YALE UNIVER

OCT &l Wo
PROBLEMS IN EXPLORATION
CENTEAL ASIA.

By Ellsworth Huntington.

£7,3

(From ' The Geographical Journal ' for April, 1910.

El 2

rsj

PROBLEMS IN EXPLORATION- CENTRAL ASIA.
By ELLSWORTH HUNTINGTON.

(From ' The Geographical Journal ' for April, 1910.)

Since the revival of geography in the days of Columbus geographers
have been zealously mapping the world and writing empirical descriptions

* In writing this paper, the author has made use of some of the material from a
paper on " Desiderata in Exploration — Asia," by the late Mr. Ney Elias, with addi
tional notes supplied by Sir T. H. Holdioh and Sir Francis Younghusband. A

396 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
of its various portions. They have been so absorbed in this important
task that they have had comparatively little time to ponder on the
great problems of why specific parts of the Earth's surface present a
certain appearance and how they can best be described. Nor have they
done much toward determining why certain species of living beings are
grouped together, or of why special modes of human life and peculiar
types of activity are distributed as we now find them. It is not sur
prising that such has been the case, for a seienoe grows like a human
being. We do not expect a child to read while it is learning to talk,
and we cannot expect a science to devote its time to arranging and
explaining its facts while it is still expending its energies in getting
them. At last, however, the purely exploratory and descriptive stage
of geography is nearly completed, and we are already embarked upon
the explanatory stage, which bids fair to be even more interesting than
, its predecessor. The explorer who has tasted the keen joy of pushing
his way through some hitherto unknown country is apt to sigh for the
days when America, Africa, Australia and a large part of Asia were
still unknown. Those, it seems, were the halcyon days of geographers,
the time when it was joy to live. A broader view, however, shows that
the most fascinating days of geography are still before us. There is
surely as much pleasure in the discovery of an idea as in the discovery
of a lake. It is a splendid thing to cross an unexplored continent, but
it is no less splendid to ascertain why the people of that continent
possess a fundamental characteristic which does not pertain to the
people of the rest of the world.
Few portions of the Earth's surface offer problems of greater interest
than those of Central Asia. For our present purpose Central Asia may
be defined as the vast area extending from the Caspian sea and the
Zagros mountains on the west to the borders of China Proper on the
east, with the Arabian sea and the Himalayas on the south, and
the plains of Siberia on the north. It consists primarily of three great
basins and two plateaus *The basins are, first, that of Iran, including
the countries of Persia, Afghanistan, and Baluchistan ; second, the
eastern portion of the Aralo-Caspian basin generally known as Eussian
Turkestan ; and third, the basin of Chinese Turkestan or Lop. The
plateaus-comprise the vast bulk of Tibet with its westward extension
into the Pamirs, and the plateau of Tian Shan lying north of the Lop
basin on the southern confines of the great Siberian plain. In addition
to these five great features there are numerous minor basins and
mountain ranges. Climatically the entire region is characterized by a
relatively high degree of uniformity, although of course there are
marked contrasts between the highlands and the lowlands. Except for
the mountainous regions, Central Asia is a land of deserts, and is either
far removed from the ameliorating influences of the sea or else cut off
from them by mountains ; hence it is characterized by broad tracts of

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 39?
arid plain ringed about or dotted here and there with oases lying either
at the base of the highlands or along the main rivers. Summer rains,
except in the northern portions, are rare ; and dry, sub-tropical
conditions prevail from spring till fall. In the winter, however, when
the climatic belts of the Earth shift southward, Central Asia lies in the
zone of prevailing westerly winds, and in most parts receives a fair
amount of rain from cyclonic storms. The ever-present contrast
between great uplands and low plains, and the climatic uniformity of
the country, give to Central Asia a distinct unity, and cause, the main
geographical problems of all parts to be closely allied.
Pioneer Exploration.
Although the work of pioneer exploration in Central Asia is almost
finished, there are still a few interesting bits to be done, and these
should be mentioned before we discuss other problems. The two chief
regions where this is the case are Afghanistan and Tibet. A generation
or two ago Afghanistan was partialty explored ; but during recent
years it has been almost inaccessible. No one has been permitted to
travel there except under close surveillance, or on the borders, where
British and Bussian boundary commissions have been at work. Thus,
although the main features of Afghanistan are known, there are large
areas which still require exploration. For instance, the course of the
Heri Bud, the river of Herat, and the only stream which escapes from
the interior of the basin of Iran, has never been completely mapped,
and the same is true of parts of the Helmund. Farther toward the
centre of Asia, in the north-east corner of the country, large portions of
Badakshan and Kafiristan are entirely unknown. It would be hard to
find any scientific task more interesting than the exploration of this
region, and a study of its wild pagan inhabitants, the bold Kafirs and
other strange tribes. At present, however, such exploration is practi
cally impossible, because the English and Bussian Governments support
Afghanistan in her complete interdiction of the entrance of European
travellers, and because the mountaineers of Afghanistan are among the
fiercest of the Earth's inhabitants.
In Tibet there are still fairly large blank spaces upon the map, but
the major features are all quite well known with one exception.
South-eastern Tibet is as yet unexplored. The river Tsangpu of Tibet
is supposed to be the upper portion of the Brahmaputra of Assam, but
no one has ever actually proved this to be the case. Its course for
several hundred miles through the Himalayas is unknown. " Even the
Tibetans know nothing as to what becomes of their greatest river.
When asked at Lhasa to give permission for a party of British officers
to return to India down the river they readily granted it, and even
deputed an official to accompany the party to the border of Tibet. But
they added that no one knew whether the river went to India, or indeed

398 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
where it went. In Tibet it is 11,000 feet above the sea, and is a
magnificent river 200 or 300 yards in width. It runs parallel to but at
the back of the main axis of the great Himalayas. Where it is next
known it has broken through the Himalayas, and is only 500 feet above
the sea. What falls and rapids there are, and what gorges the river has
cut, remain for the explorer to tell us.
"Between the limits of Tibetan territory and the boundary of
Assam therS are a number of independent tribes more or less inimical to
the presence of Europeans. It was because of the likelihood of troubles
with these tribes that the Indian Government refused permission for a^
British party to proceed down the Brahmaputra as Major Byder and
Captain Bawling proceeded up the river after the signing of the Treaty
of Lhasa. The difficulties are now further increased by the unlikeli
hood of the Tibetans again granting permission for a party to proceed
through Tibetan territory. The psychological moment was lost. But
the relations of the Indian Government with the tribes between Assam
and Tibet may improve, and an opportunity may be found of penetrat
ing this country ; and fortunately it is in that part, and before the
Tibetan border is reached, that the solution of the great geographical
problem will be found. It is on that portion, therefore, that the
attention of geographers should be focussed." *
Minor Exploration.
In addition to the two main regions of exploration in Afghanistan
and Tibet there are a number of smaller but very interesting districts
still to be explored. One of these is the Persian desert of the Lut.
This has been traversed many times, but there are still parts which
no one knows. For instance, west of Seyistan lie the barren mountains
of Palan Kuh, the Panther range, which may perhaps contain ruins of
ancient towns or other interesting phenomena. In the Trans-Caspian
desert there are also regions which would well repay exploration. No
one has ever followed the entire course of the Tuzboi and of the TJngus,
the supposed ancient channels of the Oxus river. A thorough explora
tion of these would throw most interesting light upon the fluctuations
of the Aral and Caspian seas, and upon the conditions of ancient '
civilization. Passing eastward to Chinese Turkestan, there are many
regions both in the Takhla Makan desert and in the surrounding
mountains which are only imperfectly known. For example, in the
mountains to the south-west of the basin many vast glaciers and
lofty peaks have never been properly mapped; and the source
of the Oprang river, one of the main tributaries of the Yarkand,
remains to be discovered. In the desert of Takhla Makan, which
occupies the centre of the basin, there are large tracts which have

See previous note,

PROBLEMS IN EXPLOBATION— CENTRAL ASIA. 399
never been traversed. It is not probable, however, that much will
be found here by the traveller who is simply engaged in map-making,
for the whole region is a monotonous, sandy plain, and almost the 'only
possibility of interesting results lies in the discovery of ruins. At
the eastern end of the Lop basin many maps show a number of lakes,
Toli and others, to the south of Hami ; but they are located by hearsay,
and no one knows whether they actually exist. Possibly they were
once connected with Lop Nor, or are the remnants of some other great
lake ; but it is useless to speculate until an explorer has gone there
and mapped them. He would need to go in winter, for only then could
his camels carry ice to serve for drinking-water in the great saline
desert. A third type of exploration deals with various minor problems.
Except in the few places where the Bussians have accurately mapped
their own territory, almost innumerable mountains, streams, lakes, and
villages are not accurately shown, or do not appear at all upon even
the best maps. Off the main routes one can scarcely travel a day
without finding such. For the traveller whose main interest is in
map work there is plenty of minor exploration to be done wherever
he may travel. Such work, however, is not that of the mature
geographer. It should be delegated to assistants such as the native
map-makers of the Indian Survey, or to beginners who are making
their first journeys. The recent expeditions of Stein illustrate the
high value of such map-making under the direction of competent
leaders, who themselves are engaged in studying broad problems
requiring something more than pure technical skill.
GeograpJiical Problems.
The comprehensive study of geography, which succeeds the pre
liminary work of exploration and mapping, consists of the elucidation
of various problems dealing with the distribution and grouping of
different physiographic and animate types. In other words, the task
of the geographer is to discover how and why various kinds of
mountains, climates, and other physical features are distributed as
they are, and then to see what effect these inanimate features have
had upon the distribution of animate forms. In the broadest sense
this includes not only the physical distribution of various species of
plants and animals, but also the grouping of living forms in such a
way that each geographical province is characterized by its own particular
assemblage of animate types. It includes also the distribution of human
occupations and modes of life, and also of political, mental, moral and
religious characteristics so far as these occur in one region rather than
another because of physical environment.
Central Asia is unusually rich in problems of the kind thus
suggested, but only a few of the more important can here be

400 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
considered. For convenience they will be classified as follows : (1) in
animate or physiographic problems, including (a) orography, (b) drainage,
(c) seolian action, (d) terrestrial magnetism, and (e) climate and changes
of climate ; (2) animate or ontographic problems, including (a) the
effect of pronounced climatic or topographic conditions upon man,
and (6) the distribution of human occupations and modes of life in
response to the regional grouping of vegetation and animal life as
determined by physical circumstances. The best work can be done only
when the geographer enters upon his field of work with a clear con
ception of the type of problem or problems toward the solution of which
he hopes to make a contribution, Each problem requires a distinct
mode of treatment and an intelligent preparation by means of careful
preliminary study. It is a great mistake for the traveller to set off
haphazard, hoping that he may chance upon some interesting line of
study. To be sure, he cannot avoid meeting interesting subjects, but
the man who starts out unprepared or without a definite purpose is
almost sure to be confused by the very multiplicity of subjects which
cry out for his attention. Too much stress cannot be laid upon the
value of accurate mapping, not done mechanically, but with a carefully
framed purpose. Most of the subjects discussed in this paper require
it; for whatever else geography may or may not be, it is fundamentally
the science of the distribution of phenomena of all kinds upon the
Earth's surface. It is only when the distribution of a given class of
phenomena is carefully ascertained and mapped, that the way is open for
reliable generalizations as to causes and results.
Orography.
Turning now to specific problems, let us take up those which pertain
to orography. From Persia on the west to Tibet on the east and Tian
Shan on the north, the mountains of Central Asia are rarely of the
Alpine type. Many of them show evidences of having been uplifted as
plateaus in times so recent that portions of an old surface of erosion
still remain intact at high altitudes. So common is this phenomenon
as to suggest that in general the greater mountain regions of Asia, and
perhaps of the world, owe their origin, not to the uplift of individual
ranges, but to the upheaval of broad plateaus of the Tibetan type which
have since been carved by rain and ice and wind into forms like those
of the Alps. A proper study of this problem requires a certain amount
of geological training, but it is primarily a problem for the geographer.
It would be of the highest value if some trained explorer were system
atically to traverse the great central mountain masses of Central Asia
from the Caspian sea to China, and to estimate how large a portion con
sists of genuine plateaus where remnants of ancient lowland surfaces
of erosion now remain at high levels. In the Pamirs, for instance,
many high mountains of apparently diverse geological structure have

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 401
surprisingly flat tops. In the western Tian Shan region tilted pene
plains lie at an altitude of from 10,000 to 14,000 feet, and further east
they lie at somewhat lower levels. A systematic study of these plateaus
and of the forms which they assume under the influences of erosion
would go far toward solving the problem of the origin of mountains.
In conjunction with such a study of plateaus it would be possible
to investigate two other types of mountains, both of which occur in
Central Asia. One of these is the so-called " block " type formed by the
faulting and uplifting of blocks of the Earth's crust. How widely such
mountains are distributed no one knows, because most travellers, being
untrained in modern physiographic methods, have paid no attention
to the matter. They have merely seen and mapped mountain ranges
which they describe, perhaps, as having noticeably straight base lines
and steep slopes. Further than this they have not gone, because their
purpose has usually been mere exploration and not the elucidation of
problems. In north-eastern Persia, in the province of Khorasan, block
mountains are developed upon a large scale, and the same is true to
a less degree in Chinese Turkestan west of Kashgar and in certain
parts of the Tian Shan region, where they have been seen near Lake
Issik Kul and in the peculiar basin of Turfan, which lies below sea-
level. They appear to be of rather common occurrence, although, the
individual blocks are not usually of large size. It is probable, however,
that a careful study would show that there is every grade of transition
from vast uplifted plateaus, such as Tibet, through smaller plateaus, suoh
as the western part of Kuraktagh, in the eastern Tian Shan region, to
little tilted fault-blocks only a few miles long and a few hundred feet
high, such as occur in the basin of Turfan.
Fault-block mountains are closely related to folded mountains, suoh
as parts of the Siwaliks of Northern India, and the two would naturally
be studied together. The only difference is that in the one case the
Earth's crust yields to a strain and breaks, whereas in the other case
the crust is simply bent. Begions of faulting and of folding are often
closely associated. For instance, fault-block mountains lie a day's
journey to the west of Kashgar, and folded mountains a day's journey
to the north. The latter are very young — so young that the original
form given by folding seems still to be preserved, in spite of the fact
that the low gently arching ridges are composed of unconsolidated
alluvial deposits which are very readily eroded.
The problem of the origin and form of mountains is the reciprocal
of the problem of the origin and growth of basins. A study of one
involves a study of the other. At the present time there appears to be
a tendency for the mountainous areas of Central Asia to increase in size
at the expense of the basins. Along the borders of the Aralo-Caspian
drainage area it seems as if in recent geological times a certain amount
of territory had been heaved up, and thereby removed from what was

402 PROBLEMS IN EXPLORATION- CENTRAL ASIA.
once the floor of the basin. The same appears to be true in the Lop
basin. The best way to study this problem would probably be to take
some basin and make a complete survey of its mountainous borders. It
would be necessary to zigzag back and forth many times from mountain
to plain. Such a course is naturally much more difficult than to travel
straight along the base of a mountain range, but it is far more profit
able. A properly equipped expedition might start at the Caspian sea
and travel eastward along the mountainous borders of the Aralo-Caspian
basin, and then go over into Chinese Turkestan. In Chinese Turkestan
the main caravan might follow a route close to the foot of the mountains
clear around the basin, while more lightly equipped parties could make
side trips into the mountains at frequent intervals. In this way it
would be possible to map the border of the basin and to determine the
character of the uplift by which the plateaus and mountains were
separated from the plains. It would also be possible to judge how
recently the processes of elevation have been at work; to ascertain
whether they are still in progress ; to see what connection they have
with volcanic outflows or hot springs; and to determine the relation
of folded mountains, block mountains, and uplifted plateaus. Such a
piece of work would require as much hardihood and as careful planning
as any great piece of pioneer exploration in the same region ; and its
results would add enormously to our knowledge of the origin and
structure of the irregularities of the Earth's crust.
Drainage.
Closely connected with orography are various problems of drainage.
For instance, in Northern Persia the drainage follows a peculiar
rectangular pattern. The rivers flow eastward or westward in broad
valleys between fault block mountains, and then break northward and
southward through narrow gorges for a few miles, and again turn east
or west, sometimes repeating the process several times. Again, in the
Kwen Lun mountains south of the Lop basin, there are numerous large
rivers which rise back of the mountains upon the Tibetan plateau, and
break northward in stupendous gorges in the Tian Shan mountains.
On the other hand, many rivers flow for long distances parallel to the
main trend of the plateau, and escape to the plains without crossing
any range at right angles. It would be highly instructive if careful
maps were made of these rivers, which to-day are in many cases mapped
largely by guess. It is not to be expected that such a study should
reveal new principles of such far-reaching importance as those to be
derived from the study of orography, but much of value can be learned,
and the two problems could be readily studied in conjunction.
A more unique problem in reference to rivers is furnished by the
streams which appear and disappear in piedmont deposits of silt and
gravel. The Niya river, for instance, in Chinese Turkestan, disappears

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 403
and reappears three several times, and many other rivers disappear at
the base of the mountains and come to light some miles away. In all
such cases the quality of the water where the river again comes to light
differs more or less from that of the same stream in the mountains. Some
times the water becomes so saline that it is unfit for use in irrigation ;
in other cases where, the stream does not entirely disappear, it never
theless loses so much in volume that the sediment which it contains is
concentrated, as it were, and the river becomes a stream of liquid mud.
Measurements of such rivers ought to be made to discover at what rate
they lose volume during different seasons ; and the salinity should be
tested by taking samples of the water for analysis. The practical
importance of this for the people of the lands under consideration is
great, for many a time the poor natives of Central Asia have tried
to use for irrigation water which was so saline as to ruin the crops.
Aeolian Action.
There is probably no part of the world where aeolian action can
be studied to greater advantage than in Chinese Turkestan. In
Persia and Trans-Caspia there are also abundant opportunities for the
study of the peculiar phenomena due to the erosive action of the wind.
An explorer might well spend many years on this problem. Through
out the spring season of high winds in Chinese Turkestan, or during
the summer in Eastern Persia, he might make most valuable measure
ments upon the rate of movement of dunes. Close to one another,
in regions such as that of Endere in eastern Turkestan it is possible
to find dunes which range from a height of 5 feet to 400 feet. The
differences between the motion of the small and the large must be
great and instructive. Another interesting problem would be the
measurement of the size and shape of grains carried under various con
ditions. Such studies have been carried on somewhat in Egypt, hut we
require data from widely separated regions. We ought also to know
much more about the relation of sand to vegetation. It is some
times said that in desert regions the sand of its own volition, so to
speak, encroaches upon areas of vegetation and overwhelms them. On
the other hand, others assert that this is impossible without some
previous change whereby the amount of vegetation has been lessened
or a new supply of sand is continually furnished. The question can
be solved only by careful observation in many places, especially where
rivers flow permanently in the neighbourhood of large deposits of sand.
Observations should be made to see how far the rivers, aided by vege
tation, are able to make head against the sand, and how far the sand
encroaches upon vegetation, even when the plants are kept in full
vigour by the rivers.
The rate of aeolian erosion in various kinds of rocks is not, perhaps,
so important as the relation of sand to vegetation, but nevertheless it
a 2

404 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
is of much scientific interest. No one, so far as the writer is aware,
has made any extensive quantitative study of the subject. This might
readily be done in the desert of Lop by fixing marks which should
stand from year to year, and could be visited by one explorer after
another. When alluvial deposits are eroded by the wind, they are divided
into coarser and finer portions ; the coarser portions remain in the
lowlands and form dunes, while the finer portions are carried away
hundreds of miles and are deposited as loess. In Chinese Turkestau
the process of the deposition of loess can be seen in operation almost
daily during the summer, and at frequent intervals throughout the rest
of the year. It would be of great value if some explorer could establish
a number of observation points along the northern slope of the Kwen
Lun mountains, where the deposition of loess is probably as rapid as
in any place in the world. At these stations observations on the amount
and texture of the materials deposited by the wind could be carried on
while other studies were being prosecuted in the region round about.
The observer would naturally compare the amount of deposition with
the strength and direction of the wind, and he could perhaps estimate
whether the conditions of past deposition have been the same as those
of the present. Beside this, he might well devote much time to a study
of the relation between the amount and kind of vegetation upon the
mountain slopes, and the amount of loess deposited. He would probably
find that in some areas the wind is now carrying away materials which
it deposited only a short time ago, geologically speaking. He might
also discover that, in certain areas, deposition has begun very recently,
and such areas would probably lie at altitudes higher than those of the
major deposits. The correlation of changes in the rate and location of
the deposition of loess, on the one hand, and of the changes of climate
which have occurred during recent geological times, on the other, would
probably lead to important conclusions as to the circulation of the
atmosphere during the glacial period.
Terrestrial Magnetism.
Some of the problems to be solved by the geographer require special
equipment. Such is the case with the study of terrestrial magnetism.
At present, the Carnegie Institute of Washington has in the field several
expeditions engaged in magnetic observations in the more remote parts
of the Earth. Any explorer who has the time and the ability for such
work might add materially to our knowledge of this interesting subjeot.
No place is more likely than Central Asia to afford valuable results.
The great mass of Tibet standing above the general surface of the
Earth must exert an important magnetic attraction. Turfan, too, lying
below the level of the sea in the heart of the greatest continent, is a
place where repeated magnetic observations would be of high value.

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 405
It is hard to tell what results may spring from the study of magnetism.
There are indications that variations in the intensity of magnetic
attraction are intimately related to important solar fluctuations. Other
lines of evidence have been thought by some observers to indioate that
these same solar fluctuations are the cause of the slight climatic
fluctuations to which the Earth is subject from year to year. Thus it
is possible, though this is mere conjecture, that a study of magnetic
variations may lead to far-reaching results in the field of long-range
forecasts of weather. At any rate, the subject is of such importance
that it ought to receive the attention of well-equipped expeditions to
the more remote parts of the world.
Climate and Changes of Climate.
The last physiographic problem with which we shall here deal is
climate, the most important factor of physical environment in its effect
upon life. The subject naturally falls into three divisions, namely, the
olimate of geological times, the climate of the historic past, and the
climate of the present. The study of the climate of geological times
is primarily the work of the geologist ; nevertheless, it has so intimate
a relation to the work of the geographer that he cannot afford to neglect
it. In the first place, the geographer cannot properly understand the
effect of climate upon life unless he knows how long the present
conditions have prevailed and what sort of conditions preceded them.
In the second place, the climatic conditions of the past million years,
more or less, are responsible for a vast number of purely geographical
features, such as the nature of the universal mantle of rock waste which
covers all the. lands, the forms of terraces, piedmont slopes, alluvial
plains, and other features, and the preservation or extinction of basins
which have no drainage to the ocean. Moreover, in addition to all this,
it is peculiarly fitting that the geographer in Central Asia should
engage in some geological work, because he is likely to penetrate to
regions where it may be impossible to carry on regular geological
surveys for many generations.
The study of the climate of the historic past and of the present is
pre-eminently the field of the geographer. An entirely satisfactory
knowledge of present conditions cannot, of course, be obtained until
meteorological records have been kept for many years. Nevertheless,
the fragmentary observations upon winds, temperature, rainfall, and
other phenomena made by the traveller in little-known regions are by
no means without value. His most important contributions, however,
result from careful observations of the effect of specific climatic con
ditions upon the grouping of plants and animals, and upon the occu
pations, habits, and character of man — a subject to which reference
will again be made in discussing the relation of physiography to life,

40(1 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
Lake Deposits and Old Strands.
Central Asia furnishes the best of opportunities for a study of
fluctuations of climate both in. geological and historical times. Hitherto
the climate of the geological past has been studied mainly in the
glaciated regions of North America and Europe, the most notable excep
tion being the investigation of the diminished salt lakes of Bonneville
and Lahontan in the western part of the United States. It has been
proved that the glacial period was highly complex, consisting of
alternate warm or dry and cold or wet periods, but a complete record
has never been found in any one place. The advance of a large glacial
sheet almost inevitably destroys the evidences of a smaller preceding
sheet. In dry regions, however, the records of changes of climate are
much more fully preserved. In the first place, there are numerous
salt lakes, all of which act as great rain-gauges, expanding or con
tracting in response to variations in the rainfall or evaporation. The
deposits laid down in and around such lakes vary greatly, in accord
ance with the degree of aridity and the amount of vegetation in the
regions whence the waters of the lake are derived. The deposits of
one phase cover those of another, and there is little loss by erosion.
Thus a complete record is preserved, which can easily be read, provided
we know the effect of specific climatic conditions upon the nature of
deposits, and provided the buried records can be observed. It is every
where easy to study the character of present deposits, and in many
cases buried records are accessible; for in basins, such as those of
Lop and the Lut, the edges of the various deposits have been fairly
well exposed by erosion, and in Seyistan movements of the Earth's
crust have lifted them up and exposed them to a depth of hundreds of
feet. A geologically minded explorer who could devote several seasons
to travelling about the borders of the old lake-beds of Lop, Seyistan,
and the Persian Lut could reap an amazing harvest. In a few years he
could probably learn almost as much about the occurrence of climatic
changes in recent geological times as has been learned in a hundred
years in the less favoured regions where glacial studies have been
carried on. His chief work would be the making of cross -sections
showing the succession of lacustrine clays, swamp deposits, alluvial
sand, gravel, salt, and fossiliferous beds in the greatest possible number
of old lake-beds. 'To obtain complete sections he would necessarily travel
up and down many waterless valleys, such as the one leading from the
mountains of Kuruk Tagh, in eastern Chinese Turkestan, past the ruins
of Hedin's Lulan near Altmish Bulak, to Lop-nor. The . work would
require a fairly large equipment of camels to carry food and water, but
the expense need not be great. There are scores of dry valleys on the
borders of Lop-nor and the Lut that would yield most valuable results.

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 40?
Far and away the best opportunities, however, are afforded in Seyistan
where earth-movements, accompanied by volcanic flows, have in some
places raised the bed of the lake as much as 500 feet within very recent
geological times. Under present conditions it would be no easy task
to travel for months in the desert regions surrounding these lakes, and
it would be especially difficult to visit the portion of the Seyistan lake-
bed which lies in Afghanistan. The very difficulty of the task, how
ever, makes it more attractive.
The traveller who undertakes the somewhat geological task of
studying old lake-deposits will inevitably engage in the more strictly
geographical work of investigating ancient lake-strands. In both tasks
it is essential to make accurate maps and to run lines of levels.
The highest strand of Lop-nor lies about 600 feet above the present
lake upon the south side, and apparently somewhat lower upon the
north. It probably surrounded a lake 300 or 400 miles long from east
to west, and at least 100 miles wide from north to south. The strand
has been observed at only two or three points, and its actual course
elsewhere is purely a matter of speculation. To map all the strands
completely, and find whether or not they change in level, would he a
most valuable contribution to knowledge. This would involve traversing
the Takla Makan and Lop deserts along lines where for a hundred
miles at a stretch the explorer would be on new ground . It could only
be done in winter, when ice could be carried for weeks or months, and
when camels would require the minimum amount of water. At other
times the difficulties from heat and from saline water would be almost
insuperable. No one should start on such an expedition without a
thorough equipment and without previous training in less hazardous
exploration. In addition to the mere mapping of the strands, interesting
work might be done in studying the vegetation, which seems to have
a peculiar quality in the regions once occupied by the lakes. A study
of the distribution of plants might possibly in itself be sufficient to
determine the extent of some of the successive phases of the lake.
The collection of samples of soil, and a study of the porosity of the
soil, and of its capacity for storage of water, within and without the
limits of the ancient lake might lead to interesting conclusions as to
the relation of vegetation to lakes and deserts. Naturally, the
explorer who studied the strands of Lop-nor or any of the other
lakes, would zigzag more or less from outer to inner strands in
order to map all of them at the same time. In Seyistan ihe work
would probably be less difficult than at Lop-nor, so far as physical
obstacles are concerned, although the Afghan portion would be very
hazardous because of the exclusion of foreigners from Afghan territory.
Various other lakes might well be studied in the same manner, and
exceptionally interesting results would probably be obtained in the
great salt deserts of Eastern Persia. The thorough mapping of the old
A3

408 PROBLEMS IN EXPLORATION— CENTRAL ASIA..
strands of the many lakes of Central Asia affords work enough to keep
a small corps of men at work for many years.
Changes in Rivers.
Closely related to the problem of the expansion of ancient lakes
is that of the former courses of rivers. The Keriya and Niya rivers
may serve as examples of a large number of Central Asian streams
which die away and utterly disappear not far from the foot of the
surrounding mountains. Beyond the farthest point now reached by
the water, old channels and lines of vegetation show plainly that these
streams once flowed farther into the desert. Again, in Trans-Caspia,
as has been already said, the Ungus and the Yuzboi are supposed to be
ancient courses by which the Oxus river once flowed past or through
the Sea of Aral to the Caspian. A careful mapping of these and of
many other old channels would be an arduous but profitable task. The
relation of such streams to ^vegetation, and the changes which take
place in the flora of a region when the streams decrease in size,
furnish most attractive fields of study for the geographer whose taste
are botanical.
One of the most peculiar phenomena in Chinese Turkestan is the
changes which have taken place in the course of the Tarim river. In
very recent geological times, but probably long before the advent of
man, this stream appears to have flowed nearly straight east from
Kashgar, and much to the south of its present course. Later it flowed
in approximately its present course as far as long. 87° E., and then,
as Hedin discovered, continued eastward, instead of turning to the
south-east and south as it now does. At a point near the spring of
Altmish Bulak, fully 80 miles north-north-east of its present mouth,
it discharged into Lop-nor, then apparently larger than now. Little by
little during the last two thousand years the river has been shifting its
course back toward the south. Except in the lower portion, however,
it still follows a peculiar, peripheral course, much nearer to the north
side of the basin than would naturally be expected. The Cherchen
river, which also enters Lop-nor, follows a similar peripheral course on
the south side of the basin. No one has as yet explained the peculiar
course of either stream. It is possible that during one of the inter-
glacial epochs Central Asia became so dry that none of its streams
reached a terminal lake. Under such circumstances underground water
would cause a zone of vegetation to flourish at the foot of the piedmont
slopes where the Tarim and Cherchen rivers now flow, just as such a
zone now flourishes elsewhere. In course of time the desertward edge
of the area of vegetation would become higher than the portion slightly
nearer the mountains, because wind-blown sand would accumulate
among the reeds and tamarisks on the border, where the moving dunes
would be checked by growing plants. Thus the zone of vegetation

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 409
would form a shallow trough, with slopes of piedmont gravel on the
mountain ward side and a low swell of sand and vegetation desertward.
When the climate became moister on the advent of a glacial epoch,
the rivers would naturally flow farther than before. It would be hard
to flow straight out into the sand, but easy to follow the zone of
vegetation. Possibly the chief changes in the Tarim river during the past two
thousand years are due to the tendency of the stream to abandon its
present peripheral course and flow in the normal fashion near the centre
of the basin. The theory here presented is not advanced with the idea
that it is conclusive, but merely to illustrate the problem. A sound
conclusion must be based upon lines of levels run aerobes the zone of
vegetation in various places and out into the desert. It further requires
a careful study of the relation of vegetation to the movement of sand.
It is possible that the supposed tendency of the Earth's rotation to cause
a right-handed deflection of rivers in our hemisphere has something
to do with the matter. Altogether the peripheral course of the rivers
of Eastern Turkestan presents one of the most interesting among the
minor problems of Central Asia.
Having studied the terminal lakes of a river system and the ancient
channels in the desert plains, the next step would naturally be an
examination of the upper portions of the rivers to see what evidence
they present as to climatic conditions in the past. One of the most
important lines of evidence is the alluvial terraces which characterize
innumerable valleys throughout Central Asia. They form the only
land available for agriculture, and hence are of importance econo-
micaUy. Some lie in valleys whose heads have contained glaciers ;
but in many cases the valleys havei never been glaciated in any portion.
It appears as if such terraces were due to changes of climate. Apparently
in moist periods the mountains were well clothed with vegetation,
but in dry periods the vegetation died. Then the soil which had
accumulated upon the slopes, and had been held in place by roots, was
carried off, for there was nothing now to protect it from erosion. Thus
the minor tributaries brought down such heavy loads of sediment that
the larger and more slowly moving streams were quite unable to carry
it all away. Accordingly, deposition took place, and broad alluvial
plains were built up. Sometimes the gravel and sand reach a depth of
500 or 1000 feet. When the mountains were wholly denuded of their
rock waste, or when the climate changed so that plants grew once more,
the activity of the streams changed ; being now no longer heavily
loaded, they were able to erode and to cut gorges in the alluvium which
they had previously deposited. Thus terraces appear to have been
formed. This theory needs to be carefully tested, partly because of its
importance in the understanding of such interesting land forms as
terraces, and partly because of its practical bearing on the problems of

410 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
forestation and deforestation. It would be well worth while to carefully
map the terraces of a large number of streams in various parts of
Central Asia, and at the same time to ascertain the conditions under
which they were deposited, and under which they have been eroded.
Glaciation.
In the lofty valleys near the heads of the main streams of Central
Asia, glaciers have given rise to striking phenomena. These too, like
the lacustrine strands, abandoned river courses, and alluvial terraces,
have never been adequately mapped. Glaciers many times as large as
the greatest in the Alps are still unnamed. The mapping of existing
glaciers is an interesting matter, but it is even more important to map
the moraines, which indicate the extent of ancient glaciation, and which
show how many times the ice has advanced in various places and how
far. The lakes of the Tibetan and Tian Shan plateaus are known to
be very numerous, and it is certain that many of them are due to glacial
action. As yet, however, we lack exact data as to the origin, nature,
and history of most of them. This problem should be solved in con
nection with the mapping of ancient moraines. It has often been
surmised that some parts of Tibet may at one time have been covered
by a small glacial ice-sheet, and no one has as yet disproved this possi
bility. It seems probable that such local ice-sheets may have existed,
but this can be proved only by a careful mapping of the glacial regions
of the high plateau.
The preceding discussion of lake strands, ancient river courses,
alluvial terraces, and old moraines, centres about the problem of climate.
In order to obtain the most trustworthy results, the explorer should
determine the relation between these four types of phenomena. In
other words, it is essential to discover exactly which strands, which old
channels, and which terraces or portions of terraces were formed at the
time of certain moraines. To accomplish this the traveller must deter
mine carefully the exact number of the phenomena of each kind, and
also their relative size and age. He must endeavour to find terraces
which end at the strands of lakes, and others which in their upper
portions are associated with moraines. A single instance of such
relations will not suffice. It will require long and careful work to
find enough instances to prove either that the moraines, terraces, and
strands are due to a single climatic cause, or that they are unrelated.
In this connection it may be pointed out that the study of terraces
affords scope for most careful observation and theorizing. The adjust
ment of a stream to its grade and velocity and volume is so delicate
that very slight changes may produce important results, and the results
may differ in different portions of the course of a stream. Thus it is
possible that a given dry period may cause deposition in one region
and erosion in another, and a certain wet period may do likewise.

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 411
Hence, the explorer who attempts to solve the climatic problems of
recent geological times in arid regions must be continually upon his
guard. He must follow many streams from head to mouth in order
to ascertain what changes take place coincidently in various parts of
their courses. Historic Changes of Climate.
The most interesting climatic studies are those which deal with the
climate of historic times. Opinion is much divided upon this subject.
To many travellers it seems probable that there has been a steady pro
gressive desiccation of Central and Western Asia, and perhaps of other
regions, throughout historic times. Other authorities, many of whom are
meteorologists, claim that there has been no such change. The climate
of the Earth, they say, has been practically uniform for several thousand
years ; the only changes have been slight and relatively unimportant
fluctuations about a mean which is the same now as in the past.
Between these two theories lies another. According to it, the climate of
to-day averages distinctly warmer and drier than that of two thousand
years ago ; but the change from the past to the present has taken place
in pulsatory fashion. During certain epochs there has been a degree of
aridity greater than that of to-day, while other epochs have been
characterized by relative moisture. According to those who believe in
steady progressive desieoation, or in pulsatory desiccation, the difference
between the conditions of civilization in Central Asia two thousand
years ago and to-day is due largely to deterioration of climate. Those
who believe in pulsatory changes think also that each major pulsation,
lasting for a few centuries or a thousand years, as the case may be, has
produced important effects upon the distribution of population, and
upon historic events, such as wars, migration, and the fall of empires.
Believers in the theory of climatic uniformity, on the other hand,
naturally deny that climate has been responsible for any such results.
Between these three theories choice must be made, and it is incumbent
upon geographers to gather the facts which shall lead to a positive
conclusion. In this problem, as in others, mapping plays a highly important part.
One of the greatest services which the explorer can render to science is
the careful mapping of areas of ancient human occupation and cultivation
as compared with present areas. In Trans-Caspia, for instance, numerous
ruins lie. in places which are to-day uninhabitable. In a still larger
number of cases ancient ruins are associated with modern villages, but
the ruins indicate a much denser population than to-day. The habitable
districts of Trans-Caspia have been fairly well mapped, but there are
no good maps showing in detail the areas of ancient civilization. In a
limited region it is quite possible that various accidental causes, such
as the deflection of rivers or human carelessness, may have caused

412 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
a diminution in the area of cultivation without a diminution in the
water-supply. Taking the country as a whole, however, such cannot
be the case. The Eussians to-day have introduced a degree of peace and
security which permits the full utilization of all the available water-
supply. If it should be found that for the country at large the area of
ancient cultivation was notably larger than that of to-day, it would go
far toward establishing the fact of an important change of climate. In
Persia and in Turkestan the same phenomena exist upon a still larger
scale. In those countries not even the areas of modern cultivation
have been accurately mapped except in a few cases, and much less the
areas formerly utilized.
In all such studies constant and vigilant attention must be given to
the date of the ruins under consideration. Again and again we are told
that ruins cover an area much larger than that which can be cultivated
to-day, but this, it is said, does not indicate that the supply of water
was once greater than now. The ruins date from various ages, and
those of any one age indicate a population no greater than that of
to-day. Hence it is necessary to find, not only how great an area is
covered by ruins in a given portion of the desert, but also how far
those ruins indicate that the whole area was under cultivation at once.
Much help in this problem will be obtained by a study of ruins which
lie far from sources of water or from areas now under cultivation.
Measurements of rivers and of bodies of water in piedmont plains, such
as those suggested in an earlier paragraph, will give data to show how
much water could reach a given ruin if the present stream were carried
past the oasis of to-day to the more remote sites of habitation occupied
in the past. Near Niya, in Chinese Turkestan, the most distant ruins
lie in the desert, some 60 miles beyond the present oasis. Apparently,
if the water which now irrigates the town of Niya were allowed to flow
as far as possible, it would not suffice to reach the ruins, even if none
of it were used in the modern oasis ; and of course there would be none
whatever to support the extensive gardens and fields which once occu
pied many square miles of what is now pure desert. It is probable that
there are many other places where phenomena of this sort exist, and
where accurate measurements ought to be made.
Another line of study is the investigation of ruins in connection with
lakes and swamps. For instance, tradition relates that the ruins of
Sabari lie upon the bottom of the northern portion of Lake Sevistan.
The natives say that not many years ago the lake dried up sufficiently
to expose the ruins for a month or two. Near the shores of the
Caspian sea various ruins lie beneath the present level of the water.
Other ruins at Baku and Derbent and elsewhere on the shores of
the Caspian show marks which indicate that the water once stood
much higher than now. At Issik Kul, in Southern Siberia, among the
Tian Shan mountains, ruins are reported to lie beneath the level of the

Problems in exploration-central asia. 413
water, a faot which suggests that the lake once stood lower than now.
Far to the east, Stein reports that the westward continuation of the
Great Chinese Wall passes marshes which once were lakes. Here one
can plainly see upon the wall itself marks indicating the level to
which the water rose. There are probably many other cases of ruins
either within the present limits of lakes or lying above them and
marked by the action of the water. In regard to most there are
traditions, and in many cases actual historic records indicate the
time to which the ruins belong or during which the lakes rose. By a
study of such phenomena it is possible to estimate with a fair degree
of accuracy the exact course of the rise and fall of enclosed lakes, and
consequently of changes of climate.
In all the desert regions of Central and Western Asia there are
numerous traditions or records of old roads which are not now used.
_ In the days of the glory of Kome, great trade routes passed through
Afghanistan and Persia in places where no commercial caravan thinks
of going to-day. At the dawn of history vast caravans, carrying the
trade of China to the west, followed a route which is to-day impracticable
because of the absence of water and fodder. In like manner old roads
once crossed the sands of the Takla Makan desert, the vast waste places
of Arabia, and the dry stony expanse of the Syrian desert. Some of
the roads have been abandoned for one cause, and some perhaps for
another. It would be an interesting matter for an explorer to set out
with the purpose of following as many of these old roads as possible.
He could then determine how far they are to-day practicable, and how
far a change in physical conditions has rendered them no longer passable
for commercial caravans.
In studying the problem of climate, the effect upon plants and
animals is almost as interesting as the effect upon man. In the Kuruk
Tagh of Chinese Turkestan, near Turfan, the piedmont slopes of gravel
among the arid mountains are marked by old game paths leading to
springs now dry. At some time, so recent that the paths are still
preserved, the springs must have contained a relatively permanent
supply of water, although now they contain water so rarely that animals
visit them only at long intervals. Here and there shelters for hunters
stand beside the desiccated springs, showing that wild animals fre
quented the place since the coming of man. It is highly probable that
other evidences of this kind might be found, and that a study of the
distribution of animal species in the past as compared with the present
might lead to interesting results. This is one of the many problems
which have as yet received practically no attention. It would hardly
be wise to undertake an expedition whose chief purpose should be
such studies, but the geographer who has the matter in mind is apt to
pick up facts which may be of much importance when interpreted in
connection with what is learned by others.

414 PROBLEMS IN EXPLORATION— CENTRAL ASIA.
A study of vegetation and its relation to climate is sure to yield
large results. Deforestation is one of the subjects of greatest importance
in all the drier portions of the world. We are told, sometimes, that
the cutting off of forests causes important changes in rainfall, but
meteorologists almost universally deny this. We are further told that
the cutting off of forests has been one of the great causes of the fall of
nations and the decay of civilization in various parts of Asia. Numerous
traditions and historic records unquestionably indicate the former
presence of forests in places where none now exist. This subject should
be studied by an explorer trained in forestry and in climatology. He
should go to the areas where historic records prove the former existence
of forests, and should there examine carefully the possibilities of
forestation at present. He should collect data as to the depth of soil
and especially as to the occurrence of summer rains, which are the
most important of all factors in determining the location of forests. In .
addition to this, he should study especially those areas where trees, or
other forms of vegetation, have died off and disappeared without the
intervention of human agency. Such areas are found especially in
Chinese Turkestan, where hundreds and even thousands of square miles
of the zone of vegetation are studded with the gaunt skeletons of
poplars, which have died from lack of water, and with great tamarisk
mounds, whose peculiar growth is apparently due to the same cause.
Elsewhere, for mile after mile, the stalks of dead reeds cover the ground
in profusion, showing that where now all is dead and barren, vast beds
of waving reeds once stretched green and gold to the horizon. Again,
in other regions, traces of forests are reported high among the moun
tains beyond the present tree-line. Here and there traces of ancient
cultivation and of fruit trees and poplars are found in places like
Endereh, in Chinese Turkestan, where to-day the water, which must
have been used for irrigation, is so saline as to prevent the growth of
such trees. Apparently such regions were once decidedly moister than
now, and a change in climate has increased the amount of evaporation,
thus causing the water to become more concentrated, and hence more
saline and unfit for agriculture. A journey through Asia for the
express purpose of studying the relation of forests and of vegetation to
topography and rainfall, and of investigating the changes which have
occurred in vegetation during historic times, would be as fruitful as a
journey well could be. It requires, however, a man who is not only
a geographer in the sense of being able to travel and make maps, but
also a man who has a thorough training in the ecological or geographical
branch of botany.
Before leaving the subject of historic changes of climate, attention
should be called to the necessity of systematic search, not only for
evidences that the precipitation of the past was more abundant or
the temperature lower than at present, but also for evidences of less

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 415
precipitation or higher temperature then than now. For instance, near
the lake of Son Kul, in the Tian Shan mountains, there are a number of
old canals evidently designed for irrigation, but lying in a region where
to-day agriculture is utterly impossible. Snow falls almost every
month in the year, and the brooks are skimmed with ice in mid-July.
Nevertheless, at some time, crops appear to have been grown here. On
the other side of the world, in Mexico, ruins of villages are reported
in places which are to-day so cold that agriculture is impossible. It is
possible that in Tibet, or in the high mountains of Tian Shan, or in the
Altai other such places exist. If they do, they point strongly to an
historic epoch when the climate was distinctly warmer than at present.
In all studies of historical as well as of geological changes of climate
the problem of correlation is fundamental. Just so far as possible, each
phenomenon must be accurately dated, and its relation to other pheno
mena must be investigated. For instance, the structures which lie
below the level of the Caspian sea, and are now submerged, appear to
be of two ages and to indicate two different periods of exceptional
aridity. It ought to be possible to determine whether they are con
temporaneous with the reported ruins in the lake of Seyistan and with
the canals of Son Kul.
Effect of Pronounced Environment upon Man.
To the majority of mankind the most interesting portion of
geography is the ontographic portion, dealing with the relation to
physical environment to life. An understanding of the influence of
environment upon human habits and character can be obtained most
easily by the study of primitive people, living under extreme con
ditions of any type, but especially under extreme conditions of either
climate or topography. Hence the nomadic inhabitants of deserts, the
dwellers in oases, or the people of lofty secluded mountains furnish
peculiarly profitable subjects of study. We find that in such cases
peculiar habits and customs are generally localized in small areas.
For instance, a certain tribe in the Sahara desert is noted for its
honesty, never stealing as its neighbours do. The explanation is
said to be that the tribe lives in the sand, where it is very easy to
follow tracks; hence a thief is almost sure to be detected, and honesty
has become the rule. Again, on the famous Karakoram route from
India to Chinese Turkestan, it is a point of honour with the native
travellers never to steal any articles found upon the road. Time and
again caravans on the vast uninhabited plateau, at an elevation of from
15,000 to 18,000 feet, are overtaken by storms or are incapacitated by
the death of the pack-animals. They leave their bales of goods beside
the road in full view, confident that no one will touch them, and that
they can return and get them the following season. This peculiar
form of honesty has perhaps grown up because each man who travels

416 PROBLEMS IN EXPLORATION-CENTRAL ASIA.
this baleful route is liable to suffer loss himself, and thus there has
arisen a feeling of sympathy for the losses of others. A good example
of the effect of different kinds of environment is seen in the habits
of various people as to so simple a process as bread-making. The
Turkoman of the Transcaspian desert chooses a sandy spot for his
fire, and when the sand is hot brushes away the coals, smooths off
the ground, and spreads the dough upon it. Then he pokes the coals
mingled with heated sand back upon the dough, and leaves the bread
to bake. The Afghan, living among stony wastes, oftentimes heats
a stone in the fire, pulls it out, covers it with dough, and throws it
in again. The dweller in the smooth alluvial plains of Persia makes
an oven of mud and heats it with bushes. In all these cases the con
trolling factor of environment is climate. It is the low temperature
of the Karakoram route which makes it so difficult. It is the long
prevalence of aridity which has given rise to the sandy deserts of
Transcaspia and the rocky piedmont slopes of Afghanistan. Topo
graphy, too, is highly important, for the altitude of the Karakoram
plateau causes its low temperature, and the flatness of the Persian
plains causes them to consist of fine clay rather than sand or gravel.
Another type of human response to geographical environment is found
in architecture. For instance, in Eastern Persia, and in other regions
where the climate is dry and where the mountains are so low that they
furnish practically no water for irrigation in summer, it is impossible
to raise trees, which require moisture during a long season. Hence the
inhabitants have adapted themselves to their surroundings by learning
to build houses composed entirely of adobe bricks, and having domed
roofs. On the other hand, at the foot of the Kwen Lun mountains in
the zone of vegetation, in Chinese Turkestan, there is no more rain
than in Persia. The mountains to the south, however, are so high as to
be covered with snow all the year. This gives rise to great floods which
persist well into the summer ; hence, even in the latter part of the
season, there is water enough to foster the growth of trees. Therefore
the Chanto of Chinese Turkestan uses wood abundantly in his house,
and employs clay merely for plaster, as a rule. Sometimes, indeed,
along the lower courses of the rivers, he cannot use mud even for
plaster. The salinity of the soil is so great that the clay, even though
of fine texture, will not stick together ; hence it is impossible to make
adobe bricks or to employ it as plaster to stick in the chinks between
the logs. The Chanto can keep the cold winter air out of his house
only by the poor expedient of putting layers of tamarisk twigs or of
reeds between an outer and an inner set of roughly hewn poplar-beams.
Hundreds of other examples of the localization of special habits, or of
special modes of life, because of peculiar climatic or topographic conditions
might be cited. Many such are recorded in books of travel ; but in the
majority of cases, while the peculiar habit is noted, the cause is not

FR0BLEMS IN EXPLORATION— CENTRAL ASIA. 417
explained. The geographer whose tastes are ethnological has before
him an almost unlimited and a most fascinating field.
Migrations.
The readiness with which mankind, especially in primitive con
ditions, moves from place to place in response to climatic conditions
is extraordinary. Year after year outlying areas on the edges of the
oases of Central Asia are abandoned or reclaimed in response to seasonal
variations in rainfall. Maps of these regions at intervals of a few years
would be an interesting means of obtaining a rough record of rainfall in
places where no instrumental records can be kept. Annual variations of
precipitation sometimes cause movements of population on a somewhat
large scale. During the dry years of the seventies various tribes of
Dungans and others changed their habitat because of scarcity of fodder
for their animals. In some cases they travelled hundreds of miles, and
came into warlike conflict with the people along their route. The
number of such minor movements must be great, and careful inquiry
would probably enable us to correlate them with climatic cyoles such as
the thirty-five-year periods of Bruckner. There is probably no geo
graphical problem more important than a detailed study of the effect
of cyclical variations of climate. In certain regions the rainfall of the
moist portion of a Bruckner cycle is from 50 to 100 per cent, greater
than that of the dry portion. Such a difference may cause the limits of
profitable wheat-raising, for example, to shift through three or four
degrees of either latitude or longitude. In every generation hundreds
of thousands and probably millions of the Earth's human occupants are
plunged into distress because of ignorance of the fact of climatic cycles.
Every bit of information which the traveller can gather on this subject
has a direct practical application.
In connection with movements of population, one of the subjects
which presents a broad field of study is the routes of seasonal migrations
and raids. Data as to these are easily obtained, for the nomads are
always ready to tell where they spend the winter, where they move in
spring, and where they go in summer. The geographer's part, here as
in so many other cases, is first to map the migrations. Then he is
prepared to see what peculiar characteristics they possess, and why
special routes are followed. An interesting problem would be to see
what relation exists between the summer and winter camping- grounds.
Do people who live during the summer in lofty plateaus so cold that
nothing will grow except grass, move to any special type of region
in the winter, or are their winter camps determined merely by acci
dental considerations, such as proximity ? Among tribes such as the
Afghans, who are in the habit of making raids, it would be profit
able to ascertain exactly the seasons at which raids are most common,
and the causes. This, like many other subjects, is one upon which

418 PROBLEMS IN EXPLORATION— CENTRAL ASIA,
more or less has been written, but it has never been studied systemati
cally. Among the Arabs the months of May and early June are the
great time for long raids. The reason is that at that time the mother
camels can be taken away without injury to the young, and their milk
can be used to support not only the raiders, but the fleet mares needed
for the final dash against the enemy. Similar customs, due to the
changing seasons, must exist in many places.
Distribution of Modes of Human Life.
The last geographical problem which we shall consider is the distri
bution of various modes of human life. In Central Asia, because of the
contrasts due to the aridity and great relief, we find different modes of
life in interesting and suggestive juxtaposition. One of the tasks to be
accomplished by the explorer in Central Asia is a mapping of the
location of each type, and a discussion of the reasons for its location.
Most parts of the world are divided more or less distinctly into linear
zones running parallel to the main mountain axes. This is especially
true in dry regions. The first zone consists, of course, of mountains
or plateaus ; then come slopes of piedmont gravel, and beyond these a
zone of vegetation or a strip where rivers spread out and oases are
located. Between or beyond the oases the sandy desert begins, inter
rupted sometimes by rough tracts where deposits of clay are being
dissected by the wind, and sometimes by bands of vegetation along
streams which are of such size that they have not disappeared in the
zone of piedmont gravel. Finally, beyond the sand one finds a terminal
lake located in the lowest part of a basin and surrounded sometimes by
great stretches of salt, and sometimes by vast swamps full of reeds.
Each of the zones has its own peculiar people. Among the mountains,
provided there are plateaus or grassy slopes, the population consists of
nomads, such as the Kirghiz or the Kurds. In the valleys village-
dwellers of the same races cultivate a few scanty fields for half their
living, and for the rest drive their flocks into the mountains and lead a
semi-nomadic life. Hardy people and bold they usually are, and prone
to plunder their richer neighbours. At the foot of the mountains the
zone of piedmont gravel is sometimes habitable, but never rich, and
often utter desert. If it has any inhabitants, they are apt to be
nomads or else poor villagers, who dig tunnels into the gravel and
thereby extract a scanty supply of water from its depths. The wealth
of the country is concentrated in the oases of the zone of vegetation.
Oftentimes the zone is by no means continuous, and is very irregular,
its distance from the mountains depending upon the size of the rivers.
Nevertheless, it has a genuine existence in almost every arid region.
Life in its shelter is relatively easy ; there is usually slight danger even
from nomadic raiders ; and there is little to tempt the inhabitants away
from their pleasant oases. Hence there develops not only a type of civil

PROBLEMS IN EXPLORATION— CENTRAL ASIA. 419
organization entirely different from that of desert or mountain tribes,
but also a weaker or softer type of character. Out in the sand life
once more becomes hard, and there are few inhabitants — nomads, depend
ing chiefly upon camels and sheep. Finally, in the swamps around the
lakes the most peculiar of all the races are located— the fowlers of
Seyistan, depending for a livelihood upon the birds which they catch in
nets; or the Lopliks of Chinese Turkestan, depending largely upon fish.
To map the various zones of Central Asia and the distribution of the
different modes of life is one of the most fruitful and fascinating tasks
of the geographer. When he has done so he is ready to discuss the
causes which govern the distribution, not only of man's varied modes
of life, his occupations, and his habits in respect to physical matters,
but also of mental and moral characteristics. This leads directly to
one of the most profound and fundamental of all subjects of study, the
great problem of how far geographical environment is the controlling
factor in human life, and how far other causes operate. The solution
of this problem is not, perhaps, the province of the geographer ; but it
is his duty to prepare the facts which shall lead to its solution. Having
mapped the distribution of various modes of life, he may well supple
ment his work by making anthropological measurements like those
so well made by Stein, and by a study of the distribution of races,
languages, and religions. He is then prepared to make comparison
between the inhabitants of similar regions who differ in race, religion,
or language. Knowing accurately the physical conditions of life, and
being familiar with the customs or manners of the inhabitants, he can
judge how far uniformity of physical conditions induces uniformity of
life. The number of problems which await the geographer in Central
Asia is almost infinite. It has been possible in this article to touch
upon only a few of them. Exploration, in the common sense of mapping
new areas, is to-day relatively unimportant. Exploration for the purpose
of solving the great problems of the structure and form of the Earth's
surface, the distribution of physical types, and the relation of physical
environment to the grouping and character of living organisms presents
a field whose possibilities are boundless.

PRINTED BY
WILLIAM CLOWES AND SONS, LIMITED
LONDON AND BECCLES-.

08837 1837