Olivine. . 175, 259

Ongole area of upper Gondwanas 181
Oolitic relationships of Raj-

mahal flora . . i » 193
Oolite, golden. : . + 219
Ophitic calcite in sandstone « 159
Orthoclose : + 25. 26, 30, 259

Otocerasbeds  . . : » 130

 

PAGE
P
Pachmarhi group . . + 172
Pakhal series ‘ ‘ » 89, 92, 108
»» identified with . Penganga
beds 5 . . . go2
Palzontological anomaly, creta-
ceous, Baltichistan . 21
» South India 231, 236, 238,
242, 244
»  Cutch jurassic 195, 224, 225, 229
» Kota Maléri group - 196
»  siwalik 2 . « 3061
»  Umia group ., 195, 224
Palzozoic and mesozoic, absence
of break between . 7 - 119
Palghdt group possibly an old
river valley . . . » 495
Panchet group . . 166, 170
” age 201, 204, 207, 208
5 alliances of flora 193, 194, 200,
204

3 fossil in Kota-Maléri group 189
a undecomposed felspar in 170,

207
Paniam group. ‘ . - 86
Panjal- conglomerate, glacial
origin of . 135, 136
ae »  gneissose granite
blocksin  . 3 44
Panjal system. F 7 116, 134
P4paghni group ’ . . 79
Par sandstone . . ‘ ‘ 65
Para limestone . ‘i ‘ » 130

Paradoxides beds, represented in

Salt Range . i ‘ « 133
Parihar sandstones . - « 227
Patcham group 217, 218, 221, 226
Patthargatta, supposed older allu-

vium at 3 ‘ , - 428
Pavulur group. . - 156, 181
Peat. 10, 404, 415, 421, 434
Pebbles, deformation of 49, 54> 59
Pegmatite 5 . 275 31, 58, 107
Pegu group + 336, 338, 340
Pellety laterite : . «384
Penganga beds . . . go

95 valley, glacial action in 158,

‘ 213
Peninsular area.—
rr absence of disturbance
in ‘ . » 2,53 494
SUBJECT INDEX. | 539

PAGE
Peninsular area, antiquity of drainage

system . 153, 495

55 former extension into
Assam. + 448) 488

connection with
Africa 209, 253, 493
mountain ranges 2; dy

c

o

» ”

>

a possible recent submer-
gence . . 4
Penner, alluvial gravels. + 403
Peridotite . . _ . « 346

Perim Island, upper tertiary
fauna . ; 323, 341, 429
Permian glacial period . + 207
es relationship of Bijori
fauna . . . » 195
si Gondwana floras . + 1g2

Permo-carboniferous, Salt range 121, 125
Petroleum . 292, 331, 336, 338,

343

“os in mud volcancees ‘ 20

Phlogopite . . . . 39

Pikermi fauna . : ‘ - 365

Pinnacled quartzites . ‘ : 86

Pistacite . ‘ . . 35
Plain of marine dentdatian

» €ast coast 80, 388

» west coast 386, 304

Plant fossils, uncertainty of relation-
ships in] . . . » 193
Platanista . : 7 j - 428

Plateau quartzites . . - 86

Pokaran, glacial beds and roches
moutounées . . . - 160
Poonahlite . ‘ « 260
Porcellanic beds (hacnetone) 66, 81, 95
Porebandar stone A : » 305
Potwar, erratics 14, 418, 484
Pulivendala subdivision ‘ ‘ 80
Pulkoa schists . F . . 96
Pullampet slates : . . 81
Pumice, of Puppa ‘i . . 18
Puppa, extinct volcano of . 5 17
Purna alluvial plains 306, 407
Purple sandstone 109, 113
Prehnite , . . . - 260
Productus limestone series 119, 123, 136,
206

reported occurrence in
Sulaiman hills . » 41
Prome beds é ‘ % - 339

Ee]

-Quartz amygdules

 

PAGE

Pseudo breccia or conglomerate 292, 291
. 34
Pyrites . « 26, 30, 299, 300, 303
Pyroxene . . é * 259, 266

Q

176, 260, 289
Quartz, anhydrite inclusions in. 142
» bipyramidal crystals - 10

»  trihedral 260, 28g
Quartz reefs i 27> 35, 62
Quartzose schist . ‘ ‘ 26

R

Ragavapuram shales 180, 183, 184
Raialo group 68, 69
Rain marks a . + 96
Raised beaches on Kraken coast II
Rajamahendri, infratrappeaw 268, 280
re intertrappean 270, 280
Rajmahdl group » 175
» «=o age + 203, 207
os flora os 177, 183, 188

sa » Yelation to Euro-
pean floras 192, 19}
Radjmahal series 49, 155, 17-4

pe connection in Uitenhage
series 182, 205
* outliers on east coast 152, 178.
- 232
Rajmahal trap 759 412

correlation with Sylhet

traps OI, 150, 176
» dykes in Damuda series 31, 177
o foci of eruption. - 176
"5 connection in Deccan
trap 176, 279
Ramri, mud volcanoes of - 20

Rdniganj group . . 162, 166, 167

» age. . . + 197

i coal . c . 161, 166

” flora. ‘ . - 163

Ranikot group. . . - 283
fossils in Himalayas and

Salt range . . - 354

»  felation to Kirthar 303, 396, 307

Rann of Cutch . IT, 13, 215, 4.08, 453

Red sand hills. . © + 416

Red soil . A . . - 4to

N
540 GEOLOGY OF INDIA.

PAGE
Regur ; . . . - 410

» origin ° qi q - 413
Reh . . é ‘ » 447

Resin fossil ‘ 5 ie, 297s 299. 300
Rewa group 10.0, 102, 107
Rheetic, Afghanistan . ‘. - 197
» central Himalayas + 131

relationships of Gond-
wana floras

Rilly la Montaigne, fossils com-
pared at Deccan trap . . 281

a

Ripple marks —_70, 96, 100, 113, 123, 150
Roches moutonneés, ancient, at
Pokaran F ‘ ‘ . 160
Rohtds group. 95, 96, 97, 98, 103
Ruby . . a 45, 107
Rutile . . . . ‘ 30
§ \
Sabzalkot boring ‘ » 433

St. Thomé, advance of land al ‘ 12
Sdletekri beds. . ‘ ‘ 64
Salses or mud volcanoes 5 3 17
Salt, Kohat ‘ : . . 326

x» Runnof Cutch . 3 - 454

+» Salt Range a . . 109
Salt lake, Calcutta. . . 44ql
93 Umarkot, marine shell in 430
Salt marl, . s ‘ . cg, 326
»  hypogene origin of . . Ii

Salt pseudomorph group 109, 114
Salt pseudomorphs, in dolomite of

Salt marl ‘ ; 3 110
Salt range, analogy to Assam
range . . s ‘ 8
» Carboniferous g! racial period 120, 205
» Australian fossils in 121, 493
Salt wells, Gangetic alluvium 429, 448
* Pachpadra A + 430
¥5 Purna alluvium. - 400
Sand hills 415, 455
Sandstone, flexible in Alwar group = 72
7 Talchir group . 158
Sapphire, supposed occurrence in
pegmatite. . . » 107
Saraswati river, changes of .« . 450
Sattavédu group 7 . . 156
Scapolite . ° . : e332
Schori . i ‘ : + 27, 30
Scolecite . . 3 A 260, 261
Semri beds ‘ . ‘ « 94, 96

192, 913,

 

PAGE
Serpentine rock ; . 28, 30, 80
” Afghanistan 143, 292
” Arakan Yoma 145, 146
5 Manipur . 147
Shillong quartzites . « 45, 60, 295
Shrishalam qiartzites . . 7 82
Silicified wood, Bhima - ; 88
ii Cretaceous, South
India j + 235
i Cuddalore . ° 393
i Godavari alluvium. 402
” upper tertiary 341, 346
Silurian, newer rocks coloured as,
in Himdlayas  . - 116
3 probable presence in Ten-
asserim i - 142
s supposed occurrence in
Khaibar Pass ‘ - <1
a3 sea, extent of . . - 492
Silver . . ee 8 45, 63
Simla slates ‘ 115, 117, 132
Singhe l4, nummulitic outlier . 347
Sirmur serics e : 349, 405

Sitsyahn shales . i ‘ «330

Siwalik series, age ofthe . 314, 359
” Assam, . + 332
” Burma. > 338
7 Cutch and Kéthiwar 322,
324
” fauna 316, 319; 324, 341, 359
a Sind and Balichistaén 356,
465
7 Sub- Himdlayan area 356, 465
s5 » analogy with

Vindhyan out-
liers on Ara-
vallis . » 104

3 » formed during
subsidence . 470
i » gneiss inlier in. 45

9 » relation to river
valleys. 358, 469

55 » Upper beds
cuarsest + 358
Smooth water anchorages. + 405
Soda, native sesquicarbonate of . 19
Speckled sandstone group . 120, 206
Spintangi limestone . + 307, 318

Spiti shales ‘ : + 228, 229
Sripermatur, cretaceous fossils at. 244
Sripermatur group 182, 184, 207
53 neocomian cephalopoda 182,

207, 208
Stone implements
Stony creek beds
Stormberg beds .
Striated pebbles in carboniferous

glacial boulder bed

Subdthu group

Sub-Himdlayan zone distinction
from lower Himd-

Oe aes

TAdputri beds
‘Tagling limestone

SUBJECT INDEX.

PAGE
. . . 28, 80
. + 30, 176, 260
. + 388, 399
é 7 - 198
: : + 203

120, 133, 158,
160, 199
347» 350; 357, 480

layas . - 462

encroachment on

Gangetic allu-

vium ss. » 481
inliers, hill mum-

mulitic lime-
stone . © 353
gneiss : 45
Jammu 140, 347:
: _ 3st
successive Ori-
gin of faults
in. . 480

tertiary depositsof 3 56

465

wanting in Bhutan 348
Seb-nummalitie group « 320, 321
Subsidence, evidence of — 4, 10, 407, 433
Sullavai series
Summit bed of laterite. . » 374

: + 90, 91, 108

. . + 70; 96

Sunda volcanic chain . 5 . 18

442

Sutlej river, former independent
course to the sea . “i - 449
Sutlej valley, Upper, deposits 422, 464
conditions of formation 409
Swatch of no ground

° . e 442
. ° - 260

+ 36, 44, 74s 234

See jhils, depression of the - 443

* . 61, 159, I 76

. . » 110
j i - 140
. . - 80
? e 131

. 137» 230, 480

 

541
PAGE
Talc schists, in Bundelkhand ‘ 26
Talchir group. . ‘ 157, 165
» age, : . . + 206
»  boulderbed . _ 157, 198
» flexible sandstone. . 158
» mode of origin . 151, 158, 159

» resembled by Pengangad
‘shales . ‘i . go
3 Pr transition shales 50, 62
» Western Rajputana. 106, 160
Tangis or rifts. . 7 ‘ 9
Tanawal or Tanol series. - 139
Tanr land . ‘ - 42I
Tapti and Purna alluvium x08 400, 431
Tarét . ‘ » 438
Téris or red zane hills . 5 ‘i - 416

Tertiary deposits

95 contrast of lower and upper 302
5 passage into recent 283, 318,319,
325, 391
Tesselated sandstones. Fi . 157
Thomsonite : . . + 260
Tibet, lakes . . : « 489
» peculiarities of fauna. » 476
Tin . . . : - 30, 45
Tinnevelli coast, advance of, : 13
‘i submerged forest . II, 405
Tipam sandstone . «© 324, 332
Tirohan limestone . . 96

Tongking, Gondwana pints i in » 196
Toshdm hill, felsites . . . 74
Tourmaline . ° » 30, 32
Trachyte . 17, 18, 176, 258, 278
Transition systems, age of . 72,75, 108
ss definition of term . 48

5 relation to newer
gneiss . + 24, 31, 40

. resemblance to Tal-
chirs . . « 55, 62
Trappoid beds, Ecca series . + 203
» lower Vindhyan . ‘ 95
»  Talchir group 3 . 158
Triassic, fossils in Attock slate area 115

» relationships of Gondwana
floras . ° * 192
» of Maléri fauna < 195
» supposed in Burma . 144, 336
Trichinopoli group . 233, 236, 240, 245
oA relation to Utattr 234, 235, 237
3 Ariyalur . 235, 241, 238
Tripetty group. s . 180, 186
Trivicary grits. ' . + 393
542 GEOLOGY OF INDIA.

PAGE

U

Uitenhage series ‘ «182, 205, 207
Ulster, ferruginous beds of . + 334
Ultrabasic rocks. : e 138, 346
Umballa boring . fe : 435, 476
Umia group "I . 188, 217, 223
» age : . » 207, 224.

» alliances of flora : . 164

* coal . < - 223

» relation to neocomian .152, 223, 286
Underclay, wanting in Damuda

coal seams . . . + 202
Unio, aberrant forms of ‘ - 319
Utatir group. . 233, 235, 245
relation to Ariyalir 235, 239, 241
3 9  Trichinopoli 234, 235, 237
Utatér, upper Gondwdnas . » 182

”

Vv

Vaikrita system . ‘ ~ 7,
Valimukam, submerged forest

at . ‘ ‘ 3 + 9, 11, 405
Valleys formed pari passu with

elevation of hills é . ‘ 9
Valudayur group . . + = 235
Vemdavaram shales. < 181, 184

Vempalli group . , 5 . 79
Venkatpur beds . és . - 92
Vesicular laterite . . + 383
Vindhyan era, disturbance of rocks
during . . + 491
Vindhyan system, lower. 93, 108
is debris of,in upper : 99
$5 disturbance newer than
valleysP . . . 99
on doubtful propriety of
classing with upper
Vindhyans .

e298
«94, 103

is outliers in Aravallis
53 relation to Bijdwars, 52
53 relation to upper Vin-
dhyans + 93; 96, 98, 99
Vindhyan system, upper. - 67, 98
i age . . 3 + 492
si analogy to Gangetic
alluvium . : - FO4
» local disturbance of . 102

 

PaGe

Vindhyan system, upper, outliers in
Bundeikhand . 29, 104
outliers on Ardvallis . 103

” » west of Aravalli

range . - 106

» relation to main
area. - 107

” pebbles -in Gondwana
rocks . . ° 157

3 relation to Aravalli range

71, 103
106
» system, lower Vindhyans 93, 96
98, 99

= represented by Sullavai
series . . + 92, 108
ed reputed fossils, in - 101

scarp, in Bundelkhand,
defined in Deccan trap
period. ‘ ‘ 29

Virgal group 7 F 122, 123
Volcanic action, supposed at Londr — 19
» beds, Attock slates - 116
¥ » Bengal transition . 63
” » Bijdwar + « $2, 54
a »» carbonaceous system 134
9 » carboniferous, Kash-
mir . +135, 136
‘3 », Chilpi ghat series. 64
3 » Cretaceous 148, 287, 289,
290, 292
” » Cuddapah . 3 80
” », Deccantrap. + 257
bg » Dharwar system48, 50, 499
” » Gwalior system . 65, 66
¥ ,, Jaunsar system . 117

93 » Malani. ‘. i 74

3 » Rajmahal . 175,177
us » Salt marl. 7 » 110
9 » Silurian : . 915
‘9 » Tertiary, lower 292, 293,

328, 346, 348
% » Tertiary, upper . 339

» eruption, submarine . 18
ss - neck, supposed, in
‘3 Bassein . 7 18
rr Wajra Karur 40, 107
Volcanoes, cretaceous ; » 292
3 Deccan trap . 258, 277, 279
» mud . < + "19
»  Rajmahal . 5 . 175

3 Recent 4 : < 16
SUBJECT INDEX, 543

PaGE

W

Wadhwan sandstone : + 253
Wajra Karur, volcanic neck , 40, 107
Warkalli beds A : » 299, 304
Western ghats, origin and age of

10, 397, 494
Wianamatta group . + 198,.200

Wollastonite . . . * go

 

PAGE
Wood, fossil, silcified, Bhima + 88
a % cretaceous, South India 235

7 » upper tertiary 341, 342, 355

Z
Zeolites . ‘ « 176, 260, 262, 412
Zones of amm:-nites, recognized in
Cutch jurassics . ach Ms) 224

Zoological provinces, cretaceous . 243
n Gondwana period . 193, 591