\i^^^¥^^i^^^-.'^^^^m tm.,, wm 1 1^ t ! 1 a||M°y^£^|j^W^HW HBlfflHUffl! ^HiimHi ffi^HHHHHH^K^Wnrml iiiliirili !i n nWrM^^ i % ^^ IS Digitized by tine Internet Arciiive in 2010 witii funding from Boston Library Consortium IVIember Libraries littp://www.arcliive.org/details/geologyofbostonbOOcros lY. GEOLOGY OF THE BOSTON BASIN WILLIAM 0. CROSBY. YOL. L PART I.— NANTASKET AND COHASSET. BOSTON: BOSTON SOCIETY OF NATURAL HISTOEY. ■1893. Li£ it of Plates. Plate ^ 1 in pocKet. 3, in poc. cet . 3 Frontigpieee* 4 Facin^^ page 43 5 do. do. 46 do. do. 76 7 in poc cet a in pOCi ^et. 9 in pOCi cet. 10 Facing pa^e . 232 M do. do. 244 12 do. do. 272 13 , in ,./ -^ 14 pOC: cet. 15 Facing page 328 13 do. do. 340 17 do. do. 398 ia do. do* 398 19 do. do. 398 20 do. do. 414 21 do. do. 423 Z2 do. do. 432 23 do. do. 442 24 do. do. 448 25 in pOCi riet. 23 Facing pa^^e 453 27 do. do. 483 28 do. do. 558 29 do. do. 558 Lie .^ 30 Facing pa£e 590 31 At end of volume 32 do. 33 do. 34 do. 35 do. 3S do. 37 do. 38 do. 39 do. Occas. papers Bost. Soc. Nat. Hist. IV. Plate 3. <'^y''frmT> GEOLOGY OF THK BOSTON BASIN. WILLIAM 0. CROSBY. JN TWO VOLUMES. VOL. L PART I.— NANTASKET AND C0HAS8ET. BOSTON : BOSTON SOCIETY OF NATURAL HISTORY 1893. The expense of ijublisking this part is defrayed chief y hy Mr. Thomas A. Watso7i. QOSRWCOUfQE FEB 05 NANTASKET AND COHASSET. INTRODUCTION. It is obvious upon the most casual observation that the town of Hull, Mass., which is virtually an island, being joined to the main land of Cohasset only by the narrow barrier beach con- nectino; the Green Hill drumlin with the OTanite ledg-es alono" the Jerusalem Road, is sharply divided into two districts which are very strongly contrasted in tlicir topographic and geologic features. In the order of geologic age and interest these are : (1) The highly irregular, broken and rocky tract, commonly known as Nantasket, which forms the southern extremity of the town, adjoining the mainland, and is almost completely isolated by Strait's Pond, Weir River Bay, Nantasket Harbor and the Atlantic. (2) The narrow, and, at some points, extremely « slender peninsula stretching from Atlantic Hill north-northwest for more than three miles to Point Allerton, whence, turning at a right angle, it extends to the westward two miles further to Pemberton and Windmill Point. This peninsula thus embraces Nantasket Beach and the Village of Hull ; and, with the excep- tion of the single ledo-e of slate on the south side of Thornbush Hill, near the Village, it is composed entirely of rounded drift hills or drumlins and the connecting beaches of sand and shingle. The dividing line between these two districts is where the sands of the beach rest against the rocky northern base of Atlantic Hill. Although the main purpose of this paper is to set forth the results of a detailed study of the intricate geologic structure exhibited in the magnificent rock exposures of the southern area OCCAS. PAPERS B. S. N. H. rv. 1 (Nantasket), the concludiDg pages, especially, relating to the glacial and more recent phenomena and the non-lithified de- posits of the district, will apply to the northern or beach area as well. Hence the scope of the paper really embraces the entire area of Hull, or Nantasket in the broad, original application of the name.^ The mouth of Weir Kiver, or Weir River Bay, forms, in the topographic sense, the natural western boundary of the Xantasket area ; yet, geologically, it is impossible to exclude the promontory of Rocky Neck, in Hingham. This rocky head- land is lithologically and structurally identical with the district east of Weir River Bay, while it is completely separated from the other and dissimilar sedimentary areas of Hingham by the ledo'es of o^ranite on the south and southwest and the drumlins of Planter's Hill and World's End on the west and northwest. The natural boundary line between the geologic areas of Nantas- ket and northern Hingham appears, therefore, to be, not Weir River Bay but the eastern shore of Hingham Harbor. The southern boundary or limit of Nantasket is clearl}^ marked by the continuous depression formed by Strait's Pond, Lyford's Liking and Weir River Bay, which is continued, as a narrow strip of marsh, south of Rocky Neck to the bay which sharply indents the eastern shore of Hingham Harbor. This topographic trough defines with approximate accuracy the present border of the Boston Basin, separating the sedimentary and volcanic rocks froni tlie somewhat more elevated granitic area on the south. Although the line of depression just traced is a natural and important geologic boundary, it has appeared best to give this paper greater breadth and completeness by including in its scope not only the sedimentary and volcanic rocks of Nantasket, and the granite floor upon which they rest, but also a sufficient breadth of the granitic border of the basin to nmke apparent the marked contrast in geologic structure on opposite sides of 'Tlic oilier isliiiiils of liosfoii Iliiibor (tlie Browstcrs, (!:ilf U., Green Id., etc.) belong ) tlie town of Hull. But ilic geoloyy of tlie.se formed the subject of a previous paper l^roc. B. S. N. II., ,\xiii., 450-457); and will be more fully described in a later paper f this series. ' to (1 of this series. the boandaiy. The geology, so far as the hard I'ocks are con- cerned, is as simple south of the line as it is complex north of it. Hence, although, as the map shows, the addition of the whole of Cohasset and a portion of Scituate much more than doubles the area to be described, it extends but slightly the limits of this paper, which is still, notwithstanding its title and areal scope, chiefly a detailed account of the ledges of Nantasket. The numerous islands of Boston Hai'bor, as any one may ob- serve in going from Boston to Nantasket by water, are, with unimportant exceptions, composed, like the Nantasket Beach area, wholly or almost wholly of rounded drift hills and con- necting beaches of sand and gravel ; and the fine sections of some of the drumlins due to marine erosion show that in most cases they are pure drift accumulations, at least above sea level. The Boston Basin does not afford a stronger topographic contrast than is presented when the Nantasket steamer, having rounded Bumkin Island, the graceful drumlin outline of which is abi"uptly terminated on the west by marine erosion, and followed the winding channel past the beautiful drumlins of the World's End and Planter's Hill, sweeps by the bold and massive ledges of Rocky Neck and Nantasket. We pass in an instant from an area which is deeply covered by drift and presents only the most typical drift-contours, to an essentially driftless area, where the ice-sheet left hardly sufficient detritus to fill the narrow gorges and chasms dividing the rocky hills. The problem involved in this unequal distribution of the drift, in this abrupt passage from an area in which the ice-sheet acted, at least during its later stages, as an important agent of depo- sition, to an area in which its action was purely erosive, lies beyond the scope of this paper ; although, if the drift had not been swept from the Nantasket area, there would have been little occasion for a special study of its geologic features. We may now simply accept the almost perfect exposures of these rock masses as one of the best gifts of the ice-sheet to the student of our local geology ; for there is probably no equally limited area in the Boston Basin more worthy of detailed and thorough in- vesti2:ation than this southeast corner. The structure of other districts is, no doubt, equally varied and intricate ; but Nautasket excels in the very favorable op- ])Oi-tunities Avhich its almost continuous rock surfaces present for (leci[)hering its structure lines. It is especially interesting, in this connection, to obsei-ve that, while the rocky area occur- ring- next to the north of Nantasket, in the middle line or axis of the Boston Basin — the outer islands of Boston Harbor — affords the finest and most typical exposures of intrusive sheets or interbedded dikes to be found in this region ; the characters of the contemporaneous beds or ancient lava flows of the Boston Basin and their relations to tlie interstratified conglomerate can be studied to the best advantage at Nantasket, on the border of the basin. At no other point is the evidence equally clear and conclusive that the melaphyrs and porpliyrites of the Boston Basin are, in the main, true contemporaneous lavas which were poured out on the sea-floor at different periods during the deposition of the beds of conglomerate and sandstone. The Nantasket area also shows more <^learly than any other that in some parts, at least, of the Boston Basin the conglomerate beds, as well as the as- sociated melaph3U's rest directly upon the fundamental granite, witliout the possible interposition of the slate series. It may be pointed out in this connection that the main pur- pose of the present investigation is simply to elucidate tlie structure and explain the origin and sequence of the rocks of this somewhat complicated area. In other words, this is a structural rather tlian a litliological study ; and I wish to ac- knowlwlge my great obligation to Mr. Geo. P. Merrill of the U. S. National Museum for undertaking a microsco[)ic examina- tion of t!ie newer eruptive rocks of Nantasket. His determina- tions ha\e ])revented serious errors in my Avork and enabled me {() (haw certain conclusions with a degree of confidence which it woiihl otherwise have been impossible to feel. TOPOGRAPHY. The rocky areas or ledges of Cohasset and Nantaskct arc characterized, considering their diversified geologic structure, by a remarkable uniformity of elevation, indicating tliat in.prc- glacial times this region was worn down nearly if not quite to its base-level ; and tlie deeply incised and generally abru[)t character of the valleys now dividing this old peneplain shows that it must have shared in the genei'al and marked elevation of the land at the dawn of the great ice age ; while the flict that the deeper valleys are now to a large extent occupied by the sea or its deposits, i. e., are or have been true fiords, is sufficient evi- dence of a subsequent subsidence ; and, finally, the extensive beach and marsh formations prove that the present level of the land has been maintained for a very long time. The mod- ern base-level is strongly accentuated, not alone through tlie constructive action of the sea, but marine erosion has made ex- tensive inroads upon the drumlins and other drift deposits, especially of the Nantasket peninsula and adjacent islands, de- veloping the prominent sea-cliffs of Telegraph Hill, Point Allei-- ton. Strawberry and Green Hills, etc., as well as broad submarine platforms or shoals, the outlines of which can be traced on the Coast Survey chart. Since even the most typical peneplain must slope gently sea- ward, the average elevation of the rock-surface of eastern Mass- achusetts increases gradually as it i-ecedes from the coast ; and Nantasket and Cohasset are, therefore, one of the lower portions of it. In the valleys, the rock-contours are, of course, cai'ried down to and below sea-level ; but the ledges and rocky hills separating the depressions, the real remnants of the ancient peneplain, are, even in the Nantasket area and along the Co- hasset shore, rarely below 50, and usually from 75 to 100 feet, in height; and the elevation, increasing very slightly inland, attains its maximum of 125 to, possibly, 150 feet in the broad 6 area of rocky woodland along the bouudarv between Cohasset and Hinghani. The moderate disparity of these figures does not contradict the previous statement concerning the general uniformity of tlie rocky elevations ; for the main point is that while there are hundreds of rocky hills and ridges or ledges rising 10 to 50 feet or more above the ground between them, there are none rising to commanding heights, that is, there are no rock elevations decidedly overlooking the surrounding country, as do the drift hills or drumlins. It is of course this absence of true rock hills that proves the ancient peneplain ; for the present interstream surfaces, where not encumbered by drift, are essen- tially ledgy plateaus. The absence of crags and pinacles of rock is easily accounted for by the severe glaciation which this region has suffered ; but we cannot thus exj)lain the fact that there are no dominant rock hills of rounded or glaciated outline, like the Blue Hills. That glacial erosion was not equal to the com- plete obliteration of such reliefs is proved by the survival, even when quite isolated, of the numerous hills and ledges which do not rise above the surface of the dissected peneplain. Briefly stated, then, the rock hills of this district have all been carved out of the prcglacial peneplain ; and the general equality of relief thus determined must continue until an increased elevation of the land, giving renewed energy to the ordinary agents of erosion, permits a more general and imequal effacement of the original interstream surfaces. Tliat the depressions or valleys which now^ interrupt the pene- [)lain are not due chiefly to glacial, but to prcglacial aqueous, erosion is sufficiently obvious, in many cases, from tlieir direc- tions ; and it is also seen in the fact that the rock hills and ledges of Nantasket and Cohasset, through the influence of the faults, dikes and joints by which they are bounded, are, in the main, approximately rectangular in plan, with, frequently, very sfoci) •^'' even precipitous slopes, and showing only secondarily, mid not as their primary form, the roche niontonce outline due to glaciation. Hence, cliffs, sometimes rising abnn)tly from the water or the salt marshes, and straight, narrow defiles are characteristic topogra[)liic features of tlie district, although the partial submergence of the valleys and the general plateau form of the hills causes the elevations to aj)pcar incommensurate with the otherwise rugged character of the topography. North of Atlantic Hill we descend into the great central valley of the Bos- ton Basin, now occupied by Boston Ilarbor ; and the pene- plain is lost. This is an area of much softer, sedimentary rocks, which was deeply and broadly eroded, when the reo"ion was more elevated, by the united Charles and Neponset Rivers and their affluents. The lower portions, especially, of the old peneplain, in Co- hasset, are to a considerable extent emphasized by level expanses of modified drift. The principal sand plain, through which project both the drumlins and the higher ledges, rano-es from 40 to 60 feet in height and extends interruptedly over a large part of the town, mantling but not greatly masking the contours of the underlying rocks. Although distinctly recog- nizable at many points, it has its finest development in and about Cohasset Village, between Little Harbor and the railroad, rising very abruptly 45 feet from the level marshes of Little Harbor to the kettle-dimpled plain on which the village stands. North of Strait's Pond and Weir River Bay, in Nantasket and northern Hull, the modified drift is very scantily developed and of no topographic importance. But in the terraces or elevated shore-lines traceable at corresponding heights on some of the drumlins in Hull and also in Cohasset and Scituate, and accu- rately marking the varying levels of the sea at the time when the sand plains were formed, we have a related feature of con- siderable interest which will, in its proper sequence, be described in detail. While the modified drift is, at the best, only a minor factor in the topography of this district, the unmodified drift or till, on the other hand, occurring almost wholly in the form of drum- lins, adds greatly to the topographic relief and diversity. The drumlins constitute, virtually, the only elevations in Hull north of Atlantic Hill ; in fact it is doubtful if this part of the peninsula would have any existence as dry land if it were not for these solid nuclei of till about which the beach deposits have gathered. As already noted, and as the map shows, the drum- lins are not evenly distributed. In Hull, in the direct line of the beach, like a string of beads, are Point i\.llerton and Straw- berry Hills, White Head, Sagamore Head, Hampton Hill, Rockland Hill (a lenticular slope of till) and Green Hill. Parallel with this line, on the west, is a second, including the druuilins of Hull Village — Thornbush and Telegraph Hills, Little Hog Island and Bumkin Island, this line being continued to the west of Nantasket in the double drumlin of the Worlds' End, Planters' Hill and Pine Hill. From either of these lines we pass to an area, including the Nantasket ledges and all that part of Cohasset and Hingham north of the railroad and cast of Hingham Harbor, in which drumlins, or any noteworthy accumulations of till, are almost wholly wanting ; the hard granite surface of the peneplain standing forth clear and naked. But south of the railroad, again, in the broad depressed or valley portion of Cohasset, the drumlins are thickly planted, greatly obscuring this portion of the peneplain. The series begins on the northwest with Turkey Hill (181 feet) and Scituate Hill (177 feet), two of the most typical drumlins in the Boston Basin and the culminating points of the area to which this pa- per relates. Southeastward, as the map shows, the belt broad- ens, but the drumlins, although more numerous, are also smaller and very much flatter inform, being, where they reach the level of the principal sand plain, as in Hoop Pole and Mann Hills, in Scituate, not readily distinguished from it. But farther south in Scituate, we find again, in Booth Hill, a drum- lin of respectable height and great horizontal extent. West of King Street, Scituate Pond and the head waters of Bound Brook, and south of Turkey and Scituate Hills, no drumlins liave been observed in Cohasset, the peneplain being here, as iioith of the railroad, but slightly encumbered by drift dc- ])()sits. The drainage of this district is a simple story. No part of Tltill is more than one-fourth of a mile from the salt water, and 9 there iire no streams. The plateau of North Cohassct drains directly by several short brooks, of which the most important is Turkey Hill Run, rising in the swamp between Turkey and Scituate Hills, into the Strait's Pond trough, the Atlantic and Little Harbor. The remainder of Cohasset and northern Scituate are tributary either to Cohasset Harbor and its ex- tension, the Gulf, or to Bound Brook, which is the principal affluent of the Gulf. Considerable attention has been given to tracing out not only the actual drainage lines or streams, but also the areas of obstructed drainage, the ponds, swamps and marshes ; and it is believed that these features are represented more accurately on the maps accompanying this paper than on any earlier published maps. Besides Strait's Pond, which is due to an artificial barrier at its western end and is mainly salt water, the only important pond in this entire district of Hull, Cohasset and northern Scituate is Scituate Pond (sometimes called Lily Pond) in the southern central part of Cohasset. It is tributary to Bound Brook ; and it is also connected on the northeast with a swampy tract from which a much smaller brook flows directly into the Gulf. Scituate Pond is thus very nearly an example of a basin having two distinct outlets at the same level or in a state of equi- librium. It is probably due primarily to the accumulation of modified drift in the valley of Bound Brook ; and the general absence of ponds in this district may be attributed to the lack of continuity of the sand plains. Even the small kettle ponds of other districts, with a few exceptions, as in the vicinity of Little Harbor where the principal plain has its best devel- opment, are wanting here. Neither is the form of the ground favorable to the view that any important ponds have been obliter- ated by the cutting down of barriers or the growth of silt and bog deposits. If the valley of Bound Brook were cleared out, the Gulf would doubtless be extended inland to Scituate Pond, if not considerably beyond ; and this main drainage channel and its affluents are now so nearly on a base-level basis that they ?ire bordered throughout by broad swamps and meadows. 10 Besides Bound Brook and the Gulf, the principal submerged or drowned rock valleys of this district are Weir River Bay and Strait's Pond, Little Harbor, and Cohasset Plarbor. These fiords, as remarked, are all far advanced in the process of -silting up to the present level of the sea. Even Cohasset Harbor is a harbor in little more than name, presenting with the ebb of the tide a broad level expanse of mud and sand of great but unde- termined thickness ; while the widely extended marshes on the east and south show how greatly the area as well as the depth of the harbor has been reduced. The barrier beach bordering and limiting the salt marshes on the north has nearly separated the outer harbor from the Cove ; and the soundings on the chart indicate that a similar bar is now forming across the mouth of the outer harbor, from the Grlades westward. In more ex- posed situations, the sea has ah-eady closed by barrier beaches two eastern entrances to Cohasset Harbor, the main entrance to Little Harbor, the eastern end of Strait's Pond, and connected all the drumlins of the Nantasket peninsula by beaches fi-om fifty feet to nearly half a mile in width. The numerous rocks and islets fringing this part of the coast are not the result of marine erosion ; but these, and also the ledges, sometimes of great extent, now isolated by the salt marshes of Nantasket and Cohasset, testify equally with the submerged valleys to the subsidence of the land. Continued subsidence would sub- merge these ledges and isolate others, leaving the general aspect of the coast unchanged. MAPS. This paper is accompanied by two maps (Plates I and II) . The first, on a scale of 2400 feet to the inch, embraces the entire peninsula of Hull, the whole of Cohasset, a portion of Scituate and all but the southern end of Hingham, or, in other words, the entire area to be described not only in this paj)er but also in the following paper on the geology of Hingham. To])ogra]>hi- cally, it is based primarily upon the Coast Survey chart of n Boston Harbor and the atlases of Plymouth and Norfolk Coun- ties ; l)ut certain featui-es, especially the marshes, drainage lines, (Irumlins and other drift reliefs have been greatly modified in ac- cordance Avith original observations. The aim has been to com- pile the best obtainable data, rather than to preserve a uniform degree of accuracy, and undoubtedly some parts of the map arc much more reliable than others. Even in the absence of con- tour lines, it exhibits some of the more important relief features, the kames and drumlins being clearly distinguished ; w^hile the hydography, including the shore-lines, ponds, streams, swamps and mai'shes, is represented with unusual completeness and ac- curacy ; and it shows in addition the general structure of the hard rocks, and especially the relations of the granitic masses to the sedimentary and volcanic deposits. The Nantasket area, the elucidation of which is the special object in view, forms the central part of this first or general map, which thus illustrates its topo- graphic and geologic relations to contiguous districts ; while the second or special map, on a scale of BOO feet to the inch, permits of much greater detail. No good or even approximately accurate topographic map of all that large part of the Nantasket district west of the County Road has heretofore been published, and it has been necessary to devote considerable time and labor to the preparation of a suita- ble basis for the representation of the geologic features. I am particularly indebted to Mr. Wm. M. Beaman for the sub- stantial accuracy of the western area, between the railroad and Weir River Bav, and for a general triano'ulation of the entire district from Planter's Hill to Green Hill and Black Rock ; and to Mr. Gordon H. Taylor for the contour lines upon nearly the entire map excepting the western ar.ea.^ The map of Rocky Neck is based upon the plan of a land survey kindly furnished by Mr. John R. Brewer ; and the coastal area, from the steamboat wharf to Green Hill is re- ^ Messrs. Beaman and Taylor have made a careful study of the topography of this section as students in the Department of Civil Engineering in the Massachusetts Institute of Technology. 12 tliiccd, with some modifications, from tlie ma[) ot" Nantasket in the atlas of Plymouth County, 1879 ; while I am personally re- sponsible for the outlines of the central area, between the County Road and railroad, and for the entire south shore of Weir River Bay, Lyford's Liking and Strait's Pond. The general absence of roads or other artificial land-marks west of Hull Street and the County Road, and especially west of the railroad, and the fact that the hills and other natural features in this section were still unnamed, caused it to appear desirable, in the interests of concise and intelligible description, to assign names to the more important rock masses, marshes, etc. Of course it is not expected that many if any of the geo- logic and descriptive designations which aj^tpear upon the map and in the following pages will gain general acceptance : and this is in no wise essential to their present usefulness. The actual exposures of the hard rocks are so nearly continuous over the areas where they are indicated by colors on this map that it has appeared unnecessary to represent the individual ledges or outcrops ; although this is virtually done for all the ledges in the marshes and below the high-tide line. And, ex- cept in the case of some of the fault-lines, the map may be fairly regarded as a plain i-ecord of actually observed facts, de- void of theory. THE GRANITIC ROCKS OF COHASSET. This heading, in its broadest application, covers all the geo- logical formations of Cohasset, except the dikes of diabase and porphyrite and the drift deposits. The dikes, being essentially similar in character and age to those of Nantasket will be most conveniently described in that connection ; and since the super- ficial geology forms naturally one continuous chapter for the en- tire area to which this paper relates, the drumlins, sand plains, etc., of Cohasset will not be taken up separately. But it appears best to introduce a general account, lithological and structural, of the granitic rocks of Cohasset at this point, because they form, 18 not only tlie border, but the floor of tliis part of" the Boston Basin. They are alike the foundation upon which the newer sediments and lavas of Nantasket are piled, and the principal source from which the material for building the Nantasket strata was de- rived. Among the granitic rocks of Cohasset are included chiefly the diorite and the granite proper. The diorite is in eveiy in- stance clearly the older, as well as the less abundant and less important, rock ; the relations of the two rocks being essentially the same here as elsewhere about the Boston Basin. To a large extent they are quite intimately associated, the diorite occurring very generally in the form of irregular fragments or masses, of all sizes, enclosed in the granite ; while in other cases the gran- ite forms irregular, branching dikes in the diorite. In fact, the granite, however massive it may be, is rarely entirely free from inclusions of diorite ; and the diorite, even when farthest from a main body of granite , almost invariably exhibits a net-work of granite intrusions. In other words, the diorite has been very generally fissured, and in large part completely shattered, and then injected by the granite. Hence, although the two rocks are always perfectly distinct in their chronological relations, ob- serving an invariable sequence, it is, for considerable areas, a hopeless task to trace or define the distribution of either sepa- rately from the other ; and it is for this reason alone that they are not distinguished on the map. Any boundaries that might be drawn upon the map would have but very little significance, since in no case could they be either exclusive or inclusive. The granite very largely predominates ; and the diorite, except as isolated inclusions in the granite, is quite restricted in its distri- bution. It has been observed chiefly in the ledges south of Co- hasset Harbor, north and west of Little Harbor and along the shore between Little Harbor and Nantasket. The large ledge south of the Cove and west of the Gulf — Kent Rocks — is mainly granite, but encloses considerable diorite, while the pointed, rocky hill immediately east of the Gulf — Government 14 Rocks — is all granite, coarse and pinkish. In the large ledgy tract nearly half a square mile in extent between the Gulf and the Scituate shore, diorite appears to be the prevailing rock ; but the granite is always close at hand and forms some large masses. The ledges along the Scituate shore, east of Cohasset Harbor and the Glades, and advancing from the south, are, except for an occasional inclusion of diorite, wholly composed of a beautiful, coarsely but uniformly crystalline and massive pinkish granite, which also forms the adjacent islands, from Barr's Kocks to the Osher Rocks. On reaching the prominent point a short distance north of the Osher Rocks, however, we pass abruptly from the granite to diorite with only occasional irreofular dikes of ffranite breakingf throuoh it. The diorite forms the north shore of this point ; but, crossing a short shingle beach, wc find that the east shore of Strawberry Point is chiefly coarse granite ; while across the north side there is much diorite alternating with the granite. It is partly in solid, unbroken masses, more commonly veined with granite, and to a consid- erable extent completely morcellated, yielding a very coarse breccia in which the diorite forms the fragments and the granite the cement. Gull Island presents at low tide a broad flat sur- face of diorite irregularly veined with granite ; but Shc]i])ard's Ledofc and the ledo-es southwest of Gull Island are ofi'anite with little or no diorite. These characteristic relations of the two rocks are frequently repeated on the many other ledges and islets between Cohasset Harbor and Minot's Light, although the granite usually predominates. Along the north side of Co- hasset Harbor, between the railroad and White Head, the ledges, so far as observed, are nearly all granite ; and following the shore northward from AYhite Head, around Sandy Conc, to the mouth of Little Harbor, the rock is all the most typical, coarsely crystalline, light gray graiititc, weathering pink or reddish, very massive in structure and with only very rarely a small inclusion of diorite. North of the narrow mouth of Little Harbor, on I>each Island, the granite continues, unchanged, to a point be- wecn 200 and 300 feet south of the artificial hailjor on the 15 headland opposite Brush Island. Here we pass very ahru[)tly from the coarse, massive granite to the mixed granite and diorite, tor the first hundred feet or so a very confused mixture, in which the diorite seems to predominate ; but west of that, along the entire stretch of shore, as far as Green Hill and Nantasket, the granite is the prevailing rock, with frequent inclusions of diorite. This is, structurally, an extremely interesting section, and it will be described more fully a little farther on. The essentially patchy distribution of the diorite in the granite makes the tra- cing' of this rock in the weathered and lichen-covered inland ledges rather unsatisfactory, not to say unprofitable, and only enough work has been done in this direction to show that what we can see so clearly along the shore is really characteristic of the whole town. Thus it must be evident to any one observing the ledges along the west side of Little Harbor and on Forest Avenue that the mixed granite and diorite extends inland a con- siderable distance : while in other parts of the town it is equally clear that coarse and massive OTanite covers laro'e areas. The diorite is always dark-colored and holocrystalline, but usually rather fine-grained, varying in texture from compact or aphanitic to distinctly but not coarsely crystalline, i. e., the dio- rite is rarely coarse in the sense that the granite often is. On the other hand, it rarely resembles diabase, except in the most com- pact forms, the normal diflPerence in crystalline structure or habit being readily recognized in the macrocrystalline diorite. Under the miscrocope it is usually seen to be composed chiefly of plagioclase and hornblende. The feldspar is commonly rather opaque ; but the hornblende is, in many cases, beauti- fully clear and dichroic, although usually bordered by secondary biothe, chlorite, etc. ; while in the more conjpact and highly altered diorite the hornblendic element is very largely redu.ced to hydrous silicates, which give the rock a dark greenish color. Among the secondary minerals occurring in veinlets and irregu- lar segregations epidote is most prominent, but it is frequently accompanied by chlorite and quartz. The black oxides of n'on (magnetite and menaccanite) are usually present but rarely 16 abundant, .and the rock also often contains a small amount of original quartz. The quartz diorite is especially interesting as indicating a gradual passage to the more basic granites. The diorite is undoubtedly a plutonic rock ; and it is sufficient- ly varied in character to suggest that it is possibly not all of the same age ; but no facts have been observed which point to a definite conclusion, and all that can be regarded as well deter- mined is that its relations to the granite are always essentially the same. The granite, on the other hand, belongs very clearly to two, and probably three, more or less distinct periods of igneous activity, or successive phases of the same period. First in order of time comes the granite which is most intimately associated with the diorite. This is very abundant, and partakes of the character of the diorite. It is on the average only a little more coarsely crystalline than the diorite ; and usually contains suffi- cient hornblende or black mica to make it quite dark colored , for a granite. The quartz is often deficient, and the feld- spar is partly plagioclase, so that it would be easy to mistake a portion' of the rock for either diorite or syenite. The horn- blende shows, perhaps, even more alteration than in the diorite, being very largely replaced by chlorite. Next in order comes the light gray and pinkish granite, which is usually coarsely crystalline and massive and is as a rule comparatively free from inclusions of diorite. This rock is rich in acid feldspar and quartz ; but the hornblendic element is usually quite scantily de- veloped ; and black mica (biotite) often partially, sometimes wholly, replaces the hornblende, wliilc both of these accessories are, as a rule, largely altered to chlorite or other hydrous species. It can be seen breaking through the first granite and the diorite at many points along the shore, especially between Little Harbor and Nantasket ; and some of the smaller dikes of this granite arc quite fine grained and not easily distinguished from the third type. Tills coarse acid granite is the only rock in tliis district of any particular economic interest ; and it has been quarried only to a very limited extent. There is one small quarry on the 17 Scituate shore, near the Osher Rocks, and anotlicr on tlu; Co- hasset shore, north of Sandy Cove; but inland notliino- dosci'v- ing the name of a quarry has been observed. The tliird n^ranito, in chronological sequence, forms small and irregular dikes from a fraction of an inch to several feet in width cutting all the older rocks, but occurring chiefly in the coarse, typical granite (No. 2), from which it appears to differ in composition only in containing less hornblende or mica. These dikes are always fine-grained, varying from a finely crystalline gray or pinkish oranite or micro-granite to a true felsite. The micro-granite passes into an acid felsite of the same chem- ical composition. But while the micro-granite is common, form- ing hundreds of small dikes, true felsite, which does not reveal a holocrystalline ground -mass under the microscope, is rather rare. The largest mass which has been observed is at the northeast corner of the large ledge of granite on the south shore of Lyford's Liking, south of Round Hill. This is a brown, thoroughly compact or felsitic, structureless rock; and forms a mass several yards across, just at the water's edge. The contacts with the granite are not clearly exposed ; but it is evidently a dike in the granite. Small dikes of a true quartz- porphyry have also been observed on the Jerusalem Road, eai^t of Green Hill. The great abundance of pebbles of felsite in the Nantasket conglomerates indicates that this rock was for- merly much more extensively developed in this district. It probably occurred chiefly in the form of broad surface flo\As, such as still exist in other parts of the Boston Basin ; and the dikes of both felsite and micro-granite probably date from these volcanic eruptions, being branches from the main fissures or necks throuo-h which the felsite reached the surface. Although these three types of granite clearly reveal the chronoloo-ical succession described above, it can not be shown that they are widely separated in time or belong to entirely distinct periods of igneous activity ; and it is especially obvious that the micro-granite and felsite, although cutting the coarse OCCAS. PAPERS B. S. N. H. IV. 2 18 granite, should, on account of their resemblance to it in com- position, be regarded as, in a general view, essentially syn- chronous. This coast abounds in instructive exposures of the granitic rocks, but the nearly continuous belt of wave-washed ledges along the Jerusalem Road between Nantasket and Beach Island is particularly worthy of thorough study, illustrating, as it does, the relations of all the rocks and presenting some features which are not clearly exposed elsewhere. On the south side of Beach Island, as already noted, there is a fine development of the coarsely crystalline, massive granite ; but from the junction of this with the mixed granite and diorite just south of the artificial harbor to Kantasketthe finer-grained, dark-colored, older granite is the principal, and for a considerable portion of the distance, excepting the dikes, almost the only, rock. This older granite is characterized throughout the entire section by a remarkable gneissoid structure due to the flowing of the material while in a viscous condition. At many points this flow-structure is so perfectly developed that the rock presents a distinctly gneissic character even in small hand-specimens ; and where the fluidal lines are least obvious, they are still traceable on the broad, clean surfaces of the ledges. The trend of the flow-structure or pseudo-stratification varies somewhat. It is about N.E.— S.W. on Beach Island ; while farther west it ranges usually between N.-S. and N.W.-S.E., but becomes N. 60° W. as we approach the l:)cach leading to Green Hill. The dip is usually nearly vertical, although sometimes as low as 60° or less. This gneissoid granite encloses at most points numerous frag- ments of diorite of varying lithological character. These exhibit a great range in size ; and they are usually more or less elongated in form, being often distinctly lenticular and indicating in their smooth and somewliat indefinite outlines a partial fusion of the diorite in the melted granite. The elongated fragments coincide in direction with the flow-structure of the granite, and this struct- ure is always most perfect where the diorite is most abundant. The Ibrm of the diorite fragments means, probably, that the 19 diorite also possessed a gneissic structure of some sort I^efore tlic eruption of the granite ; in fact this can frequently be observed in the more lenticular masses of diorite. The diorite, then, has been injected by the granite chiefly along its own original struct- ure-planes ; and the flow-structure of the granite is probably due largely to its having come up between these parallel walls of diorite. Breaking through this gneissoid granite and the enclosed dio- rite at irregular intervals are many small and some large masses of the coarse, light-colored granite. These usually conform in direction, at least approximately, with the flow-structure of the older granite, and exhibit in this direction a similar but, as a rule, less distinct flow-structure. We thus not only have a somewhat gneissoid diorite of probable igneous origin broken through very profusely by, and enclosed in, a highly perfect gneissoid granite — a true eruptive rock which, but for its rela- tions to the diorite, might readily be classed as of sedimentary origin, a true gneiss ; but both of these terranes are traversed again and again in a way possible only to an igneous rock, by a second and more acid pseudo-gneiss. Again, all the preceding- rocks are injected by occasional small and irregular dikes of micro- granite and felsite, which, for the most part, are devoid of gneissic structure. And, finally, this entire sequence of granitic eruptions is divided by a well-defined series of porphyrite dikes and no fewer than three distinct systems of diabase dikes ; all of which will be described in connection with the dikes of Nan- tasket. This shore undoubtedly presents the best general section of the plutonic rocks of the Boston Basin. Certainly at no other point do we find plutonic masses of so many different ages so clearly exposed in their normal relations ; and in this connection it should be noted that the passage (in natural sequence) from the diorite through the more basic to the more acid granite affords some indication that the diorite is not widely separated in time from the granites, the entire series constituting but one complete igneous cycle. THE ROCKS OF NANTASKET.. GENERAL RELATIOXS AXD ORIGIX. Although the surface exposures of the rocks of the Nantasket area are completely isolated by drift deposits and the sea, these strata are probably continuous to the north and west with the great body of sediments occupying the Boston Basin. And it is certain, as will appear later, that they are terminated on the south by profound dislocations ; so that the sharply defined boundary between the Nantasket sediments and the broad area of granite can not be regarded as marking the true original border or maximum extension of the Boston Basin in this direc- tion. On the contrary, the facts point very plainly to the con- clusion that the basin rocks formerly extended a considerable but undetermined distance beyond this line ; being here, how- ever, on the upthrow side of the great faults, they were lifted above the present plane of erosion. We may, nevertheless, en- tertain the hope that future investigation will reveal upon the granitic plateau outlying remnants of the sedimentary series, and thus indicate more exactly the original limits of the basin. The Nantasket rocks, above the fundamental granite, and omitting the dikes, consist chiefly of the conglomerate (Roxbury pudding stone) and the interbedded lavas and tuifs. The incompleteness of the Nantasket section is plainly shown in the entire absence of the great slate series, which elsewhere in the Boston Basin overlies the conglomerate ; and it is probable, as will appear when the facts are presented, that the upper members of the conglomerate series are also wanting. The student will, however, find compensation for these deficiencies in the magnifi- cent development of the basal beds of the conglomerate. At no other point are the floor of the Boston Basin and the strata rest- 21 .<»e?f^/o-amyg- dules, like the true amygdules, are filled with a variety of secondary minerals, as feldspar, quartz, epidote and chlorite, often without definite order, or again showing a quite regular concentric arrangement. I have not, however, been able to find the marked regularity in the arrangement of these minerals in the pseudo-amygdules that I was led to expect from a peru- 1 Proc. B. S. N. IL, Vol. xx., page 41C. 33 sal of Mr. Benton's paper. The chlorite of the amygdules is quite different from the so-called viridite of the diabases. With a low power and by ordinary light it appears like a- continuous sheet of a light green micaceous mineral viewed perpendicularly to its cleavage ; but with a power of 400 or 500 diameters, it is seen to be a compact mass of minute scales which is almost black between crossed nicols, but which polarizes faintly as the stage is revolved and shows at the same time that the individual foliae have a fan-shaped arrangement. It very likely represents the final stage of the viriditic alteration, and is presumably the same as the dark green compact chlorite which occurs in masses of some size in the amygdules of the melaphyr at Brighton. Numerous other minerals of secondary origin are found in the cavities and crevices of the melaphyr ; but as they occur only in minute quantities in the material at hand and usually without well defined crystallographic outlines, their correct de- termination is a matter of great difficulty, and in some cases would be impossible. For this reason and because they do not appear to have any important bearing upon the problems under consideration I have omitted mention of them in my descriptions. Fourth Melaphyr, Coastal Area. — This is a fine grained greenish rock, with few macroscopic constituents. True amyg- dules are rarely abundant in the melaphyr of this ai^ea, but it is very commonly distinctly brecciated, considerable masses having a highly fragmental aspect. Under the microscope it is found to consist of the usual lath-shaped plagioclases with step-like ends imbedded in a grayish amorphous ground-mass and showing^ a marked fluidal arrano-ement. Much secondary epidote is present, occurring in scattered grains, in strings or veins and in aggregates filling the numerous small cavities re- sulting from the decomposition of some of the original constitu- ents of the rocks. There is also present a considerable amount of viridite, which results from the alteration of the augites, and is also derived in part from the ground-mass. This at times occupies the entire interspaces of the feldspars, and, act- OCCAS. PAPERS B. S. N. H. IV. 3 34 ing but faintly on polarized light, gives them the appearance of beino- imbedded in a light greenish glass . In the limited bed on the north side of Atlantic and Centre Hills, bomb-like masses of melaphyr are highly amygdaloidal, the amygdules varying from 1 to 5 mm. in diameter. These I regard as true amygdules. They are filled mainly with quartz and chlorite, though often a little epidote is present. In such cases the quartz occupies the outer zone with the chlorite and epidote interiorly. The ground-mass of this rock is an extremely dense, dark brownish gray, amorphous material bearing minute feldspar microlites. The pseudo-amygdules of this melaphyr are commonly much elongated, and contain epidote and chlorite with more or less quartz. The epidote occurs as minute crystals, a single row of which often forms a border around the wall of the cavity ; while the entire central portion is occupied by the very compact, light green chlorite noted above. Outside of the amygdules or seg- regations of both kinds, scattered through the entire mass of the rock, are the common decomposition products, ferrite, opacite, viridite and kaolin. A typical sample of melaphyr from this area, but bearing some secondary quartz, which could not be avoided, yielded 50.47 per cent, of silica. Second and Third Melaphyrs, Central Area. — Themel- aphyrs of this area are usually of a deep greenish color and often highly amygdaloidal ; but they vary greatly in botli color and texture. A common type is a dark purplish gray rock, which is also often profusely amygdaloidal. Under the microscope its color is seen to be due to the large amount of opacite and magnetite in the ground-mass. Secondary epidote is abundant in comparatively large grains, which are often surrounded by a dense black border of opacite. The amygdules are filled with quartz, epidote and feldspar. The quartz and epidote prevail and occur in the form of an extremely fine grained and compact aggregate having a specific gravity of about 2.!). The iron 35 oxides, which are much decomposed, are in part titan iferous, as shown by theii* peculiar club-shaped forms so characteristic of" menaccanite. Moreover, the pulverized rock gives a distinct re- action for titanium, after prolonged digestion with hydrochloric acid. The green melaphyr differs mainly in carrying a larger proportion of epidote and in the feldspars having undergone the viriditic alteration. The amygdules of the rock show first a zone of epidote, and interiorly quartz, feldspar and chlorite. A typical sample of the melaphyr from this area, non-amygda- loidal and carrying no free quartz, yielded on analysis 41.89 per cent, of silica. First Melaphyr, Western Area. — The structural features of this rock are greatly obscured by decomposition, but after the examination of several sections I am disposed to place it with the melaphyrs. The microscope shows it to consist of a large number of greatly decomposed porphyritic plagioclases imbed- ded in a groundmass so filled with secondary decomposition products that its original character is wholly obscured. The dull green macroscopic areas are epidote resulting from the feldspathic decomposition. The original iron-magnesian constituents have quite disappeared, even their outlines being no longer recogniz- able. I assign the rock to the melaphyrs with consider- able hesitation. Chemical analysis shows its basic character, three samples from different localities affording the following percentages of silica: 47.29, 47.97 and 51.05: the latter, at least, probably including some free secondary quartz. Porphyrite. The rock which I here have called porphyrite differs from the melaphyrs in showing under the microscope a larger pro- portion of irresolvable ground-mass, in being of a more pro- nounced porphyritic structure and in never so far as observed being truly amygdaloidal. They are also more compact and fresh appearing and show in none of the sections examined 36 traces of augite, olivine or other original iron-magnesian sili- cates. Western Area. — The porph^^rite of this area varies from light greenish to dull red and purplish in color and is well de- scribed by the term felsitic. So far as observed none of its mineral constituents are of such dimensions as to be recog- nizable by the unaided eye, nor is it at all amygdaloidal, a feature so pronounced in the melaphyrs. The purple variety, in the thin section, shows a densely microlitic base, with much opacite, carrying numerous porphyritic feldspars, which are mostly, if not all, triclinic. A few of the feldspars show no twinning stride ; but I cannot obtain such measnrements as would prove these to be certainly monoclinic. One of the sections examined shows a single corroded bleb of original quartz, carrying flu id al cavities and moving bubbles. Large dnsky apatites are not rare. In the greenish variety there is also present a little viridite from the feldspars, and a few granules of secondary epidote. The porphyrite presents, usually, a well-marked fluidal structure. Subjected to chemical analysis, the green variety yielded 58.02 per cent, of silica, and had a specific gravity of 2.73. The red variety is more felsitic and also more decomposed. Two samples, light red, and dark red in color, yielded silica as follows : light, 56. (j() per cent. ; dark, 5G.25 per cent. Black Rock. — Some of the black rock porphyrite more re- sembles the melaphyrs in external appearance than does that of the western area. Macroscopically it is a very compact and hard, greenish gray to dark purplish and nearly black rock, bearing small whitish feldspars, and secondar}' epidote in crys- tals and granular aggregates of sufficient size to be distinguish- able by the naked eye. Thin sections, under the microscope, show it to consist of a dense ground-mass of needle-like and short, stout feldspar micro- lites, interspersed with numerous brilliantly polarizing epidote 37 granules, and witli little if any truly aniorplious base. Im- bedded in this feldspathic ground-inass are numerous large, clear and often greatly corroded plagioclases. A few small apatites are present. The section shows a well-marked flow-structure. Chemical analysis yielded 58.25 per cent, of silica ; and an aver- age of four determinations gave a specific gravity of 2.78. The Dikes or Intrusive Rocks. The dikes, as has been noted by Professor Crosby, are all of normal form — sharply defined, wall-like masses making high angles with the horizon and invariably cutting across the strat- ified rocks ; no intrusive beds or sheets having been observed in the Nantasket district, owing, probably, to the absence of thin bedded or slaty sediments. As previously stated, the dikes consist mainly of diabase ; but they also include an important series of porphyrite dikes and at least one large dike of mela- . phyr. The dikes of melaphyr and porphyrite are believed by Pro- fessor Crosby to be contemporaneous with the surface flows of those rocks ; while he finds that the numerous diabase dikes are certainly newer in most, and probably in all, cases than the melaphyr and porphyrite. Hence, following the chronological order, we begin with the last named rocks. The Melaj^hyr Dikes. The dike of melaphyr ( 1 ) on the north side of Centre Hill is so evidently identical with the melaphyr which it intersects that it was not examined microscopically. The, large dike (2) in the western area, which Professor Crosby regards as contem- poraneous with the first flow of melaphyr, is, like that flow, of doubtful character. Macroscopically, it bears a marked resem- blance to this flow, and is contrasted in appearance with all the diabase dikes of the district. Under the microscope it is seen to be a fresher and more crystalline rock than the melaphyr, showing in the section a ground-mass of feldspar microlites and 38 granules injected with epidote and the usual ferruginous decom- position products. In this ground-mass are numerous phen- ocrysts of feldspar and an occasional nearly colorless augite. There are numerous granular areas of secondary epidote ; but no trace of olivine or its decomposition products ; and 1 am inclined to doubt its presence as an original constituent. Rec- ognizing the uncertainty attending the determination of such rocks from a single section, I can only say that I am inclined to believe the rock more nearly related to the augite-porphyrites or olivine-free diabases than to the melaphyrs. It is classed here with the melaphyrs simply because the true nature of the rock is not yet free from doubt, and that arrangement best accords with its field relations as worked out by Professor Crosby. The analysis of a single specimen gave 54.47 per cent, of silica. The Porphyi'ite Dikes. The porphyrite dikes, so far as observed, are, lithologically, essentially similar to the great flow of porphyrite on Black Rock, rendering a detailed description unnecessary. The Diabase Dikes. These dikes are, as a rule, exceedingly compact, tough and hard, of a dark gray, nearly black or more frequently greenish color, and carry few if any original macroscopic constituents, although pyrite and secondary epidote are often developed in granules and aggregates of sufficient size to be detected by the unaided eye. To this secondary epidote and the abundant viri- dite is due in all cases the green color of the rocks. I find nothing whatever in the slides, with possibly a single exception, to indicate that the different dikes belong to distinct periods of eruption. Slides from dikes which plainly belong to the same system often show all the variations in texture, struct- ure and stages of decomposition to be ol)served in those which 39 from their position belong iinmistrtkahly to distinct systems. Tn grouping the rocks as below I have, therefore, but followed the notes of Professor Crosby. .n ■[ First Series. Dike 3, — Rock finely porphyritic with greenish feldspars ; macroscopic pyrite ; color greenish. This is a very typical diabase, with irregular and sharply wedge-shaped augites, lath- shaped plagiocTases and numerous grains of iron oxide. The augites are undergoing a chloritic alteration ; and the feldspars are so badly kaolinized that the twinning strife are completely obscured. The iron oxides are in part magnetite and in part show the whitish alteration characteristic of menaccanite. Pyrite occurs in brassy yellow irregular clumps. Small, irregular, brown and strongly dichroic scales attached to the altered au- gites are evidently secondary hornblende. The porphyritic feldspars are so thoroughly kaolinized that nothing whatever can be learned of their original nature. Scattering grains of epi- dote, kaolin, opacite and viridite complete the list of recognizable constituents. Dike 5 — Very fine grained and compact ; macroscopic pyrite ; color greenish. This presents no distinctions in any way essential. Both augites and feldspars are more decom- posed ; and there is a corresponding increase in the propor- tions of epidote, viridite, calcite, etc. A few small apatites appear in this slide. Dike 22 — Medium fine in texture but distinctly crystalline, and color greenish. This differs from the preceding only in being slightly coarser in texture. The iron oxides, judged by their irregular forms and whitish decomposition products are largely titaniferous. Dike 25. — Very fine-grained and compact ; a few greenish feldspars porphyritically developed ; and color greenish. The 40 aiigite is in most cases completely altered ; and there is much viridite, epidote and iron oxide. This rock is undoubtedly a diabase, although I am unable to find a particle of recognizable augite or even hornblende, the viriditic alteration being complete in all the sections examined. The viridite is very abundant and is considered of undoubted augitic derivation, from its sharp straight and very angular outlines, which are so characteristic of the augites in this class of rocks. Minute epidotes are in some cases very abundant and impart to the rock a yellowish green stain. The section is traversed by veins of secondary quartz. Dike 37, — Very fine grained and compact ; macroscopic pyrite ; color, greenish. This is of finer grain than any of the preceding and so highly altered that none of its original constituents are now recognizable, excepting by pseudomorphs. The section shows only a fine, compact, fibrous or scaly ag- gregate of viridite, opacite, epidote, calcite and kaolin, with numerous grains of iron oxide. Although so highly altered, I have no hesitancy in referring this rock to the diabase group. Second Series. Dike 12. — Very fine grained and compact ; no macroscopic constituents ; color, dark greenish gray. Under the micro- scope it is seen to be highly altered, but with portions of augite still fresh and showing pleochroism. Iron oxide very greatly altered to a grayish amorphous product. Other alteration pro- ducts as in preceding sections. Dike 11. — Differs from 12 only in being of slightly coarser texture and in having suffered more from alteration ; augites completely changed. Dike 10. — Fine grained and compact ; porphyritic with small greenish feldspars. Plainly a diabase, although the feldspars 41 are so miuldiod by decomposition products as to he scarcely recognizable. A few augite [)articles arc still unchanged. No essential differences. Dike 9. — Like 10, but a triffe coarser in texture. The hand specimen shows small black segregations ot" what is a])parently hornblende ; but these do not appear in the section. Dike 8. — Very compact, with no macroscopic constituents ; color, dark gray, nearly black. The sample was taken from the edge of the dike at the contact with melaphyr. It offers no characters worthy of note to distinguish it from 9. Dike 7. — Coarse and distinctly crystalline ; color, greenish. This is one of the coarsest rocks of the series. Under the micro- scope it shows large plates of reddish brown augites, often en- closing the lath-shaped plagioclases and producing the typical ophitic structure. The feldspars here are sufficiently fresh to show twinning stria?. Iron oxide is very abundant, in large grains, with much pyrite and some apatite. The section shows to excellent advantage the various stages of augitic alteration into viridite. The rock resembles 22 more closely than any of the preceding. Dike 26. — Very fine, compact and homogeneous. Color, dark gray, somewhat greenish. This rock is greatly altered ; being essentially an aggregate of viridite scales, and fibres and gran- ules of iron oxide, interspersed with epidote. A few cavities or pseudo-amygdules occur, the cavity wall being lined with vir- idite, while the central portion is occupied by calcite. The feld- spars are very obscure ; and the iron oxide in part titaniferous. Dike 29. — Texture of medium fineness, with large segregations of black hornblende. Color, uneven, black, greenish and yellow- ish green ; the last tint being due to epidote veins. The augite is greatly altered ; and there is much viridite and epidote, the latter in scattered granules and veins. The feldspars are ob- 42 scure ; and quite large grains of iron oxide and pyrite occur. The segregation patches already alluded to are found, under the microscope, to consist largely of deep brown, strongly dichroic hornblende, in compact and well-defined crystals ; sometimes perfectly fresh or again somewhat altered into the inevitable viridite. I am not able from the sections prepared of this rock to state whether this hornblende is original or a product of para- morphic alteration from the augite. Dike 31. — This is of medium fineness, with macroscopic pyrite and segregations of black hornblende. The color is greenish, the rock being blotched with epidote. The augite is almost completely altered to viridite, although a few remnants are still recognizable. This rock also contains, as stated, coarser grained segregations in which black hornblendes are readily distinguished by the unaided eye. In thin sections these are of a deep brown color and strongly dichroic. Some of the individuals are so compact and well defined in crystalline outline as to indicate that they are original constituents of the rock ; while in other cases the presence of a hornblende border of varying width about an augitic core indicates unmistakably a paramorphic origin. The hornblende has, in its turn, undergone in certain cases the viriditic alteration. This is the only one of the Nantasket rocks in which I have been able to find paramorphic hornblende so distinctly characterized as to leave no doubt of its origin, al- though chis is suggested by sections from dike 29. A few small, grayish, wedge-shaped crystals of sphene are present. Feldspars, in most part, are considerably altered ; but in a few instances are still clear and show the twinning striae very plainly. There is much viridite, which obscures everything. Dike 34. — Differs in no essential particulars from 29 and 31, though the section does not show the hornblendic segreo^ations. Dikes 40 and 41 — These two dikes are without doubt iden- tical with the preceding, though nothing in the present compo- 43 eition goes to show tliat tlioy were not originally diorites rather than diabases. So far as examined, not a particle of" fresh au- gite remains ; but an occasional fragment of brownish hornblende still exists so far unchanged as to be recognizable. The lijjht green viridite is everywhere present, interspersed with magnetite grains, particles of epidote and apatite needles. A few of the feldspars are still fresh, though here also the viriditic alteration has prevailed. My reasons for calling the rocks diabase is that the products of decomposition, their form and their arrangement, are identical with those of the augitic rocks throughout the en- tire area under examination . Third Series. Dike 62. — Very fine grained and compact ; nearly hlack. Under the microscope this rock shows a marked deviation from the prevailing types of structure as already described. In the thin section it shows beautifully fresh and clear porphyritic plagioclases, in single lath-shaped forms and in cruciform aggre- gates : tosrether Avith altered olivines and occasional clear but corroded augites, imbedded in a dense fine-grained ground-mass of small plagioclases, augites and iron oxide. The olivine, though changed completely to a chloritic product, is readily recognized by its form and irregular fracture-lines. The augites are con- fined almost wholly to the ground-mass ; but the section shows a few corroded porphyritic forms, as above noted. All are fresh and free from enclosures. The iron oxide occurs abun- dantly in rod-like forms, crossing one another at nearly right angles, producing grate-like or barred structures. This dia- base is the least altered of all that has been examined. Dike 66. — Of medium coarseness, and to the naked eye ap- parently granular crystalline and fresh. This rock is quite dif- ferent in general appearance from any yet described. Macroscopically it is a well compacted, quite fresh-appearing rock, of a deep purplish black color, in which hornblende or an- 44 gite is apparently the chief constituent. Under the microscope, however, the section is found to present, aside from the apatite needles, not a sino-le orisrinal mineral in a fresh and unaltered state ; but the entire mass of the rock is filled with small amor- phous grains and dust-like opaque particles of a black color, in- terspersed throughout a dirty grayish groundmass, with only here and there a small fragment of an augite crystal or colorless portion of feldspar, in which in a few instances twinning stria? were still apparent. Scattering grains of epidote and shreds of a brownish mineral, evidently hornblende, complete the list of determinable minerals. The black opacite and the gray amorphous material are evidently derived from the decomposed augite ; it being not infrequent to find the border of a crystal irregularly outlined by the larger grains, while interiorly is the gray material and other black grains. The form and size of the black grains is such as to render their identification by the microscope alone impossible. The pulverized rock is, how- ever, strongly magnetic ; and after long digestion with hydro- chloric acid the solution gives a fiiint reaction for titanium. The aug^itic alteration would in this case seem to be similar to that of the hornblende in certain New Hampshire diorites described by Hawes.' That the altered mineral in this case is augite and not hornblende is proved by a few unchanged parti- cles still remaining and the form of the outlines still preserved. Viridite, so abundant in the other rocks described, is here en- tirely lacking. DETAILED STRUCTURE OF NANTASKET. The main purpose of this section is to set forth as fully as may seem desirable the facts upon which the generalizations of 'Geol. of New Hampshire, Vol. III., part IV., pages 4H and (5(i. 45 the ])reccding sections are based ; not forgetting- that ihc highest interest attaches to the relations and cspeeially to the contact phenomena of" tlie bedded rocks — conglomerates and lavas. For present convenience, as well as in the interest of fiituro students of the Nantasket ledges, each principal mass or area will, as regards the sedimentary rocks and associated lavas, be described somewhat independently and in topographic order, reserving the general correlation, the complete elucidation of the stratigraphy, until the close of this systematic itinerary. The structural details of the dikes, or the intrusive igneous rocks, will then be presented, in both chronologic and topographic order. The A^tlantic Shore oi' Coastal Area. This area, extending from Nantasket Beach to Black Rock, includes the most accessible and the most frequently visited of the Nantasket ledges ; and it is a fortunate circumstance that, although the melaphyr largely predominates here, some of the more characteristic features of Nantasket geology are well ex- hibited at this natural starting point. The granite, however, is wholly wanting ; and its interesting relations to the con- glomerate cannot be observed in this part of the field. Long Beach Rock Conglomerate. — The most northerly part of Long Beach Rock is a somewhat isolated, half-tide ledge of conglomerate. It contains many pebbles, mostly small, of granite and felsite, and also many larger pebbles of different varieties of melaphyr. From the latter we may infer that this bed is underlain by melaphyr, and that it is probably not the lowest or basal conglomerate. Eight or ten feet below the top of the conglomerate is an intercalated layer of fine, hard, red sandstone, eighteen inches thick ; and conformably overlying the conglomerate, are six to twelve inches of a beau- tifully banded, greenish slate of flinty hardness. The sandstone and slate show that the strike is N.65° E.\ and the dip S.E. 15°-20.° 1 All azimuth directions given in this paper are referred to the true meridian. 46 CO 'c^ :■/::::■:■ :x:;^ •!':/.'■ ■o'jiDllJr-jol -,-..'..vco"(l.?W, J>'0.»Oo o'o Ooi Melaphyr and Tuff .—The Long Beach Kock sediments are overlam conformably by a compact green melaphyr. The contact of the mel- ^ aphyr and slate is exposed for about ^ one hundred feet along^ the strike : ^ and at several points there are fine y transverse sections of the contact, g which IS accessible only at low tide. §. Fig. 1 is a general section of both II the conglomerate and melaphyr. 5 ^ The irregularities of the actual con- 2 I tact, and especially the bending of y ^ the subjacent laminas of slate, as well £ < ■? as the intense induration of this rock, ■^ ' ►§ which has been baked to a o-ood semi- •^ • .... li u ^ porcelainite, indicate that it was but ^ ^ imperfectly consolidated, if not en- g g tirely plastic, when the melaphyr M g flowed over it. But, on the other o ^ hand, the exquisitely beautiful brec- ° :g ciation and miniature faultins^ ex- X ^ hibited on some of the wave-washed ■J} "Vi § ^ surfaces show that before the dis- o § ^ ■::j turbance ceased the clay had be- 2 >^ come brittle enouo-h to break. This ^ ^ fascinating exposure is the gem of Nantasket geology ; and it is "^ hoped that it may long remain un- ^ ^ mutilated, to delight and instruct S future students. That the erujjtion I was submarine, or at least essentially ^ contemporaneous with the deposition ^ of the sediments, is beyond reasona- ble doubt ; although it is probable that the thin layer of green slate on which the melaphyr immediately Occas. pspers Bost Soc, Nat. Hist. IV. Plate 4. Wave-worn surface of the intensely hard, greenish, banded slate or TUFF separating THE CONGLOMERATE AND MELAPHYR ON LoNG BeACH Rock. Linear scale, one half natural size. [Reduced by photography from a tracing made directly from the stone.] Tliis impalpably fine sediment was deposited quietlj' and nniformlj' over the uneven, pebbly surface of the conglomerate; and consequentlv where erosion iias cut down nearly to the top of the conglomerate, each pebble determines the center of a series of concentric lines, the outcropping edges of the laminae of slate. In the portion of the eroded surface here represented, the banded slate has been worn away sufficiently to actually expose only two pebbles of the conglomerate. In the upper part of this thin bed of tuff, where it experienced more dis- tinctly the mechanical stresses of the overflowing lava, the layers are, to some extent, finely brecciated and faulted, as shown in the illustrations on the next plate. Occas. papers Bost. Soc. Nat. H'St. IV. Plate 5. o ^ Z ^ '^ 3; o o 47 rests should be regarded as an exceedingly fine volcanic tufi, the eruption of liquid lava having been preceded by an out- burst of volcanic ashes. The influence of the lava flow has also penetrated the conglomerate, the layer of sandstone eight feet below being almost as thoroughly indurated as the slate. The breadth of the melaphyr is fully two hundred feet and its actual thickness probably not far from sixty feet. It is, for the most part, rather compact and of a dark green color. But toward the top especially it is full of irregular segregations of quartz and shows, in a rather large way, a very distinct and somewhat wavy flow-structure, resembling bedding and parallel with the stratification of the underlying conglomerate. Irreg- ular veins and masses of compact and impure epidote, occa- sionally of considerable size, also occur in the upper part of the melaphyr. It is very plain that the entire thickness of mela- phyr is to be regarded as forming one simple and normal flow homogeneous and massive below, but superficially brecciated and scoriaceous. This great lava-flow is undoubtedly continued eastward in the outermost part of the ledge exposed at low tide north of Centre Hill ; and it appears probable that further east the line of strike changes so as to connect Little Black Rock with the same bed. This islet, which is about four hundred feet long and half as broad, is entirely composed of melaphyr very similar to that on Long Beach Rock, except that it is in part amygdaloidal. The prominent ledge at the northwestern base of Atlantic Hill, and partially isolated by the sands of Nantasket Beach, is chiefly a distinctly and evenly bedded greenish gray sandstone, alternating, especially in the upper part, with layers of greenish and rather indistinct, small-pebbled conglomerate. This con- glomerate is made up of much smaller pebbles, and is in every way very distinct from that on Long Beach Rock. The peb- bles are mainly more or less angular fragments of melaphyr, are not well assorted, the coarsest and finest material being 48 mingled indiscriminately, and, with the exception of a rare pebble of felsite or granite, the entire ledge, sandstone as well as conglomerate, appears at first to have been derived from the Long Beach Rock flow of melaphyr. There is much, however, in both the composition and texture of this mass to suggest that it may be a true volcanic tuff, a record of explosive volcanic action either at the close of the eruption already described or at the inception of the similar littoral or submarine eruption by which this bed was covered. It will be observed that this in- terpretation is adopted on the map, the lava-flows being thus supposed to have followed each other so closely over this marine area as to preclude the deposition of any appreciable amount of the ordinary or normal sediments between them. This bed of tuflP strikes N. 65° E., and dips S.E. 15°. Its full breadth appears to be exposed, viz., 120 feet, equal to a thickness of about 30 feet. It can be traced for about 300 feet along the base of Atlantic Hill, rising at the highest point some 20 feet above the beach. The line of strike carries it directly across the shingle beach connecting Long; Beach Rock and At- lantic Hill. I have been unable to positively identify it on the low-tide ledge north of Centre Hill ; but it seems impossible to doubt that it crosses here, although probably with diminished thickness. It appears necessary, also, to suppose that the tuff overlies conformably the melaphyr of Long Beach Rock, al- though there is no exposure of the contact. If the tuff extends so far to the eastward, it must pass to the north of Gun Rock and between the two Black Rock islets. This old ash-bed is overlain clearly and conformably by a second bed of melaphyr, the contact being well exposed at sev- eral points along the northern base of Atlantic Hill. The con- tact is similar to that between the green slate and overlying melaphyr on Long Beach Rock — conformable in the general view and yet with many minor irregularities. Fig. 2 shows the contact as it appears at the edge of the beach, below the bath house and near the foot of the Atlantic House steps. It is clearly such a contact as would naturally result from the 49 flowing of liquid lava over uncousolidated sediments. And the attention of those who may see in the minor uneonformities of the contact evidence that the melaphyr is really intrusive is called to the fact that no detached fragments of the tuff or sedi- ment are observed in the base of the melaphyr. M£i4^p'^y^ J. M 7? "^^^TT! j^* ' * \ *' I '"V* ' • ^ . J **^- •• /C\ ■•.*.■.*.•.■>*•'• •*•> ^.'' .•'•".^". /• Vf-^' !" •' * „• -',•*•*. • \^ • .' • • • * ^^1* • '.^H* \ /■'*■'•'. '■ ',•'"■'. •^' .* •*. •* • » . • •*.*•' / * •■•.-•* •-,•<•' '. fc** * '.* * • * • • Tuff /S??4^ --' *••• ,,. ."*•' ..*'*•" '•5~^ "'t"-sl. ^-<^2^-$'Se~ ^5s-c^.'. \ • r* I''*- • \ ' *.'4 v.*. I .' . •* Beac^ ^ Fig. 2. — Contact of Tuff and Melaphyr at the northern base of atlantic hill. Scale, i inch = 12 feet. The thickness of the melaphyr, near the bath house, is about twenty feet ; and it appears to increase eastward. This is very clearly a second lava-flow, and not an intrusive bed ; and in lithological character it is very similar to the first, being chiefly dark green and compact. It is sometimes, however, more brec- ciated ; but exhibits the same ill-defined segregations and veins of quartz and epidote. This flow can be traced eastward along the base of Atlantic Hill to Valley Beach, between Atlantic and Centre Hills ; and it undoubtedly forms the landward end of the low-tide ledge north of Centre Hill. Valley Beach is bounded by two good transverse sections of the formations. Atlantic Hill ends here in a small clifl', which shows the following thicknesses from below upwards (Fig. 3) : The second melaphyr, already described, about 30 feet; volcanic agglomerate and tuflf, 7 feet ; melaphyr, 9 feet ; ag- glomerate and tuif, 25 feet; and then melaphyr extending southward indefinitely, or beyond the end of the section. The first fragmental bed or tufl" in this section measures 7 OCCAS. papers. B. 8. N. H. IV. 4. 50 L V. > \i- X ~i 1 ft. V 1 -«. JiMI /< A VA d IE Fig. 3. — Section on the west SIDE OF Valley Beach. Scale, i inch = 20 feet. 51 feet on the cliff, but grows narrower westward and appears to die out entirely between 200 and 250 feet from the cliff. It is probable, however, that the tuff really persists until it passes the edge of the overlying melaphyr, and is merged with the next bed of tuff, as shown on the map. The boundaries of this tapering edge of the fragmental bed are quite irregular. This is especially true of its lower surface, where it rests upon the second melaphyr. The form of this contact for 130 feet west from Valley Beach, or as far as it is clearly exposed, is shown in Fig. 4 ; for the sake of convenience, however, the curves are represented as following each other more nearly in the same direct line than is actually the case. The upper surface of the melaphyr presents smoothly rounded hemispherical protuberances or swellings one to thi-ee or four feet in diameter. The actual boundaries of adjacent protu- berances may usually be traced below the surface of the mela- phyr ; and some of these remarkably regular and graceful curves thus describe two-thirds or three-fourths of a complete circle. Probably the majority of the curves are really sections of rounded ridges or rolls, — the surface flow-structure of the lava. But it is very noticeable that these prominences exhibit semicircular pro- files to some extent on both the transverse and longitudinal sec- tions of the bed, suggesting hemispherical tumefactions or superficial bubbles formed on the liquid lava. It seems impossible to regard these features as the product of erosion ; they must be entirely original ; the actual surface of the fresh submarine lava-flow, which was covered almost in- tact either by the ashes accompanying a second flow or by the debris worn from adjacent masses. The fragments are mainly angular, ranging in size from dust to three inches in diameter ; and this is clearly a thin, local deposit on the surface of the sheet of lava. The well-marked north-south depression separating Atlantic and Centre Hills, and terminating at the lower end in Valley Beach, does not seem to be occupied by a dike, for the melaphyr shows half way across the bottom of it, and there is not a ves- 52 tige of dike rock to be seen in situ ; nor does it appear to be chiefly due to faulting, but ratiier to rapid erosion along a series of close parallel joints. There has been a little slipping, how- ever ; for on tracing the thin bed of tuff across the one-hundred feet of beach, we find its outcrop shifted about twenty feet to the south, indicating an upthrow on the east of perhaps five feet. The melaphyr between the two walls, may, of course, have dropped down almost any amount, producing two con- vero'inof and compensatino- faults. East of the beach the bed of tufi" already described crosses the ledges diagonally to the shore in about one-hundred feet, passing out of sight under the water without any sensible change of dip or thickness. Returning to the west side of Valley Beach, the melaphyr, nine feet in thickness, overlying this tuif is found to be quite conformable, the contact showing only the minor irregularities that would naturally be developed where a stream of lava flows over unconsolidated sand and gravel. The contact is not always a sharply defined line, but the lava is enough mixed with the sand so that the two rocks are blended through a thickness of several inches. The swelling curves charactei'izing the lower surface of the tufi" are not observed above it. The structure of this melaphyr, which is really the third flow, reckoning from the conglomerate on Long Beach Rock, is quite peculiar. A compact, greenish matrix encloses iri'egu- larly rounded amygdaloidal masses from two inches to two feet in their longest dimensions. The amygdules are usually ar- ranged in concentric lines or zones parallel with the exterior ; and the coarsest amygdules are sometimes towards the periphery and sometimes in the centre. These masses are quite clearly distributed in irregular lines parallel with the bed ; and this tact, as well as the great number of the masses, is decidedly unfavor- able to the view that they are true volcanic bombs. But the best explanation which has occurred to my mind is, perhaps, not wholly satisfactory, viz., that, during the flowing of the lava, vesicular layers and crusts were, by the unequal flowing and revolving motions, broken up and the fragments rounded 53 into the forms we now see. East of Valley Beach, along the north front of Centre Hill, this melaphyr expands rather rap- idly ; and it is very obvious that while the rounded amygdaloid- al masses or pseudo-bombs are thickly and pretty uniformly scattered through the middle and upper parts of the flow, they are almost entirely wanting in the lower part. They continue to form a prominent feature of the upper part of the melaphyr about half way across Centre Hill, and then die out rather gradually. Beyond this point the identity of the melaphyr as a separate flow is lost, since it is indistinguishable from the over- lying melaphyr, with which it is here in direct contact. That the third melaphyr actually extends quite across Centre Hill, if not across Gun Rock, there can be no reasonable doubt. This flow is of exceptional interest, not alone on account of the pseudo-bombs, but also because it aflbrds, perhaps, the only instance in the Nantasket area where the tapering edge of a bed of lava can be clearly traced. It certainly thins out rapidly toward the west, and, as stated, can not be followed much more than two hundred feet west of Valley Beach, the under- lying and overlying tuffs appearing to come together at this point. It will not escape attention that the amygdaloidal masses are thus found where we should most naturally look for them — in the marginal and superficial portions of the flow. The second bed of tuff in the Valley Beach section, or that overlying the amygdaloidal melaphyr, is lithologically similar to that below the melaphyr ; but it is three times as thick, and can be traced westward the entire length of Atlantic Hill, with no sensible change of dip or thickness, passing under the beach on the west where the County or Beach road first reaches the base of the hill. It is well exposed near the Atlan- tic House steps, just above the bath house ; from here it con- tinues around the hill a little above the middle of its height ; and then gradually descends to the level of the beach on the east. The continuity of the bed is clear ; but its strike must change from N. 65^ E. on the west to at least N. 75° or 80° E. on the east side of the hill. Its upper and lower contacts are 54 clearly exposed at several points, and are favorable to the view that the melaphyrs are contemporaneous. It consists, like the other beds of tuff, almost w^holly of more or less angular and imperfectly assorted fragments of melaphyr imbedded in a dis- tinctly stratified cement of the same character. On the west side of the hill it rests directly upon the second melaphyr ; while in the vicinity of Valley Beach the second tuff and the third or amygdaloidal melaphyr are interposed, as already ex- plained. East of Valley Beach it is scantily exposed, and can not be traced much beyond the Waverly House. But it is here, seemingly, somewhat interstratified or mingled with the underlying amygdaloidal melaphyr. Above the last described or third tuff comes the great body of melaphyr forming all the remaining portions of Atlantic and Centre Hills and extending south to Conglomerate Plateau, as shown on the map. The outcrops are so numerous over this area that there is absolutely no room to doubt the essential continuity of the melaphyr. Its breadth, measured from the third tuif, along Valley Beach Avenue, is nearly 1400 feet, cor- responding to a probable thickness of more than 300 feet. Or, classing the three beds of tuff as fragmental lavas, as the facts appear to warrant, and thus regarding the volcanic series as essentially continuous back to the conglomerate on Long Beach Rock, the breadth of the entire series of three tuffs and four melaphyrs is, in round numbers, at least 1800 feet, and thick- ness 450 feet. The great mass of melaphyr above the third tuff is fairly uniform lithologically. It is usually more or less brecciated, with numerous highly iiTcgular and limited segregations of vitreous quartz and chalcedony, as well as epidote. But sometimes it is much more distinctly brecciated, especially im- mediately above the third tuff. On Atlantic and Centre Hills it is rarely amygdaloidal or quite compact in texture. Near the southern border of the melaphyr area, on the north side of Willow Ledge Hill, it encloses about twenty feet in thickness of greenish tuff and agglomerate. The green arenaceous tuff ia 55 interstratified, especially in the lower part of the section, with a bright red, slaty rock, which is somewhat contorted ; and in the upper part of the section with the agglomerate, which is chiefly composed of quite small, but mostly angular fragments of melaphyr. These beds, which are very distinctly stratified throughout, strike about N. 80° E., dip S. 15'^, and ai-e overlain conformably by the melaphyr ; but the lower contact is not ex- posed. On the west they are quite certainly cut oflTby afault, and this, as shown on the map, is probably also their fate on the east. At any rate, the melaphyr clearly crosses their strike in this direc- tion, and it is impossible to trace them for more than three hun- dred feet. A red slate or tuff similar to that found here, crops out just west of the Rockland House, accompanied, appai'cntly, by agglomerate, although this may be only the ordinary con- glomerate or puddingstone. The correlation of these red out- crops would require us to postulate dislocations which are not indicated on the map and for which there is no independent evidence. It appears wiser, therefore, to regard them as local and non-synchronous accumulations of volcanic dust and lapilli. The melaphyr immediately northeast of the Rockland House is similar to that on Atlantic Hill, and is undoubtedly a con- tinuation of it : and the same is true of the melaphyr forming Gun Rock and its neighboring ledges. The bedsof tuff on the north side of Atlantic Hill cannot be traced west of the abrupt western slope of the hill, perhaps for want of outcrops ; nor can they be followed eastward more than half way across Centre Hill. But in this direction the various flows of melaphyr are, apparently, united ; and in the Gun Rock district there is not a trace of any sedimentary rock, but the true conglomerate and the fragmental lava or tuff are alike wanting. It would be easy, of course, to explain the absence of the tuffs by faulting, but it is, perhaps, more probable that these fragmental lavas were never spread over this area. The submerged or half-tide ledge east of Gun Rock (see map) is about three hundred feet long at low tide, and consists wholly of melaphyr similar to that of Gun Rock. There is, apparently, no reason to doubt that all the mela- 56 phyr described up to this point, with the inchided tuffs, on Rockland, Atlantic, and Centre Hills, Gun Rock, the sub- merged ledges and Little Black Rock, belongs to one great bed or sheet having an approximately east-west trend and southerly dip and overlying the conglomerate forming the north end of Long Beach Rock. Throughout this great mass, however, there are many indications that it is really composite, consisting of a succession of flows ; which either were not submarine, or followed each other so rapidly as to preclude the formation of any sensible thickness of intervening sediments. The included beds of tuff show this very clearly. The zones of brecciated melaphyr, the fractures being marked by segrations of silica and impure epidote, seem to show it. And there are often curving lines, such as have been described at the base of the second tuff (Fig. 4), which appear to mark the contact of two successive flows. On the north side of Centre Hill, near the end of Centre Hill Avenue, is another proof, in the form of a dike of mela- phyr (No. 1, on the map), that this great bed is not throughout of the same age. Its compactness and general resemblance to the enclosing melaphyr, as well as the irreg- ular contacts, mark this dike as probably contemporaneous with some later flow of this immediate series, although, of course, of later date than the flow which it is seen to intersect. The general absence of the amygdaloidal texture in this mel- aphyr, except in the flow separating the second and third tuffs, is very noticeable ; and this negative character, as well as the prevalent brecciation, contrasts it strongly with the most of the other masses of melaphyr or basic lava in the Nantasket area. In fact, as Mr. Merrill's descriptions show, both the structure and composition of this rock assign it a position between the ultra basic lava or typical melaphyr, and the more acid lava or porphyrite of this region. Porphyrite. — The remaining rocks and islets, east of Gun Hock, constitute a very typical area of porphyrite, bounded on 57 the north, apparently, by the eastward extension of the i^reat belt of melaphyr indicated by Little Black Rock, and on the south by the conglomerate of Green Hill. Although, as Mr. Merrill states, this rock resembles the melaphyrs more in its general aspect than does the porphyrite of the western part of the Nantasket area ; chemical analysis shows that it is as acid as any porphyrite in this district. It is somewhat variable in its microscopic features, but usually presents a compact, or apparently felsitic, dark gray or purplish to nearly black, base, enclosing numerous porphyritically developed feldspars or minute aggregates of epidote. It is never amygdaloidal, and rarely distinctly brecciated ; but exhibits at several points a well-developed striping or flow-structure. Black Rock, which is about seven hundred feet east and west, and half as broad, is a continuous and almost perfectly bare mass of porphyrite. Fluidal lines are very plainly marked in a por- tion of the porphyrite, with usually an approximately east-M'^est trend and a gentle southerly dip, thus proving that the normal attitude of the rocks, so clearly exhibited about Atlantic Hill, is preserved to the very eastern limit of the Nantasket area. Near the middle of the west half of the island, a considerable mass of porphyrite is split up by close , parallel east-west and nearly vertical joints so as to present a very shaly appearance. The character and especially the attitude, or dip and strike, of this structure indicates that it is possibly true cleavage rather than jointing, and analogous to the foliation of the diabase on Calf Island.^ The small island north of Green Hill (Green Hill Rock) is another mass of gray and purplish porphyrite, with flu- idal lines ; and the half-tide ledge between this and Black Rock exhibits a similar constitution. Since the porphyrite of this area is exposed only in these three insular masses, it is mani- festly impossible to determine with certainty either its bounda- ries, its thickness or its relations to the bordering melaphyr and conglomerate. If, however, we may follow the lithologic indica- tions and correlate it with the porphyrite of the western area, it must, as will appear later, be referred to a horizon below this iProc. B. S, N. H. xxni., 455, 58 melaphyr and conglomerate and thus owe its exposure to exten- sive faulting. The map, it will be observed, has been constructed in accoi'dance with this view ; and reference to the table of Nan- tasket strata on page 24 will afford some idea of the magnitude of the displacements which it involves, — at least 50 feet on the side toward the conglomerate and 400 feet on the side toward the melaphyr. Indications are not wanting that Black Rock is near the point or centre of emission of the Nantasket porphyrites. One evi- dence of this is the seemingly great thickness of the porphyrite in the Black Rock area, and its more crystalline character as compared with the porphyrite in the western part of Nantasket. The most important fact pointing to this conclusion, however, is the series of porphyrite dikes on the Cohasset Shore east of Green Hill. The most casual observation shows that the ap- proximately north-south dikes on the Cohasset Shore can not all be referred to the third system of diabase dikes ; but besides the three well-defined systems of dark -colored, finely crystalline diabase dikes, there is evidently a fourth system having a gen- eral north-south trend, consisting of a distinctly greenish rock, which varies in texture from apparently felsitic and porphyritic to visibly holocrystalline or nearly so. The dikes of this kind are not only lithologically but chronologically distinct, for they are repeatedly cut by both the east-west and north-south dia- base dikes ; and hence, although agreeing approximately in trend with the newest series of dikes, they must be regarded as the oldest system exposed on this shore. My interest in this oldest system of dikes was not arovised so much by the clear proof of their age as by the marked litholog- ical resemblance which they bear to the more acid lavas or porphyrites of Nantasket, and especially to the porphyrite of the Black Rock area. Except that the porphyrite fiows, al- though often visibly epidotic, arc rather rarely distinctly green- ish in color, while the dikes of this system arc always so ; and that the dikes are more crystalline than the effusive rocks, as we should naturally expect ; while the latter exhibit commonly 69 a striping or flow-structure which is wanting in the dikes ; the superficial resemblance is certainly very obvious, suggesting at once that the dikes may be the channels or vents througli which the effusive porphyrite reached the surface, or at least referred to the same eruptions — the same period of volcanic activitv. Mr. Merrill has found (page 38) that, although somewhat more crystalline, these dikes are essentially similar in micro- scopic characters to the Black Rock poryhyrite ; and in order to further test the validity of this hypothesis a typical example from one of these dikes was submitted, through the kindness of Dr. T. M. Drown and Mr. G. F. Eldridge of the Massachusetts Institute of Technology, to partial chemical analysis, with the following result, the mean of four accordant determinations : silica, 60.84 per cent. ; alumina and ferric oxide, 20.12 per cent. ; with which may be compared the percentage of silica — 58.25 per cent. — afforded by the Black Rock porphyrite. As in the case of the melaphyr (page 38), we find that the dike is slightly more acid than the effusive portion of the eruption ; its more crystalline character causing it, apparently, to yield less completely to the deep-seated alteration, thus reversing the normal relative proportions of silica observed in recent erup- tive rocks. That the classification of these dikes by the superficial char- acters is unsafe is shown by a second analysis, by Mr. Eldridge, from another dike of the series. This afforded, as the mean of four accordant determinations, silica, 48.47 per cent. ; alumina and ferric oxide 31.53 per cent. ; and if not a true diabase, it should, apparently, be associated with the more basic melaphyrs of Nantasket. Obviously, then, a systematic microscopic and chemical examination of these dikes will be required for their accurate classification ; and we can now only assume that, as appearances indicate, they are chiefly porphyrite. A more detailed description of these dikes will be presented later, in the systematic account of the dikes of this region ; and all that I desire now is simply to call attention to the following points ; — (1) That they are probably of the same age as the 60 surface sheet of porphjrite and connected witli the vents through which it was extruded. (2) That no dikes of this character have been discovered west of Green Hill or more than one mile east of it. (3) That they all have, without exception, a northerly trend, i. e., they either run toward the Black Rock porphyrite or so as to intersect the eastward extension of its line of strike. And yet (4) none of them have been found breaking through the porphyrite itself, these dikes being thus comparable with the great melaphyr dike (No. 2), since in both cases the complete erosion of a portion of the effusive accumu- lations was essential to expose the intrusive masses — the roots of the eruptions — in the underlying granite. G-reen Hill Conglomerate. — The prominent ledge of con- glomerate which has been laid bare by the wearing away of the northern slope of Green Hill, is probably, as the fault-lines of the map indicate, a part of the great bed forming Folsom's Island and Conglomerate Plateau ; but it appears more con- venient to describe it in its geographic than in its geologic sequence. The ledge is about four hundred feet long from northwest to southeast, and shows throughout a southeasterly dip of 20° to 30°, giving an apparent thickness of perhaps 175 feet. Beginning on the southeast, there are 10 or 12 feet of medium conglomerate with distinct layers of reddish sandstone, showing the bedding well. This is underlain by about 60 feet in thickness of very coarse conglomerate, many pebbles, espec- ially of granite, ranging from six to eighteen inches in diameter. The remainder of the section is conglomerate of a more normal character, varying from medium to rather coarse in texture. The conglomerate is composed throughout of well rounded masses of granite (coarse and pinkish), felsite of different vari- eties, and porphyrite ; but, appai-ently, little true melaphyr. The porphyrite is mainly of compact, dark gray and brown va- rieties. 61 The Central Area. This area embraces all that part of Nantasket east of the rail- road and south of Atlantic and Centre Hills. The oeoloj-ical relations of the central area to the preceding or coastal area can be most easily traced in Willow Ledge Hill, which belongs in part to each. The summit and main mass of this elevation is a nearly square block of melaphyr and tuff. These rocks have been described in the preceding section, and the limits of this block are sharply defined on the map by the bounding fault- lines. The similar block of conglomerate on the west lies at a lower level, and its outcrops are separated from the volcanic block by a small artificial pond. Both of these blocks are separated from the narrow ridge of conglomerate between them and Hull Street by the well-marked east- west depression occupied by Spring Valley Road. This depression is an evident fault-line, for the beds of conglomerate are so nearly horizontal that they must abut directly against the melaphyr ; and the contact is undoubtedly oblique to the strike of the conglomerate. As indicated on the map, this is one of the great faults of the Nantasket region, traversing the entire area, from Strait's Pond to Weir River Bay. Few of the dis- locations of this district are more obvious, and none are of greater structural importance. East of Valley Beach Avenue it is marked by the well-defined escarpment terminating the melaphyr of Centre Hill on the south, the conglomerate lying at a lower level here, under a meadow ; while to the westward it coincides with the similar escarpment along the north side of Melaphyr Plateau, and, crossing the bay, defines approximately the abrupt northern shore of the western area. This line of fracture is thus clearly the natural boundary line between the coastal and central areas. But when we pass to the considera- tion of the direction and amount of the displacement along this line, and the stratigraphic relations of the two areas, a much more difficult problem confronts us. The escarpments, which 62 face now to the south and then to the north, evidently signify nothing but unequal erosion, appearing always on the side of the harder and more resistant rock, which is usually melaphyr. Facts will, however, be accumulated in the following pages pointing to the correlation of the conglomerate of Conglomerate Plateau with a horizon below that of the conglomerate on the extremity of Long Beach Rock ; and hence to the conclusion that the downthrow is not only on the north, but that it must exceed the entire thickness of the melaphyr series of Atlantic and Centre Hills, — four hundred if not five hundred feet. Although melaphyr occurs abundantly in the central as well as the coastal area, it is of a diflferent and distinctly more basic type. It is very generally and often profusely amygdaloidal, rarely distinctly brecciated and never, so far as observed, ac- companied by beds of tuff. The abundance of conglomerate south of the great fault, as well as the numerous faults, still further contrasts the two areas. It is, of course, quite possible, if not probable, that some of the faults of the coastal area have escaped discovery on account of the uniform character of the melaphyr in Atlantic and Centre Hills ; and the great apparent thickness of this melaphyr ought, perhaps, to be regarded as largely due to repetition by faulting. The alternations of the conglomerate and melaphyr in the central area are so frequent that it will be more convenient to describe these rocks together, following the order of the ledges or t02:)ographic features. East shore of Nantasket Bay, Crescent Hill and Marsh Island. — These localities, forming a narrow belt across the western end of the central area, between the railroad and Mela- phyr Plateau, may be conveniently described together, since, as the map shows, they present, as the result of parallel faulting, simply repetitions of certain beds of conglomerate and melaphyr. This tract is bordered by two well-marked north-south faults, and this primaiy fault-block is divided by no fewer than twelve transverse fractures, witli the slips alternating in direction, and indicating the unequal rising and sinking of V-shaped blocks 63 (Fig. 8). Commencing at the north end, the first ledges of conglomerate and melaphyr rise somewhat abruptly from the marsh and the bay immediately south of the great east-west fault, and about five hundred feet, following the railroad, south- west of the broad, low outcrop of red and green slate or slaty tuff already described ; and, as the map shows, the section is almost uninterrupted along the east side of the railroad from this point to the south side of Marsh Island. Melaphyr. Conglomerate. ■ Fig. 5. — Section showing the fault-dike (25) and the contact between the third conglomerate and second melaphyr, south of the fault. scale, i inch =^ 30 ft. It begins with the typical melaphyr of the central area, a dark green, evidently basic and highly amygdaloidal variety. The amygdules are not uniformly distributed ; but they are especially abundant in a superficial layer of the rock from two to three feet thick. They are usually somewhat elongated and have a distinctly linear arrangement parallel with the layer ; so that the flow-structure of the melaphyr as a whole is very marked. This mass of melaphyr has an exposed breadth of about forty feet north and south and is divided midway by dike 26 ; while it is abruptly terminated on the south by dike 25. This dike un- doubtedly marks an east-west fault ; for it is bordered by con- glomerate on the south (Fig. 5). The fault probably hades with the dike to the north, and the melaphyr is clearly on the downthrow side. Not more than eight feet in thickness of the conglomerate are exposed before it is overlain by melaphyr. 64 The contact is very sharply defined, and most admirably exposed ; and it is conformable to the flow-structure in the melaph^'r, dipping south-southeast 5°-l(y. This melaphyr appears to be identical with that north of the fault, the amyg- dules and flow-structure being even more pronounced ; and they may be fairly regarded as parts of the same flow. The displace- ment here has been suflficient to conceal the conoflomerate and to prevent its reappearance at the northern edge of the mela- phyr, perhaps fifteen feet as a minimum. It should be stated, however, that the melaphyr on the north or downthrow side appears to be about horizontal or to have a slight northerly dip. Above the conglomerate, on the south side of the fault, the melaphyr has a breadth across the strike before we come to the overlying conglomerate of nearly one hundred feet and an appar- ent thickness of less than twenty feet. The true thickness, how- ever, is pi'obably not less than twenty-five or thirty feet, the apparent diminution being due to faults with the downthrow to the south accompanying dikes 29 and 31. Southwesterly or parallel with the railroad the sloping edge of the melaphyr flow is exposed broadly and almost continuously as far as dike 35. Near the railroad this dike marks the contact of the melaphyr and overlying conglomerate. The exposure is not very satis- factory ; but there seems no reason to doubt that the melaphyr passes under the conglomerate conformably and wit'nout appre- ciable faulting. Following the contact to the northeast, it sweeps around in a curve concave toward the melaphyr. The two rocks are seen near together, but not in actual contact, un- til we come to the large inclined dike (31). Here a slight fault with the upthrow to the north brings the melaphyr into sight directly under the conglomerate (Fig. 6). This conglomerate overlying the melaphyr is similar to that underlying it. It is, however, coarser and more irregular in composition, containing many rounded pebbles of granite and felsite 6 to 12 inches in diameter ; and it is especially distin- guished by holding many large pebbles of melaphyr, some of which are coarsely amygdaloidal and apparently identical with or, that on which the conglomerate rests. The conglomerate outcrops broadly south of dike 35, forming a low hill ; and a few streaks of sandstone show that the bed dips southeast about 10°, so that the thickness can scarcely exceed 25 feet. It is distinctly overlain on the southeast corner of the outcrop by a second flow of amygdaloidal melaphyr. The surface of the ledge slopes with the dip, and thus a thin layer of melaphyr covers quite a breadth of the conglomerate. A small triangular patch of the overlying melaphyr is also seen resting on the conglom- erate immediately north of dike 35, indicating, apparently, a slight fault with the downthrow to the north on the line either of this dike or of the small dike (3G) which is, probably, a branch of the main dike. Comrlotnerate. ^ Melaphyr. Ca/j^-'omerafe. Fig. 6. — Section showing the fault-dike (31) and the contact BETWEEN THE SECOND MELAPHYR AND THE FOURTH CONGLOM- ERATE, NORTH OF THE FAULT. ScALE, I INCH = 30 FEET. Between dikes 29 and 31 the conglomerate crosses the north- south fault bounding Melaphyr Plateau and extends eastward into the melaphyr about 100 feet with a breadth of 75 feet. This extension, being on the upthrow side of the fault, is ele- vated above the rest of the bed some 20 feet. At its south-east corner the extension is overlain by amygdaloidal melaphyr of the same character, in precisely the same manner and with the same southeasterly dip (10'-15°) as at the southeast corner of the lower area. The extension is undoubtedly cut off b}^ a fault on the south, with the downthrow in that direction, so that the contact of the conglomerate and overlying melaphyr in the OCCAS. PAPERS. B. S. N. H. IV. 6. f.f. extension can not be traced far in the direction of the same contact in the main area. The conglomerate extension is bonnd- ed on the north by dike 29, which clearly marks a fault with the downthrow also, probably, to the south. Hence the mela- phyr north of the conglomerate extension appears to be that normally underlying the conglomerate, and therefore distinct from, and separated by a fault from, that overlying the con- glomerate. The small meadow or marsh sepai'ating the conglomerate south of dike 35 from Crescent Hill must cover a fault with the downthrow to the north, as indicated on the map ; for the bed of conglomerate first described and the underlying and over- lying melaphyrs are repeated, with a stronger topographic re- lief and fewer dikes, in the hill. . The conglomerate forms the summit of the northwestern part of the hill and extends en- tirely around the northern and western slopes, its outcrop thus havino; the form of a crescent. It is the irreo-ular or mixed coarse and fine conglomerate, containing pebbles of amygda- loidal melaphyr as Avell as of granite and felsite. The contact with the overlying melaphyr is very plainly exposed on the summit. It is a true effusive contact and has a southeasterly dip of 10° — 20°, being conformable with the bedding-planes of the conglomerate. This overlying melaphyr shows several dis- tinct zones of amygdules, as well as other indications of flow- structure, and altogether exposes a thickness of 25 to 35 feet. The contact is also clearly exposed for fifteen feet at the base of the hill on the east side. The thickness of the con- glomerate can scarcely exceed 30 or 35 feet. The underlying melaphyr exposes a thickness of about fifteen feet at the north- west corner of the hill, the contact descending both east and south. Near the middle of the west side of the liill the under- lying melaphyr appears again, being elevated some ten feet above the railroad by an east-west fault with the upthrow to the north (Fig. 7). This fault gives the conglomerate and underlying melaphyr a distinctly synclinal structure in the fiT north-south line, the axis of the syncline clipping east 10° or more ; which prevents the mclaphyr from reappearing on the east side of the hill. Consrlomerate. Melafhyr. Fig. 7- — Section across the fault On the west side of crescent HILL. scale, I inch = 50 FEET. Crescent Hill is, in fact, divided by a series of east- west faults. Commencing on the north, the first fault is 65 feet from the northwest corner of Crescent Hill and exactly opposite the noi'thcrn end of Great Hill. It downthrows to the north about eicjht feet, iogofino- the contact of the cono-lomerate and lower melaphyr ; and, since it does not now cut the upper melaphyr, the effect is to increase the apparent thickness of the conglom- erate. The second fault is tliirty feet farther south, it hades S. 25° and downthi^ows in the same direction about eight feet, being compensating with reference to the first fault. Twenty- five feet farther south is a small dike (43) and the only one observed on the hill. Accompanying the dike, but not coin- ciding in plane with it, is the third fault, which hades and downthrows to the south, the throw being about eight feet, and the beds sloping steeply from the fault on the north. The next or fourth fault, seventy feet farther south, is that already re- ferred to, on the west side of the hill. This is one of the plainest and most instructive faults in the Nantasket district. Figure 7 represents a general view of the fault as seen from the railroad, taking in the whole height of the hill. The fracture is still a well-marked and somewhat open fissure ; and the jogging of the contact of the conglomerate and upper mela- phyr shows that the downthrow is to the south and about 68 eighteen feet. The lower nielapliyr very clearly rises to the south or toward the fault, and again to the north, giving the syncline already described. The fault trends N. 75° E., being in line with dike 37 in Great Hill ; and it can be readily traced directly across the entire breadth of the hill, following approxi- mately the boundary line between the conglomerate and upper melaphyr, and breaking this contact almost as distinctly on the eastern as on the western face of the hill. About midway be- tween this fault and the southern end of the conglomerate is a fifth fault, which is the counterpart of the last, the downthrow being to the north and about 15 feet. It plainly jogs the contact between the conglomerate and upper melaphyr; and trends N. 75° W., being in line with dike 44 on Great Hill. Finally, at the ex- treme southern end of the conglomerate there appears to be a sixth fault, probably parallel with the last, but with the down- throw to the south. The amount of the throw, if to the south, is certainly 10 if not 15 feet, being enough to throw the con- glomerate entirely out of sight. If the throw were to the north it would need to be 25, and more probably 50, feet to do this. The fact that the melaphyr under Crescent Hill, although rising to the west, can not be found, not even the slightest trace of it, on Great Hill, is alone sufficient to prove an important north-south fault between the two hills, with the downthrow to the east ; and the further fact that not one of the dikes in Great Hill can be traced in Crescent Hill points to the same conclu- sion. Indeed, the contrast presented by the opposite walls of this narrow defile is one of the especially surprising and puz- zling features of Nantasket geology. The very abrupt way in which the conglomerate of Crescent Hill terminates on the east proves that the hill is bounded in this direction also by an im- portant north-south fault, the downthrow probably being as before on the side of Crescent Hill, or to the west. Marsh Island is a large ledge of conglomerate and melaphyr completely isolated by the eastern marsh. Geologically it is but a repetition, at a lower level and without important dis- placements, of Crescent Hill. The conglomerate, which forms 60 the main mass of tlie ledge, is of the coarse, irregular type seen in Crescent Hill ancVon the east shore of the bay, ■^ n containing many large pebbles of gran- m ^ ite, felsite and melaphyr. The breadth of the conglomerate outcrop, from northwest to southeast, is about 140 feet ; and the thickness of the bed, the upper and lower surfaces of which are clearly exposed, is about 25 or 30 feet. It is underlain on the northwest, at the very base of the ledge, by a highly scoriaceous and amygdaloidal green melaphyr, of which only a slight thick- ness is exposed ; and is overlain on the southeast by a very similar flow of melaphyr, which at one point reaches up over the conglomerate nearly to the top of the ledge. These are undoubt- edly true effusive contacts, and the up- per one especially shows the dip of the conglomerate to be S. E. 10° — 20°. The melaphyr contains abundant seg- regations and veins of epidote and red jasper. Since there appears to be no reason whatever to question the. strati- graphic equivalence of Crescent Hill and Marsh Island, it is necessary to suppose that the marsh between them conceals an important fault ; the down- throw being to the north and , fully equal to the combined thickness of the conglomerate and upper melaphyr, about 50 or 60 feet. The foreo-oino; details, which are summarized in the north- south section (Fig. 8), show that the various stratigraphic 1) 70 features of this belt of ledges are readily explained by one bed of conglomerate between two very similar flows of melaphyr, with only such dislocations as are actually seen or may be rea- sonably inferred, until we come to the conglomerate underlying the lower melaphyr south of dike 25. This is a partial expo- sure only of a lower and older conglomerate which does not appear elsewhere in the section. Although the transverse faults of this block are frequently re- versed in throw and compensating, the northerly throw clearly prevails, the sum of all the displacements giving 161 feet to the north and 71 feet to the south. Hence the entire series is equal to one northerly slip of 90 feet. Melaphyr Plateau. — This rectangular block of melaphyr is sharply defined by the bounding fault lines and escarpments, and may be regarded as a very characteristic feature in Nan- tasket geology. With the exception of the small patch of con- glomerate on the western edge, it consists throughout of the most typical, basic, green variety of melaphyr, with frequent highly amygdaloidal layers. The flow-structure is very perfect at many points, and shows that the sheets or flows are still nearly horizontal, with slight undulations to north and south and, perhaps, a very gentle general dip to the southeast. The facts already stated in connection with the mass of conglomerate pro- jecting into the Avestern edge of the plateau indicate that the plateau embraces both of the flows of melaphyr observed in the Crescent Hill section. The lower flow forms the northern bor- der of the plateau, or that portion north of the conglomerate ex- tension : while all the remaining area, or fully three-fourths of the whole, must be referred to the upper melaphyr. The con- glomerate, which normally separates the two flows, has been en- tirely cut out of the section, so far as the surface development is concerned, by the oblique strike ftiults. At the extreme east- ern end of the plateau there are indications, in the form of thin outliers of sandstone, that the upper melapliyr was also for- merly covered by conglomerate. It will appear later that this n conglomerate overlying the upper mclaphyr should not be cor- related with that forming the adjacent ledges of Conglomerate Plateau, but rather with that outcropping on Long Beach Rock and underlying the melaphyr series of Atlantic Hill. Conglomerate Plateau. — This broad mass of conglomerate? the largest in the Nantasket district, is of the normal type, well rounded pebbles of felsite and granite predominating, although pebbles of porphyrite and compact melaphyr are also common. The texture varies from moderately coarse conglomerate to sandstone ; but the bedding can be made out at only a few points as on the north side of Spring Valley Road. The dip seems, however, to be always gentle (5° — 10°) to the south-southeast, diminishing toward the granite. The prominent ledges ex- hibit some well developed joint-planes ; and the plateau is un- doubtedly bounded on all sides, and not alone on the north, by fault-lines. It rises abruptly from the marsh on the west ; Hull Street and Spring Valley Road traverse it in narrow defiles ; while the straight and solid wall of conglomerate 15 to 30 feet high in which it terminates on the south is an exceptionally typ- ical fault scarp. Facts will be presented later tending to show that the downthrow in the case of all of these faults is on the side away from the plateau, and hence that this conglomerate is older than and normally underlies the surrounding formations ; but, as already indicated, the displacement is mticli greater on the north side than in any other direction. Of the thickness of the con- glomerate it is impossible to speak with certainty, since neither the base nor, probably, the original summit of the bed are ex- posed. It can not be less than the height of the plateau, about fifty feet, and may exceed one hundred feet, this being, prob- ably, the thickest of all the Nantasket beds. Round Hill. — This hill, which, as the map indicates, is named for its curved outline, is a solid mass of conglomerate rising very abruptly from the eastern marsh on the north and west to to a height of fiftj feet and sloping gently down to the south- 72 east beneath a bed of melapliyr. The conglomerate of Round Hill is connected with that of the plateau by essentially contin- uous outcrops along the eastern border of the marsh, and must be regarded as a part of the same bed. It is of medium texture above, passing in the lower part of the hill into fine conglom- erate and a considerable thickness (10 to 20 feet) of reddish sand- stone ; while in the ledges rising from the marsh on the southwest we have a coarse and irregular conglomerate with layers of sandstone, which show that in these outlying ledges as well as in the hill the dip is mainly to the east or east-southeast and very gentle — 5° to 10°, but rarely exceeding 5°. The exposed thickness of the conglomerate, including the outlying ledges, is probably between 75 and 100 feet. The low ledge of melaphyr protruding from the marsh on the north side of the hill is prob- ably a part of the bed underlying the mai'sh ; and the fault sep- arating this melaphyr from the conglomerate is thus given a general east-west course. Since both this fault and that bound- ing the marsh on the east cut across the strike of the conglom- erate, the breadth of the outcrop of tliat x'ock, as the map shows, has been g-reatlv diminished at this southeast ansfle of the marsh where the fault-lines meet. According to the view here pre- sented, the melaphyr forming the floor of the marsh belongs to the same bed as that overlying the conglomerate, and hence the conolomerate must also underlie the marsh, but at a sfreater depth than the melaphyr. More probably, however, as the iso- lated ledge of conglomerate in the channel east of Marsh Island suggests, the entire section of Marsh Island and Crescent Hill is represented beneath the marsh, the latter being bounded by faults on all sides and separated from Round Hill by a slip of at least 150 feet. South of the hill, on the bank of the river (Lyford's Liking), the granite is well exposed, with conglomerate and sandstone resting against and upon it. The sandstone is really a fine, red- dish and intensely hard quartzite ; and both it and the conglom- erate appear to fill fissures and depressions in the surface of the granite, in the same manner as the basal conglomerate of the 73 western area to be described later. Tf we make tbis correlation, then, since it is reasonably certain that other beds of" niela|tliyi- as well as conglomerate underlie the coni2;lomerate of Hound Hill, and we can not connect the latter with the basal c! traced about half-way across Granite Plateau, or until it 121 meets the great dike of the second system (4()). Ft is some- what branchinp; ns well as poi-[)hyi-itic, and cross(;s the channel without sensible displacement, although it is somewhat jogged in the plateau. The E.— W. dike (oD) in the granite south of" the west marsh is imperfectly exposed ; and its chief point of interest is that it cuts the great melaphyr dike (2). The dikes beyond the east marsh (40 and 41) have rather noncommittal trends, not belonging distinctly to either system. These are, possibly, extensions of the Great Hill dikes ; but it is impossible to prove it. They converge eastward and upward. The dike in the granite southeast of Round Hill (42), on the other hand, is, so far as can be seen, a very typical example of the first system. The most northerly dike of the second system in this belt is the smallest one, the solitary dike on Crescent Hill (43). It can be traced continuously across the conglomerate and iuelaphyr from the western to the eastern face of the hill. If it were con- tinued across the railroad in the same direct line it would strike Great Hill about 40 feet from the northern end ; but no trace of it could be found in the bare ledo-es of conij-lomerate and sand- stone. Advancing south across Great Hill, we come first to dikes 44 and 45, which, as abeady stated, appear to be cut by 37 of the first system. These two dikes are somewhat con- verging eastward, and 44 ends very abruptly in the conglom- erate before reaching the face of the hill. The principal dike of this entire belt is 46, which begins on the western shore of Granite Plateau with a breadth of about 12 feet. It is jogged to the north 8 or 10 feet in ascending the slope, and cannot be clearly traced more than half-way across the plateau, passing under the grass ; but there is a natural path for it under the narrow meadow along the north side of the high, precipitous ledge of granite and conglomerate to the western marsh. Here it is directly in line with what is clearly the same dike in Great Hill. On the precipitous western face of Great Hill, just north of the junction of the conglomerate and granite, it 122 measures 11.5 feet ; and about half-way across the hill it appears to divide into two nearly equal dikes, which are separately numbered C47 and 48) . Near the point of bifurcation, which is, unfortunately, concealed, a much smaller branch (49) starts from the north side and can be traced across to the western cliff. The similar small dike (50) just south of 46, on the western shore of Granite Plateau is probably also a branch of this main, parent dike. Parallel with 46 on the south is 51, which can be traced for about the same distance across Great Hill, crossing the boundary fault between the conglomerate and the granite without sensible displacement. The most southerly dike on Granite Plateau (52) is, perhaps, a continuation of 51. It is first seen about half-way across the plateau and can be followed to the western shore of Weir River Bay, crossing and probably cuttins: 38 under the water. East- West Dikes South of Weir River Width No. Trend. Hade. in feet. Remarks. 5:^ N. 80° E. N. 5°-10° 1 In iirauite ^waat of railroad. 54 N. 80"^ E. N. 5°-10° 3 " 55 N. 80° E. N. 10°-15° 3.5 " 56 N. 80° E. N. 10° 4 ? " " 57 N. 80° E. east " 58 N. 80° E. N. 30°-40° 10-15? " " 59 S. 80° E. N. 35° 4 west " 60 S. 85° E. N. 10° 7 " " 61 E.-W. S. 15° 2.5 " east " No special or systematic searcli has been made for dikes among the granite ledges south of Weir River Bay ; and the few that were casually observed have been noted chiefly to render the special map more uniform and complete, and not on account of their intrinsic interest. It sln^uld l)e understood, however, that these outlying non-sedimentary areas of the map have not been uniformly treated in this respect, and un- doubtedly many dikes have been overlooked, especially west of 123 the biiy and :ilon<>; Rockland Street and Jerusalem Road. Tlie dikes of this section may be veg-arded as forming* a fourth east- west belt, and it is noteworthy that, for the first time, the older system predominates. The outcrops are so imperfect that no attempt will be made to describe these dikes in detail. North-South Dikes. Width No. Trend. Hade. ill feet. Remarks. 62 N.-S. W. 3'J-5" .75-1.75 Cats 13, 14, and 15. 63 N.-S. Imperfect exposure. 64 N.-S. W. lO'-' 1.5 Irreiiular and interrupted. 65 N.-S. Imperfect exposure. 66 N.-S. Vertical 40 Composite, six parallel dikes. 67 N.-S. Vertical 3 Probably ends against fault on the soutli. 68 N.-S. E.3° 2.-2.25 69 N.-S. Vertical 1-1.5 70 N. 15° E. W. 5^^-10° 2.5 In iiranite south of Weir River Bay. The dikes of the third or newest system, with the normal trend N.-S., have their best development in the central and western ai'eas, only one having been observed in the coastal area and none on Rocky Neck. Unfortunately, the only clear and satisfactory intersections with the east-west dikes are those afforded by the small dike in the Grreen Hill ledge ; but that the dikes of this system are all of about the same age, and newer than the east- west dikes, there can, I think, be no reasonable doubt. Tliey are, as previously explained, darker colored (less chloritic and epidotic) than the older dikes, more prone to weather brownish and to disintegrate on exposure, and more generally characterized b}^ transverse or columnar jointing. The single small example of this system in the coastal area (62) breaks through the Green Hill conglomerate and cuts all the east-west dikes (13, 14, 15) in this ledge, without sensibly displacing them. But it is itself slipped or jogged to the west 4 feet near its intersection with 14 ; and north of 15 it is 124 double, giving off a branch parallel with itself. It runs directly toward the middle of Green Hill Rock, but no trace of it can be found in this bare mass of porphyrite. The fault bounding Melaphyr Plateau on the west is accompanied by several small dikes, one of which (64) can be traced, with some interruptions and dislocations, the entire breadth of the plateau, crossing and probably cutting all the east- west dikes. The most important and decidedly the most interesting of all the dikes of this system is the parallel series (66) separating East Porphyrite Hill and Cliff Plateau. Near the shore these afford the followinaf section from east to west : — Separating melaphyr, 3. feet. " couaiomerate. 5.75. " 50. 74.75 feet. This composite dike or series of dikes is thus equivalent to about 40 feet of diabase and 75 feet of separating conglomerate and melaphyr, or 115 feet for the entire fractured zone. The series is actually exposed for only a few rods ; but there is a clear path for it between Conglomerate Hill and Cliff Plateau to the northern end of the western marsh. Almost in the same direct line at the southern end of the marsh is the largest of all the Nantasket dikes (2), which might be regarded as due to the union and continuation of the composite dike. This view is precluded, however, by the lithological contrast, the southern dike being a true melaphyr and undoubtedly contem- poraneous with the first flow of that rock, while the northern or composite dike, representing the latest period of igneous activity at Nantasket, is very much newer. Probably both the melaphyr dike and the composite dike are intercepted by the boundary fault under the western^marsh. (1) Diabase, 9.75 feet. (2) 4.25 " (3) 7.5 (4) 6.5 (5) 11.25 " (6) .5 foot. 39.75 feet. 125 The three-foot dike (f)7) so clearly exposed on the north shore of Mclaphyr Peninsula and traceable across West Porphyrite Hill certainly does not extend so far in this line as Granite Plateau, probably ending under the water against the N. E.-S. W. fault. Nearly 200 feet west of this line, on the north shore of the plateau, is 68, which can be traced about 150 feet back from the water ; and about 500 feet farther west, following the shore, is 69, of which only about 50 feet in length are exposed. DIKES ON THE COH ASSET AND SCITOATE SPIORE. Dikes are abundant in the granitic rocks south of Nantasket and the boundary faults ; in fact, almost every large ledge or considerable exposure of the granite and diorite shows upon careful examination one or more dikes of diabase. As a rule, however, except along the shore, where the almost continuous ledges are clean and bare, the outcrops of the dikes are obscure andean be correlated only to a limited extent; /. e., the same dike can rarely be traced with certainty from one outcrop to another. And when we further consider that the dikes naturally tend, through their more rapid erosion, to follow the depressions and the drift-covered portions of the surface ; and also that the drift-deposits and marshes are continuous over large areas, especially toward the south and south-east, the futility of attempting to trace out or map the dikes of Cohasset and Scituate becomes apparent. In the more limited Nantasket area the conditions are comparatively favorable for such thorough work ; but it is probable that even here many dikes are wholly concealed, and the map claims to represent with only approximate accuracy and completeness the dikes which are actually exposed. Particular attention was given to the Nan- tasket dikes, because it was seen that the outcrops are sufficient to permit of their correlation with the faults of the district and in systems of diffisrent ages; and about all that it has seemed wise to attempt in the study of the dikes of Cohasset and ^ 126 Scitiiate is to determine to what extent they can be referred to the same systems as the Nantasket dikes. Systematic observa- tions for this purpose were confined to the immediate vicinity of the shore ; and the following paragraphs and accompanying lists give the results of a complete canvass of the littoi-al dikes from Green Hill to the shore beyond the Glades. No attempt has been made to trace any of the dikes inland or beyond the clean exposures afforded by the waves ; but such casual observations as have been made away from the shore indicate that the littoral zone is, in the number and trends of the dikes, representative of the entire area. No dikes have been represented on the general map, partly on account of the inconveniently small scale, but chiefly to avoid giving the impression of great inequality in the distribution of the dikes. It is readily apparent that the dikes of Cohasset and Scitnate exhibit a general agreement in character and trend with those of Nantasket. No dikes of melaphyr have been certainly identified ; but in their place, as previously explained (page 105), there is a well-defined series of porphyrite dikes. These are clearly older than all of the diabase dikes and probably date from the Nantasket flows of porphyrite. Each of the tlu'ee systems of diabase dikes is clearly represented, and the ex- posed intersections show that their relative ages are essentially unchanged. But the oldest system, having the normal trend N. 75° to 80° E., largely predominates, though, as at Nantas- ket, it cannot always be clearly distinguished from the second system, with the normal trend S. 75° to 80° E. Only a few dikes can be referred to the second system. The dikes of the newest or N.-S. system also are relatively less numerous than in the Nantasket area ; and it is especially noteworth)' that they become less numerous eastward, not a single clear example having been observed beyond the Cohasset Rocks or moi-e than three fourths of a mile east of Green Hill J)ea<'h. This apparent limitation of the newer dikes to the Nantasket area and its inunediate vicinity is an inlcrcstiiig and suggestive fact. 127 but a more extended study would be required to discover its full significance. An occasional dike only, such as 92, 109, 114, 136, etc., is distinctly aberrant or ambiguous in trend; and some of these afford intersections determining the relative age and hence the system. Although faults are, perha[)s, as numerous in Cohasset as in Nantasket, none have been definitely located ; and we can only conjecture that their relations to the dikes are unchanged. The dikes are often observed to coincide with prominent joint-planes in the granite ; but as at Nantasket, the joint-structure is evidently mainly of more recent origin than the dikes. The Porphyrite Dikes. 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 N. N. N. N. N. N. N. N. N. N. 5°E. 30° E. 15° E. 5°E. , 18° E. , 50E. , 40° E. -S. -S. 20° E. N.-S. N. N. N. N. N. 20° E. 20° E. 15° E. 15° E. 15° E. N. 10° E. Hade. Width in feet. E. 15°-20° E. 10° Vertical 4 3 18 Vertical 2 E. 5° 8 Vei'tical 13-14 N.W.10°-15° Vertical Vertical Vertical 4-5 4-6 10-15 8-10 Vertical 8-10 Vertical Vertical 1-1.5 2-3 Vertical Vertical Vertical 6 25-30 3-4 E. 10° 8-8.5 Remarks. In granite on Green Hill Beach. 75' E. of 71 ; irregnlar. 15' E. of 72; imperfectlj' ex- posed. 250' E. of 73; near Black Rock House. Near Black Rock House ; cut by 88. Behind Black Rock House; cut by 88 and 89. Branch of 76 ; cut by 88 and 89. About 450' E. of 77 ; cut by 89. About 2550' E. of Forest Ave. ; cut by 150. About 3000' E. of Forest Ave. ; cut by 90 and 91. About 450' E. of 81. 6' east of 82 ; probably con- nected with 81. Near 83 ; cut bv 92 and 93. About 100' E. of 84: cut by 94. About 200' E. of 85; also cut by 94. Wesi} end of Pleasant Beach ; cut by 97 and 98. The dikes of this series have not been observed along the entire shore, but only for about one mile, between Green Hill Beach and Pleasant Beach or Walnut Angle. This is the comparatively straight and eminently rock-bound part of the 128 shore known ns Cohasset Rocks. Although it is convenient to class all the dikes of the above list, provisionally, as porpliy rite, it is well to bear in mind that a few at least are more basic and that the list quite possibly includes some melaphyr dikes. The chief points of interest presented by the porphy- rite dikes have been noted in connection with the Black Rock flow ; while most of the details of any importance, including especially those required for the identification of the individual dikes, are presented in the table. We learn from the table that nine of the seventeen dikes are visibly cut by the oldest system of diabase dikes. On account of the approximate agreement in trend, intersections by the north-south diabase dikes Avere scarcely to be expected ; but, fortunately, one such is presented ; and it is, perhaps, the prettiest and most unique phenomenon to be observed among all the dikes of this district. The relations of the two dikes are shown in Fig. 1(3. The diabase dike (150) advances obliquely from the south until it strikes the east wall of the porphyrite dike (80), follows this wall for 25 feet, amputating a branch of the porphyrite dike, and then passes in a graceful double curve diagonally through the latter and follows the west wall as far as cither can be traced, — 24 feet. The intersecting dike is a typical example of the third system of diabase dikes — black, brownish-weathering, and beautifully cross-jointed. This porphyrite dike (80) is the one aflfbrding the analysis given on page 59. On the north ■rrvTTTTTTUX^ , "/%!/.+ + V* +* +Yt+ + +-^ + »■ *■ + »-* + + + *-- P"iG. i6. — Plan showinc the intersection of a porimiy- RITE dike (So) by a DlABASfC DMvl': (150) ON THE COIIASSET RoCKS. SCALE, I INCH =: 25 FEET. 129 side of Beach Island, near Sandy Beach, is a small exposure of a porphyrite (?) dike — N. 5° E., 4 feet, vertical — which is cut off by a fault. The Diabase Dikes. No attempt is made in the list of E.-W. dikes to distinguish or separate the first and second systems ; but in both this and the N.— S. list the dikes are described in topographic order, beginning in each case at the western end of the Cohasset Rocks, near Green Hill Beach and Forest Avenue. The data of the lists are sufficient for the identification of the individual dikes, if taken in order; but in the brief notes which follow the lists, some additional landmarks and bearings are oiven. East-West Dikes. Width No. Trend. Hade. in feet. Remarks. 88 N. 80° E. S. 5° 3.5 Coarsely porpliyritic; cuts 75, 76, and 77. 89 N. 80° E. Vertical 6.5-7 Exposed aljout 750' ; cuts 76, 77, and 78. 90 N. 72° E. N. 5° 25-30 Branches toward east end : cuts 81; cut by 152-3-4. 90a N. 85° E. N. 5° 2-3 A northern brancl; of 90. 91 N. 85° E. S.5° 1.5-2 A south, brand) of 90 ; cuts 81. 92 E. -W. Vertical 2-3 Exposed over 300'; probably unites with 93. 93 N. 85° E. S. 5° 2-3 8' S. of 92; 92 and 93 cut 84. 94 N. 80° E. Vertical 9 Exposed about 300' ; cuts 85 and 86 ; mai>netic. 95 N. 80° E. S.5° 3-3.5 60' S. of 94; cut by 156. 96 N. 80° E. Vert.-S. 5° 8 75' S. of 95 : parallel with flow- structure of granite. 97 N. 80° E. N. 15°-20° 4-5 125' S. of 96; cuts 87. 98 N. 80° E. Vertical 1-2 50' S. of 97; irreiiuhir; cuts 87. 99 N. 65°-70° E. N. 20° 2-4 E. end Pleasant Beach ; cut by 100; irregular. 100 N. 80° E. Vertical 5 E. end Pleasant Beach ; ex- posed long distance. 101 N. 75° E. N. 5° 1 150' S. of 100. 102 N. 80° E. S. 5°-10° 3-4 About 250' S. of 101. 103 N. 85° E. 25' S. of 102; W. end of Sandy Beach ; irregular. 104 N. 75°-80° E. Vertical 3-4 N. side of Beach I. ; probably cut by 105 and 107. 105 N. 85° E. Vertical 10 N. side of Beach I. ; exposed long distance. OCCAS. PAPEKS B. S. N. H. IV. 9. 130 N. 70°-80° E. N. 85° E. N. 80° E. E.-W. N. 80° E. E.-W. N. 82°-90° E. S. 75° E. E.-W. N. 85° E. N. 80° E. N. 80° E. N. 80° E. N. 80° E. N. 80° E. N. 75° E. S. 75° E. N. 80° E N. 70° E. Irregular. N. 70° E. N. 70° E. N. 70° E. N. 77° E. N. 700-80° E. N, 80° E. N. 80° E. N. 80° E. N. 82° E. N. 80° E. E.-W. N. 70° E. N. 75° E. N. 82° E. N. 80° E. N. 80° E. E.-W. N. 75° E. N. 70° E. Vertical Vertical 4-5 3.5 S. 5° 15 N. 15°-20° Irreg'lar N. 10° 1 Vertical 2.75 Vertical 5-6 N. 15° 5 S. 5° N. 5° N. 10° N. 10° N. 10° N. 5° 2.5 5 2 1 3 1-1.25 N. 5° 1 Vertical Vertical Vertical N. 15° Vertical Vertical Vertical N. 15° N. 10° 8 6-7 6 3.5 10 3-4 9 1.5 6 Vert.-N. 10° 2-5 N. 5° N. 5° Vertical Vert.-N. 3° N. 10° Irregular 2.75 11 3-4 4 .5-.75 2.5 N. 3° 3 N. 10° .5-.75 N. 15° N.5° Vertical 3 2.75-3 9 Vertical Vertical 5.5-6 2 Vertical .5-3 1 70' S. of 105 ; very irregular. 50' S. of 106 and 75' N. of arti- flcial harbor ; faulted. S. side of artiticial harbor: branching. E. side of Beach I. ; compos- ite and branching. Cuts 111 obliquely, both longi- tudinally and vertically. 30' N. of 112; interesting in- tersection by 110. First dike N. of outlet of Little Harbor; branching. First dikeE. of outlet of Little Harbor. 40' N. of 113. Dikes 113 to 118 18'N. of 114. are all on shore 150' N. of 116. E. of the outlet 5' N. of 116. of Little Harbor. 25'N. of 117. 60' N. of 118; 150' from point; exposed 300'. Point nearest Quamino Rock ; exposed 200'. N. W. shore of Sandy Cove. Extreme end of Hominy Point. South side of White Head. ] In coarse gran- 130' S. of 124. ite on E. side 75' S. of 125. [-of Deacon 115' S. of 126. I Bourne's Is- 12' S. of 127. J land. On Hog Rock in Cohasset Harbor On shore W. and E. of Glades Hotel ; coarselv porphyritic. 6' N. of 130 ; near Glades Hotel. About 200' N. of 131 ; in diorite. About 90' N. of 132 ; in diorite. 30' N. of 133 ; in coarse granite. 20' N. of 134 ; in coarse granite. 15' S. of 130; E. of'"Glades Hotel. 30' S. of 131; E. of Glades Hotel. 70' S. of 135; on Strawberry Point. About 100' S. of 138. About 25' S. of 139. About 175' S. of 140; irregular, branching. About 75' S. of 141 ; branching. Headland E. of Strawberry Point. About 300' S. of 143; in old quarry. 131 145 N. 70° E. Vertical 1-1.25 About 100' S. of 144, in quarry ; porpliyritic. 146 N. 75° E. N. 5° 1-1.5 Separated by short beach from 145. 147 N. 75° E. N. 5° 2-3 Near 140. 148 N. 55° E. N. 30° 8 150' S. of 147; coarsely por- phyritic and branching. As a rule, the east-west dikes cut the shore ledges very obliquely ; and in going eastward we cross the first system from north to south, and the second system from south to north. The outcrops begin in the rear of the Black Rock House, where 88 is readily recognized by its coarsely and distinctly porpliyritic character. About 50 feet south of 88 is 89, a splendid dike, which can be traced almost continuously for fully 750 feet. About midway of the outcrop it is some- what irregular and split up ; and the well-preserved glacial striae upon it, and especially the freshness of an inscription bear- ing the date 1852, testify to the stable and resistant character of those diabase dikes which have, by deep-seated alteration, been largely changed to such almost indestructible species as epidote, chlorite, magnetite, and quartz. A little more than half a mile (2860 feet) from Forest Avenue brings us to 90, the largest of all the east-west dikes ; and its outcrop is quite a center of eruption (Fig. 17). It cuts one of the porphyrite dikes (81) and is cut by no less than four of the north-south diabase dikes. The small dikes (90a and 91) are probably branches of 90. The magnetic dike (94) is so highly charged with magnetite as to make the compass observa- tions quite unreliable. It not only cuts 85 and 86, but just before entering the sea is cut by 156 of the third system, which also cuts 95. The next three dikes of this series (96, 97, 98) follow in reo-ular order and brino- us to Walnut Angle or the eastern end of Cohasset Rocks. No dikes have been observed in the ledges on Pleasant Beach ; but near its eastern end we strike 100, which can be traced the entire length of the head- land, trending just north of the extreme north end of Brush 132 Island. In like manner, 103, on the east side of this headland, is exactly in line with a large dike near the southern end of Brush Island. ISi I5i Fig. 17. — Plan showing a group of intersecting dikes on the cohasset rocks- scale, i inch =50 feet. Crossing Sandy Beach, we come at once, on the north side of Beach Island, to 104 and 105, which can be traced for long- distances, the latter enclosing, near the eastern shore, a long narrow mass of granite. The second dike beyond (107) is slipped about twice its width. The large dike south of the artificial harbor divides eastward into two unequal branches separated by from 8 to 10 feet of granite ; and 15 feet north of it is a one-foot branch. The composite dike (109) on the east side of Beach Island is near the contact of the coarse typical granite and the mixed OTanite and diorite. The intersection of 111 by 110 (Fig. 18) is a clear case of the confusion of the first and second systems. The intersection is very oblique, extending over 20 feet along 111. Immediately north of 110 is an unenumerated six-inch dike. The next dike (112) is a fine and instructive example of a branching fissure, and shows flow-structure parallel with the walls. The dike on Hog Rock (129) is exposed, again in the re-entrant angle of the shore south of Strawberry Point. The next dike (130) is beautifully porphyritic, with large clustered 133 Fig. iS. — Plan showing the intersection of two DIKES on the east SIDE OF BeACH IsLAND. Scale, i inch = 8 feet. crystals. It outcrops west of the hotel, east of the hotel, and then, after some faulting and branching, on the east side of Strawberry Point. The remaining dikes on Strawberry Point (131-142) are easily found and identified by taking the ledges in order from west to east, one of them (135) cropping on three distinct ledges. North- South Dikes. 149 149« 150 151 152 153 154 155 156 157 Trend. N. 10° W. ,-S. 50-10° E. .-S. .-S. .-S. ,-S. 8° W. 5°-10° E. 15°-45° E. Hade. Width in feet. Vertical 3-4 Vertical Vertical 2-3 Vertical 3-4 Vertical Vertical Vertical Vertical 3-6 6-7 7.5 3-3.5 Vertical 2 S. E.30°-45o 2-8 Remarks. On Cohasset Rocks, 400' E. of Blaclv Eock House. In tlie valley immediately E. of the Black Rock House ; imperfectly exposed. On Coliasset Rocks ; cuts 80 obliquely. Interrupted and faulted; cuts 90. Cuts 90 ; irregular. Cuts 90 ; unites witli 152. Cuts 90; brandling northward. About 200' E. of 154, on Cohas- set Rocks. On Cohasset Rocks; cuts 94 and 95. On Hominy Point, near White Head. On the short, stony beach immediately east of the Black Rock House there are traces of several — at least three and possibly more — dikes (149a) of this system, recalling the composite dike of the Nantasket area (66). The last of the north-south dikes 134 (157) is the only really doubtful one. This is the dike shown in Fig. 19. In tracing it across the ledge to the mainland, its easterly trend, hade, and thickness are all greatly increased ; and it would, perhaps, be best to class it as a highly aberrant member of the first system. Fig. 19. — Dike 157 cutting a ledge of granite, near White Head, on Hominy Point, Cohasset. SEQUENCE OF EVENTS EECORDED IN THE NANTASKET LEDGES. It is unnecessary to review or summarize farther the evidence supporting the conclusions that have been reached in tlie preced- ing pages concerning the succession of the beds of conglomerate and the flows of lava ; but we may fitly close this section with a general statement of the geological history of Nantasket so far as it is recorded in the hard rocks. The granite, Avitli the associated diorite and felsite, is the fundamental rock in tliis part of the Boston Basin ; and it had, a[)parently, been exposed to erosion for a long time before the basal conglom- erate was deposited over it, The character of this conglomerate 135 is, however, a sufficient indication that outflows of lava and, very probably, the deposition of conglomerate had already begun in the deeper parts of the basin ; and we may fairly suppose that both conglomerates and lavas older than any exposed in southern Nantasket underlie at a great depth the northern part of the peninsula. Eventually the slow subsidence of the land then in progress carried the shore-line over and beyond the Nantasket area, and the so-called basal conglom- erate was formed upon the uneven and fissured surface of the granite. This conglomerate had attained a maximum thickness of perhaps 50 feet, when the first flow of melaphyr was spread over it. This was probably originally an ordinary compact, black lava ; but through subsequent alterations it has become greenish and purplish, compact and jaspery. The lava-flow being submarine, the sedimentary process was uninterrupted ; and the uneven and scoriaceous upper surface of the melaphyr was slowly covered by the second conglomerate, which is largely composed of debris derived from the melaphyr, and like this melaphyr is especially characterized by the segregations of red jasper. When the second conglomerate had attained approximately the same thickness as the basal conglomerate, a flood of more acid lava (porphyrite) was spread over the sea- bottom to a depth of from 50 to possibly more than 100 feet ; and all the porphyrite in the Nantasket area probably belongs to what was once one continuous flow, increasing in thickness eastward, or toward the vent from which it issued. Over the porphyrite was gradually accumulated the third conglom- erate to a thickness of from 50 to 100 feet or more, but termi- nated at last by a comparatively thin flow of highly basic and vesicular basaltic lava, which is recognized now as a typical green and amygdaloidal melaphyr. The fourth conglomerate, with a thickness of from 20 to 30 feet, separates this second melaphyr from the very similar third melaphyr. The latter consists, however, of two flows, having an aggregate thickness of from 40 to 50 feet. Once more the beach conditions prevailed, 136 and the fifth conglomerate was formed. When the volcanic activity was again renewed over this area, the eruptions were, for the first time, partly of an explosive character, forming, as the Atlantic Hill section shows, beds of tuff alternating Avith beds or flows of compact and brecciated melaphyr. This marks the culmination of the volcanic energy, flow succeeding flow, until they attained an aggregate thickness of several hundred, possibly five hundred, feet. This series of eruptions completes the conglomerate and melaphyr series as now developed in the Nantasket ledges ; but on the extremity of Rocky Neck we find evidence that the great melaphyr was finally covered by a sixth bed of conglomerate, and that by a flow of green, amyg- daloidal melaphyr resembling the second and third sheets of that rock. The volcanic energy finally died out ; and these alternating sheets of conglomerate, melaphyr, and porphyrite were probably covered by a great thickness of conglomerate and sandstone without interbedded lavas. But of this upper conglomerate series there are now no visible traces in the Nantasket Penin- sula, although Harding's Ledge, as already explained, aftbrds some evidence that it underlies the middle part of the beach, in the vicinity of Strawberry Hill. As the subsidence progressed and the water became deeper and the shore more remote, the deposition of the coarser fragmental rocks over this area gradually ceased, the conglomerate changing through sandstone to slate, which, we may fairly suppose, underlies almost the entire peninsula north of Atlantic Hill, and has a thickness of several hundred, possibly a thousand, feet. The deposition of the slate probably occupied a much longer time than that of the conglomerate series ; but it was finally terminated by the period of disturbance during which the sediments of the Boston Basin were strongly folded, faulted, and elevated to form dry land. It is to this geological revolution that we owe nearly all those structural complexities which make the stratigraphy of the Boston Basin such a difficult problem. This was also 137 probably a period of intense igneous activity, tlie nunieroiis dikes of diabase traversing the ledges of Nantasket and other parts of the Boston Basin appearing to date from this time. The dikes were probably, in many cases, the feeders or channels of supply of effusive eruptions ; but these surface lavas of every form, not being protected by later sediments, have, at least in in the Nantasket district, been long since completely swept away by the agents of erosion, which daring all subsequent time have worked unceasingly and so efficiently that not only have these ancient volcanoes been destroyed but the very foundations on which they stood. In southern Nantasket the entire thickness of the slate series and probably, on the average, half of the conglomerate series have been removed ; and over broad areas erosion has exposed the original granite floor or cut deeply into it. AGE OF THE NANTASKET ROCK8. Up to the present time no fossils have been found in the Nan- tasket ledges. It is, of course, improbable that any ever will be found in the conglomerate and volcanic series ; but we may reasonably entertain the hope that the slate ledges will yet afford us some clue to their geological age. Fragments of a some- what calcareous slate have been observed in the drift of Straw- berry Hill, and suggest the possibility that beds of impure limestone (of organic origin, if not actually fossiliferous) underlie the northern end of the peninsula. It seems impossible to feel quite sure that the hard massive slate on the railroad, northwest of Rockland Hill, which has been referred provisionally to the bed of tuff outcropping at the base of Atlantic Hill, is not really an isolated exposure of the Cambrian slate of Braintree and Weymouth. The Weymouth beds are now referred by Walcott, provisionally at least, to the Lower Cambrian ; and the Braintree beds to the Middle Cambrian^ The granite and diorite are undoubtedly here, as elsewhere in the Boston ' Tenth Ann. Keport U.S. Geol. Survey, p. 567. 138 Basin, younger than these Cambrian strata ; and it is perfectly clear that the conglomerate and melaphyr series is newer than the granite, and therefore distinctly more recent than the Middle Cambrian. They are quite certainly Paleozoic, but whether as late as the Carboniferous strata in the Narrasfansett Basin is at least doubtful. The conglomerate and melaphyr and newer slates of the Nantasket area may be safely correlated with the same rocks in other parts of the Boston Basin ; and it is hoped that, before the revision of the geolog}'^ of the basin is completed, evidence will be forthcoming which will finally and definitely settle the question as to the age of these strata ; but for the present the problem must be regarded as unsolved. In January of this year (1892), Mr. T. A. Watson found on Pleasant Beach, Cohasset, a smoothly rounded and evi- dently water-worn bowlder, between five and six inches in diameter, of a highly fossiliferous, compact, red (ferruginous) limestone. The fossils appear to belong wholl}^ to two species, which have been identified by Mr. C. D. Walcott^ as the two Lower Cambrian types StraparolUna remota Billings and Hyolithes communis Billings. Although Mr. Watson was unable, after the most thorough search, to find a second speci- men of this limestone on the Cohasset shore, its Cambrian age made it seem highly probable that it had been derived from some point within the Boston Basin, and possibly within the Nantasket and Cohasset district, thus encouraging the hope that it would yet aiFord us the desired clue to the geological age of the Nantasket strata. But this hope has been dispelled ; for Mr. Watson has recentl}^ found several water-worn fragments of precisely the same kind of limestone, holding the same fossils, but not quite so abundantly, on the beach at Bass Point, Nahant. We can no longer doubt that tliis rock is a part of the Cambrian limestone of Nahant, although representing a more ferruginous and more conspicuously fossiliferous bed than any now exposed on that peninsula ; and the Cohasset speci- men must be regarded as a solitary glacial erratic. 1 Proc. Biological Society of Wasliiiigton, VII., 155. THE SURFACE GEOLOGY OF NANTASKET AND COHASSET. As a starting point for the surface geology of this district, including both its glacial and post-glacial history, we must accept the preglacial peneplain, the evidence for which has been presented in sufficient detail in the general description of the topography (pages 5 to 7). The Mesozoic and Cenozoic eras are, apparently, unrecorded here, except in the erosion of the hard rocks. But, while it is to these lono; ireoloo-ical cycles that we must refer the removal of a great thickness of stratified rocks from this area, and the wearing down to the base-level of tlie hard granitic rocks, lavas, and conglomerates, the rugged contours presented today by the preglacial pene- plain undoubtedly date chiefly from the marked elevation of the land which, it is probable, ushered in the great ice-age. On account of the intense hardness of the rocks, these deeply incised lines were only partially effaced by the powerful abra- sive action of the ice-sheet : but, as we have already seen, they are to a large extent obscured or concealed by the non-lithified deposits of glacial and post-glacial origin. The principal events recorded in the surface geology, since the development of the preglacial peneplain, are : (1) A con- siderable elevation of the land at the beoinnins; of the o^lacial epoch. The elevation was, probably, a principal cause of the subsequent glaciation, and is proved by the deeply eroded character of the peneplain. (2) Long-continued glaciation, during which the divided peneplain was strongly eroded and the ground-moraine or bowlder-clay accumulated irregularly upon it, chiefly, at the last, in the form of drumlins. (3) A marked depression of the land, accompanied by the final melting and retreat of the ice-sheet, and the accumulation, either in the sea or in temporary lakes and streams, of the ex- 140 tensive deposits of modified drift. (4) A slight re-elevation of the land immediately after the disappearance of the ice and, probably, during the formation of the modified drift. The elevation is proved in other parts of the Boston Basin by the occurrence of clay beds above sea-level, and generally, it is believed, by cut terraces on the drumlins and marine deltas of sand and gravel. (5) The development of the modern shore, including the growth of the beaches and marshes, accompanied by a slow movement of subsidence. There remains now, in order to complete this paper, simply the task of setting forth more fully the tangible and material facts upon which the foregoing statements rest, taking up the topics, so far as possible, in the chronological order. DEUMLINS, GLACIAL STRIAE, AND BOWLDEES. The unmodified drift or till of this district, so far as it now exists above the sea and is not covered by the modified drift, occurs almost wholly in more or less typical drumlins. The positions, outlines, heights, and names of these are indicated on the general map, so far as they have been determined. Rockland Hill, however, is only an incipient drumlin, and might be more properly classed as a lenticular slope of till. Telegraph Hill, World's End, Planter's and Pine Hills, Tur- key Hill, Bear Plill, and others are examples of composite drumlins ; what were originally separate accumulations of till becoming more or less perfectly united by their continued growth. As the map shows, the longer diameters of the drumlins have without exception a strong easterly trend, varying between south-southeast in the southern and east-southeast in the northern part of the district, and agreeing closely at all points with the glacial striae on the ledges. The following observatious on the directions of the striae em- brace the extreme range and are sufficiently numerous to show 141 the normal courses. It is very obvious tli;it in cro.ssint^ what is now the Nnntasket Peninsula the ice was stroni>lv infiucnccd by the eastward trend of the main valley of the Boston Basin, being held to this course partly, perhaps, by the eastward slo[)e of the ground and the proximity of the sea in that direction, and partly by the continuous barrier of granite 200 feet or more in height presented by the south shore of the harbor. But when the ice had escaped from this influence by flowino- out of the harbor into the bay, or by scaling the ledges to the peneplain of Cohasset, it resumed more nearly its normal south-southeast trend, as shown by the drumlins and the glacial striae along the Cohasset shore. Directions of Glacial Str-iae. Slate ledge south of Thonibusli Hill S. 720-75° E. Northeast base of Atlantic Hill S. 43° E. Double dike on Little Gun Rock S. 37°-47° E. Dike 13 in Green Hill Ledge S. 80°'E . Slate on the railroad south of Nantasket Station. . . S. 3o°-40° E East side of the bay, near the railroad S. 35°-40°E. North base of East Porphyrite Hill S. 47°-50°E. West end of Granite Plateau S. 530-57° E. East side of Rocky Neck S. 43° E. West side of Rocky Neck S. 41° E. South side of Weir River Bay S. 40° E. Jerusalem Road, south of Straits' Pond S. 36° E. " near Green Hill Beach S. 80° E. Cedar Street, Cohasset. S. 30°-35"E. West side of Little Harbor S. 22°-25°E. Cohasset Rocks, east of GreenHill Beach S. 70° E. " " one fourth mile east of beach. . . S. 46° E. " " near Pleasant Beach S. 40° E. Beach Island, near the artificial harbor S. 32°-S5° E. " near Pleasant Beach S. 25°-30°E. Strawberry Point, Scituate shore S. 2.5°E. Scituate shore, south of Strawberry Point S. 25°E. Doubtless the most interesting feature presented by the composition of the till is the occurrence of fossil shells in some of the drumlins. Mr. Warren Upham has noted^ the finding of fragments of shells in the sea-cliffs at several points about Boston Harbor, including the northern member of Tele- ' Proc. B. S. N. H., XXIV., 127-14L 142 graph' Hill and Sagamore Head in the Nantasket area. He has also called attention to the fact that, so long ago as the Revolutionary War, a fort was built on the top of Telegraph Hill, and a well was dug inside the fort of which the commander. Gen. Benjamin Lincoln, wrote as follows^ : "There is a large fort on the E. Hill, in which there is a well sunk 90 feet, which commonly contains 80 odd feet of water. In digging the well, the workmen found many shells, smooth stones and different stratas of sand and clay similar to those on the beach adjoining to the hill. These shells and appearances were discovered from near the top of the ground to the bottom of the well." Of more than thirty species of shells collected by different observers from the till of the Boston Basin, Mr. Upham has, in a hasty examination, found but three — V^enus mercenai'ia (the round clam or quahaug) , Cyclocardia borealis, a.ndClioua sulpJmrea (a boring sponge) — in the section at the north end of Telegraph Hill; and traces of the round clam, which is by far the most abundant species in all the sections, in Sagamore Head. More recently, by careful searching, with the assistance of Mr. H. D. Card, I have found the shells in the eastern scarps of Point Allerton Great Hill, and Strawberry Hill ; and have extended the list of fossils known to occur in the till of the Nantasket Peninsula to eleven species, as follows : — List of Species in the Nantasket Drimilins. Species. Telegraph Point Strawberry Hill. Allerton. Hill. Balanus {sp. f). * * Tritia trivittata, Adams. * Ilyanassa obsoleta, Stimp. * Crucihtilum striatum. Say. * Buccinum undatum, Linnfi. * Mya arenaria, Linng. * Venus mercenaria, Linne. * * * Cyclocardia borealis, Conrad. * * * Astarte undata, Gould. * * * Scapharca transversa, Say. * * Cliona sxdphurea, Verrill. * * * 1 Geographical Gazetteer of the Towns in the Commonwealth of Massachusetts, 1875, p. 56. (Only a small part of this work was published.) 143 I have failed to find even the slightest truce of shells in :iiiv of the druinlins south of Nantasket Beach. This is an instance, however, where negative evidence is of very little value, the apparent presence or absence of shells depending very largely upon the character of the exposure. A deep section is usually essential ; for, as Mr. Upham has explained, the shells are very likely to have been dissolved out of the upper, oxidized or weathered layer of the till. It is probably for this reason that they are not found in the shallow sections afforded by Point AUerton Little Hill, Little Hog Island, White Head, World's End, and Planter's Hill, as well as all the inland drumlins. We may reasonably hope, however, that shells will yet be found in Green Hill and on Bumkin Island. That shells actually occur in the drumlins of Cohasset we have satisfactory evidence in the fact, to which my attention has been called by Mr. H. W. Nichols, that fragments of the round clam were found by Mr. Titus Burbank some years ago in digging a well near the summit of James Hill, northeast of Scituate Pond. The well is 45 feet deep : and the shells were observed only near the bottom. They cannot, therefore, be referi'ed to any post- glacial source. Although shells are now so generally wanting in the buff or oxidized till, evidence of their former existence is afforded by the calcareous material which may, in certain cases, be detected in the till. This evidence is especially clear in the buff till of Telegraph Hill. The carbonate of lime has, in part at least, been dissolved and segregated, locally but firmly cementing the till, and forming in the finer parts regularly rounded concre- tions, from a fraction of an inch to several inches in diameter, which are sometimes attached to the flat surfaces of the large stones or bowlders. These nodules effervesce freely with acid, and no other source of the carbonate than the shells is apparent. Mr. Upham has correctly explained the shells in the till, which are always in a fragmentary form, as having inhabited the bottom of Boston Harbor before the coming of the ice-sheet, 144 by which they were gathered up and incorporated in the till ; and he has shown that some of the species, at least, like the round clam, are now found only in more southern waters, indicating that the sea in this vicinity just before the glacial epoch was warmer than at the present time. The shells are also of special interest as proving not only that Boston Harbor was in existence in preglacial times, for otherwise we could not know but that this basin was above sea-level before as well as during lis occupation by the ice-sheet ; but they also show that the harbor has now approximately its preglacial outlines and hence that the sea has regained, on this coast, very nearly its preglacial level. The peneplain, as already stated, and especially when we make an allowance for the glacial erosion upon its surface, proves that this coast is now one hundred feet or more above its preglacial level. But this greater elevation is off-set, so far as the harbor is concerned, by the excessive erosion which, being a valley in comparatively soft rocks, it suffered in glacial and especially in immediately preglacial times. Although the fragments of shells are unquestionably the most interesting feature of the till observed in Nantasket and Cohasset, other components of more normal character — the ordinary stones and bowlders — demand brief mention. No particularly striking instances of glacial transportation have been noted. The till of the Nantasket Peninsula, as may be most readily seen in the marine sections of the drumlins, consists very largely of fragments of slate, which is undoubtedly the under- lying rock not only of this })eninsula but of the main part of the harbor to the northward. Bowlders of conglomerate are not wholly wanting nortli of Atlantic Hill, but they are so few and small as to prove tiiat it can occur only very s[)aringly, if at all, in the harbor area or beneath the extensive drift-deposits north of the harbor. In fact, pretty much all of the congloni- ei-ate in the till north of the Nantasket ledges might be referred with considci-able probability to the known exposures of that M5 rock in Mcdford, on the extreme northern niaroston Basin, or at least to the probabk; eastward extension of those beds beneatli the drift ; although the strong easterly trend of the glacial movement in the harbor allows us to regard the ])roniinent conglomerate ledges in the central part of the basin as a possible source of some of the conglomerate erratics in the Nantasket Peninsula and Cohasset. In the Nantasket drunilins, granitic rocks, felsites, quartzite, etc., froiu the high land north of the Boston Basin, form a small but very obvious or noticeable fraction of the till, and in some cases exhibit lith- ological peculiarities enabling us to refer them to the parent ledges, as in the case of the bright red felsite from Saugus Centre. A portion of what miglit be mistaken for conglom- erate is brecciated felsite, and another portion is brecciated or pebbly slate. Over the granitic area of Cohasset and Scituate, as a matter of course, the crystalline rocks predominate in the till, and the slate, which does not bear glacial transportation well, is subordinate, except, perhaps, in the finest part of the till. There is, however, among the larger masses or fragments in the till, a quite liberal sprinkling of both melaphyr and con- glomerate from the Nantasket ledges. These are especially noticeable along the shore, the number and average size, with occasional exceptions, diminishing rapidly as the distance increases. The surface distribution of the larger bowlders indicates that they are not equally abundant in all parts of the till ; for while the majority of the drumlins are remarkably smooth, pre- senting on the sui^face but few bowlders larger than those to be seen in the stone walls, others, like Booth Hill in Scituate, are, at least on certain slopes, thickly strewn with masses from three to eight feet or more in diameter. The bowlders are, in some cases at least, clearly most abundant along those lines where the till has been eroded by standing or running water and which may thus be regarded as in some sense ancient shores or OCCAS PAl'EKS B. S. N. H. IV. 10. 146 channels, resembling the modern shore in this respect ; or over areas where there is at least evidence that the till has been sub- jected to the sorting action of water and the finer materials removed. It is reasonable to suppose that the bowlders are most abundant in the lower parts of the drumlins, and in gen- eral where the till is thinnest, as may be observed at many points over the rocky peneplain of Cohasset ; and it is certainly not improbable that, as suggested to me by Mr. Upham, this is owing, not wholly to the fact that the source of the bowlders is near at hand, but partly to an actual combing of bowlders out of the till or ground moraine as it was swept past the ledges by the movement of the ice, the ledges thus serving ;is gathering points for bowlders. An occasional bowlder merits special attention on account of its size ; and some of these are rather striking also in their sit- uations. As might be expected, the largest erratics consist almost exclusively of granite, the most massive and resistant rock of the region. One of the most impressive examples observed in the Nantasket-Scituate area is the block of granite on Booth Hill, in Scituate, known as Hatch's Rock (Fig. 23). It is very regular in form, approximately twenty feet square with an average height above the ground of at least ten feet. It stands on the summit of the hill and some fifteen feet above a broad level terrace or platform. Near the eastern base of Booth Hill, on the grounds of Mr. Silas Peirce, is Toad Rock, a block of diorite about twenty feet long resting on a local deposit of modified drift. A granite bowlder about twenty feet long, twelve feet wide, and ten feet high lies in the woods close on the south side of Beechwood Street, about one mile northwest of Beech wood Village, with several others nearly as large near it ; and east of the Village a block of granite which is approximately a ten- foot cube lies on the west side of Bound Brook, a few rods south of the street ; while in the rocky woods south of Scituate Hill may be found a striking example of a large bowlder which has been disrupted by the action of the frost. 147 In a walk along the shores ofNantasket, Cohasset, and Scit- iiate, one observes occasional large bowlders, sometimes local and sometimes traveled or truly erratic, but the only one that need be specially mentioned is an angular block of coarse, massive, pinkish granite, one among a large ninnl)er of smaller masses, on the eastern shore of the Glades, in Scituate, and nearly opposite the Osher Rocks. It is about twenty-five feet long ; but being of precisely the same lithological character as the neighboring ledges, it would, perhaps, escape particular notice, except for its position, being supported by smaller bowlders and the ledge in such a manner that when the tide is out one can pass under it from one side to the other. The fact that it maintains this insecure position, although exposed to the full force of the breakers, testifies to its great weight. The best example, however, of a perched block, and one of the largest bowlders in Cohasset, is the mass of granite on the estate of Mr. Edward Wheelwright west of Little Harbor, which has long been known as Tittling Rock (Fig. 20). Going north from Cohasset Village on Jerusalem Road, we come in about half a mile to the residence of Mr. Wheel- wright, whence a private road runs nearly due west for a considerable distance through beautiful rocky woodland, bringing us, when about two thirds of a mile from the highway, to Tittlino- Rock. The bowlder, which lies close to the road on the south, is a rude and somewhat oblique parallelopipedon measuring approximately 20 feet in extreme length (north- south), 12 feet in breadth, and 12 feet in greatest height. Its location is almost the highest point in this part of the peneplain, probably 100 feet above the sea. It rests upon a low, glaciated ledge of granite sloping gently to the north, seeming to lie in a shallow depression or glacial trough ; and the actual base or supporting surface is not more than six feet long and from two to three feet wide. But although seemingly so nicely poised, considerable force would probably be required to disturb it ; and it is not properly a rocking stone. 148 Fig. 20 — TiTTLixG Rock, Cohasset. On the northwest slope of Telegraph Hill, just above the junction of the main hill with the lower, flat-topped part, is a bloclv of coarse granite about fifteen feet in diameter and from six to seven feet hioh above the ground. A verv similar block rests in a similar position on the north side of Strawberry Hill. It is about fifteen feet square and eight if not ten feet thick. Two bowlders of granite, essentially similar to these, are similarly situated on the northern slope of Otis Hill, in Hingham. The occurrence of these isolated blocks on the northern aspects of the hills, and at approximately the height of one of the principal sand plains, is certainly suggestive of their transportation by floating ice rather than the ice-sheet ; and it appears necessary to reg^ard them as havino- been derived from the o-ranite ledo-es north of the Boston Basin, the nearest possible source of granite of this chnracter, in the direction of glacial movement, being- twelve to fifteen miles distant. GLACIAL POTHOLES. Four years ago, my attention was first called by Mr. T. T. Bouve to the existence of a fine series of glacial potholes on Cooper's Island in Little Harbor, Cohasset, and I had the 149 pleasure a little later of examining the locality in company with him and Mr. Warren Uphani. In the followinf year Mr. Bouve published a full account of the potholes/ from which the greater part of the following description is taken, together with the illustrations (Figs. 21 and 22). V^^^^ J"' ,„7>' Fig. 21. — Pothole, Cooper's Island. Cooper's Island, so-called, is the peninsula consisting of Sfranite ledsres and salt-marsh which extends into Little Harbor from the west side. The eastern end, which reaches nearly to the middle of the harbor, is an approximately north-south rid^e of o-ranite one fourth of a mile lona^ and from 15 to 25 feet high. This ridge is divided transversely near the southern extremity by a short stretch of grass land ; and it is near the 1 Proc. B. S. ISr. H., XXIV, 219-228. 150 southern end of the main ridge, on the east side and quite close to the water that the potholes occur. The most southerly and most perfect holes are on the north side of a slight indentation of the rocky shore and may be readily found when the tide is out (Fig. 21). Of the lowest and best preserved of these (No. 1) the bottom is still intact and perfect, holding water to a depth of 21 inches and having a well-defined rim just at the surface of the water. The diameter at the rim is 25^ inches ; below the rim, 30 inches. Above this rim the whole southern side of the hole is wanting ; but on the northern side the rock is smoothly concave and the characteristics of the pothole are plainly discernible for a height of four feet, with a breadth of from three to four feet, making the total depth about six feet. The whole has the appearance of a large, imperfect or one-sided pothole with a smaller and perfect pothole in the bottom of it. Exterior to this pothole, the tide sinks below the level of its bottom, but at high tide all is covered. The second pothole has its bottom three and one half feet above that of No. 1 ; and their centers are three feet apart hori- zontally. The whole southern side of this hole (No. 2) is wanting and water can now stand in it only to the depth of about two inches. Above the bottom, the granite is smoothly concaved for a breadth of three feet and a height of five feet ; and this hole, when entire, was evidently about as large as the first. The slope of what remains of the walls of these holes shows, apparently, that the flow of water over the rocks was from the west and northwest. Of the third pothole there is but little to be said, except that it is small and shallow. It is about five feet above the bottom of No. 2, still in a northwest direction ; and there may be traced from it in the same general direction a narrow water-worn channel about six feet in length. The hio-h-tide mark is about midway between the second and third potholes. It is quite obvious thtit these potholes were once entire. Their sharply-defined vertical edges are seen to coincide with 151 one well-marked joint-plane sloping steeply to the south ; and the southern sides of all three of the holes have, apparently, been carried away, by the removal of a single large joint-block of the granite. Since this block appears to have been torn away bodily, leaving an angular, unglaciated surface, and no trace of the block itself can be seen in the vicinity, we may fairly regard its removal as the last work accomplished by the ice-sheet on the lee side of this ledffe. Fig. 22 — Pothole, Cooper's Island. The fourth pothole is or was the largest of all, and hence has been commonly called the "Well." Passing over the rocky elevation in a northerly direction, it may be found about a 152 hundred feet distant from the others ; and it is a little farther than the others from the shore, its bottom being about four feet above high tide (Fig. 22). The bottom of this hole, which is still entire, so as to hold water to a depth of about one foot, is cut in the solid granite at the base of a small scarp. It is somewhat oval in form, the horizontal diameters varying from two feet ten inches to four feet. The ledge rises abruptly nine feet from the margin of the Well and ten feet from its bottom ; and the Well itself was probably as deep at least as ten feet, the curvature and wearing of the rock clearly showing this. The beautifully rounded and water- worn form of the granite on the southeast side of the Well, above the water, is a plain indication, however, that it has not been entire on this side since the water ceased to swirl through it. On the north side the concave wall is vertical or slightly overhanging ; but on the west side the wall is widely flaring, receding upwards in such a way as to indicate that the water entered at this point, the Well agreeing in this respect with the group of potholes ( 1 to 3 j . It seems probable that the pothole was entire in its early stages ; but as it increased in size the outer or lee wall was either worn through or carried away bodily by the ice, after which the rush of water continued long enough to develop the graceful contours of its natural exit, as shown in the figure. Besides these four potholes, there are other depressions which are evidently incipient or embryo potholes. A linear group of these may be observed about twenty feet north of, and parallel with, the southern series (1 to 3). The upper one is shallow, like the bowl of a spoon, about a foot across, showing, extending from it, a water- worn channel sloping easterly about ten feet to the brink of the ledge, connecting the first depression in its course, with two others of similar form ; while on a lower surface there is a larger depression just where the water from the first might descend. Considering the shallowness of the entire portions of all the 153 potholes, and tlie probabilitv ot" tlieii- visitation hv fenerations of both the Indian and the white man, it is not snr|)risino- that nothing of their contents is left in or al)out them. There is, however, one rounded stone in the possession of Mr. Charles S. Bates, the owner of Cooper's Island, which tradition states to have been taken from the deepest pothole (No. 1). It is a smoothly rounded and nearly spherical ball of granite about four inches in diameter, — a typical pothole bowlder ; and there seems to be no reason to question the truth of the tradition. On one of the higher ledges south of Little Harbor and Beach Street, and near the village, there is a smooth hollow in the granite, about a foot across and nine inches deep, which, from its form, has been called the "Devil's Armchair." Although not well-defined, it is clearly a small pothole ; and in this instance, as in others, an apparently water-worn channel extends east -southeast on the rather abrupt slope of the ledge for several feet, ending in another and smaller hole. Durino- the past year I have discovered still another pothole. This is on Cohasset Rocks, immediately behind the Black Rock House, on a bare surface of granite sloping down into the sea, and only two or three feet above the high-tide level. It is a smoothly- worn and sharply-defined basin, somewhat pear- shaped in outline, and measuring 41 by 33 inches in maximum length and breadth, the major axis trending about S. 15° W. It holds about six inches of water, but the depth below the well-defined rim varies from (3 to 18 inches, being shallowest on the side toward the sea. The potholes of Cooper's Island, although on the shore, do not directly face the sea, and the Well, especially, is quite shut in by the granite ledge on the seaward side, as the cut shows. They are also partly above the high-tide level ; and the lower ones are bathed only by the quiet waters of Little Harbor. If the sea were high enough to sweep over Beach Island and the other barriers between Little Harbor and the Atlantic, the pot- holes would be completely submerged and so still beyond the 154 reach of the surf. In short, both the situation and the forms of the potholes make it impossible to regard them as in any sense the product of marine erosion. It is equally clear that no ordinary or surface river ever flowed over these ledges. We are forced, therefore, to ascribe their origin to the action of glacial streams. But here the question arises as to whether they have probably been formed by subglaciai rivers rushing along over the ledges, or by streams which, flowing over the surffice of the ice-sheet, plunge through crevasses to the solid rocks below. Such a waterfall in the ice is called a moulin or glacial mill ; and since the crevasses must often be of great depth — hundreds or even thousands of feet — these o-lacial mills are generally recognized as the most efficient of all agencies in the formation of potholes. Subglaciai streams, it is believed, must usually, like ordinary rivers, follow the depressions of the rocky surface ; while moulins — the chief source of the subglaciai stream — may strike the ledges with resistless force at almost any point, and especially on the crests and southeastern or lee slopes of prominent ledges and ridges, where the conditions are most favorable for the formation of profound crevasses. A strong presumption is thus created in favor of the glacial mills as a cause of the potholes of Cohasset. One difficulty, however, still remains. As the ice-sheet moves continuously forward, carrying the crevasses and moulins with it, how is it that the potholes escape elongation in the direction of the movement? Of course a mo2ilt7i cannot move forward indefinitely, but only until a new crevasse is formed behind the one through which the water first fell, a few feet or yards at the most ; and the moulin then returns to its starting point. But why do the potholes not show even this small amount of elonga- tion ? Mr. Bouve has discussed this point in the following words^ : " It has, indeed, been thought strange that, as the ice moved continuously on, the holes were not found generally elongated 1 Proc. B. S. N. H., XXIV, 224. 155 in the direction of the movement of the ghicier rather than circular. Such thought, however, is only consistent with the presumption that the holes were made just where the water first fell upon the rock surface below. Far more reasonable is it to suppose that the holes were formed somewhat distant from this place, where the masses of rocks borne by the waters found a lodging in some depi'ession and there by rotation worked out the potholes. The ice might move on and the waters descend through the moulin far from where they first fell, yet continue their flow in the same direction as at first and go on with the work of rotating the contents of the hole through a whole season. In such case there could be, of course, no reason to expect the elongation." In explanation of potholes in close proximity to others and yet seemingly independent, as in the case of the Well, Mr. Bouve says : " Observation upon Alpine glaciers shows that as a crevasse is carried forward by the general movement of the ice it closes. Subsequently a new one is formed just where in relation to the land at the margin of the glacier the former one existed ; and the waters again descend upon the rock surface near where they before fell, but not often, probably, in exactly the same place ; and thus pot- holes are formed contio;uous to each other and vet far enouo-h distant to make it evident that they were not ])roduced by the same flow of water." This is undoubtedly a true explanation of many glacial potholes ; and its weakest point, as applied to most of the potholes of Cohasset is that it requires the water to flow up over instead of around the prominent ledges. Mr. Upham has suggested a diflTerent explanation. He says^ : " The time of the excavation of these glacial potholes was probably the early part of the epoch of glaciation, when the ice-sheet was being formed upon the land by snow-fall. Upon any hilly country the ice must have attained an average depth somewhat exceeding the altitude of the hills above the adjoin- ing lowlands before any general motion of the ice-sheet could [iProc. B. S. N. H., XXIV, 226-228, 166 begin. During this process of slow accumulation of the ice- sheet, the summer melting upon its surface would produce multitudes of rills, rivulets, and brooks, which might unite into a large stream, and this, pouring through a crevasse and melting out a cylindric moitlin, might fall a considerable depth to the bed rock, perhaps one or two hundred feet or more upon an area so moderately uneven as Cohasset, while yet the ice- motion, though sufficient to permit the formation of the crevasse, might not have gained a definite current to carry the crevasse, nioulin, and water-fall away from the spot where they were first formed. We may thus explain the continuation of a glacial water-fall in one place while it was excavating one of these " Giant's Kettles " or potholes. After the ice-sheet acquired a current because of the greater thickness and pressure of its mass, such deep cylindric excavations in the bed-rock could not be made ; and during the recession and final dissolution of the ice- sheet, it seems probable that its receding border had steeper gradients and consequently even more rapid motion than in the culmination of the glacial epoch." In criticism of this view it may be pointed out, first, that while the movement of the ice-sheet was still practically nil, it would probably have accommodated itself to the irregularities of its rocky floor without cracking ; second, that the face of the country, before it had been swept by the long-continued advance of the ice-sheet, was probably buried beneath a considerable thickness of soil and half-decomposed rock, the product of chemical and mechanical decay during long preglacial ages ; and third, that it makes no adequate allowance for glacial erosion of the hard rocks during the long periods of the maximum development and waning of the ice-sheet, which would inevitably have obliterated the potholes. In short, Mr. Upham's view carries as its logical consequence the reduction of glacial erosion not only to a minimum, but almost to nothing. While seeking for additional light upon this problem it has occurred to me, first, that a moulin may remain approximately nr 157 stationary, while the ice moves on, through the bafkvvard ero- sion and melting of" its up-stream side; and, second, that when a pothole is formed at the bottom of a rnonlin it is not the direct impact of the water upon the face of the ledge that does the work, nor do the stones carried down by tlie water wear the ledges appreciably by their direct fall, but the pothole is due to their subsequent movement and especially their rotation by the water. This rotation implies an antecedent depression or hollow to hold the stones, and thus the conditions are seen to be essentially the same as for ordinary river potholes. Since the rotation of stones in a pre-existing hollow is an essential condition of the glacial as of other potholes, and the moulin simply supplies the power, it would seem to make little or no difference whether the water plunges into the up-stream side, the middle, or the down-stream side of the hollow. The pot- hole will be made where and only where the hollow is ; and during the progress of a ojioidioi across the hollow there wOuld not, apparently, be any mai-ked tendency to elongate it. In the case of a linear group of potholes on the lee slope of a ledge, it is reasonable to suppose that the upper one, which, on Cooper's Island, is always the smallest and most indefinite, marks the shifting position of the moulin, and that the others were formed by the subglacial flow of water from the bottom of the moulin. Reverting once more to the question as to what phase of the continental glaciation — the beginning, the maximum, or the waning — was most favorable for the formation and preservation of glacial potholes, it may be noted : First, that during the de- velopment of the ice-sheet precipitation was mainly in the form of snow, and greatly exceeded the Avaste by melting ; the true glacier ice was covered by a great thickness of neve and snow ; the water escaped largely by seepage ; and the superficial streams were small. Second, that while the growing sheet of ice, neve, and snow must have been virtually stationary for a long time, it must also have attained quite a high velocity (for an ice-sheet) 158 before crevasses extending up through the thick layers of yielding neve and snow to the surface could be formed ; and the latter condition must, of course, be realized before glacial moulins are a possibility. Tiiird, that it is in tlie highest degree improbable that during the period of maximum glaciation such comparatively slight inequalities as the Cohasset ledges could have produced crevasses traversing the entire thickness of the ice- sheet ; just as in the case of inequalities in the bed of a river, the surface disturbance diminishes with the depth of the water. Fourth, that since the decay and final disappearance of the ice-sheet were due to a general amelioration of the climate, resulting in part it is })robable from a depression of the land, it was probably accomplished by such a general ablation of its upper surface as would have caused only a slight increase in the gradient, an increase that may have been neutralized, so far as its effect upon the rate of movement of the ice was con- cerned, by the greater depression of the land toward the north, and the steadily diminishing thickness, weight, and power of the ice. It certainly seems more reasonable to suppose that the rate of movement of the ice-sheet diminished as its weight and power declined, than that it continued to increase until the final disappearance of the ice. Fifth, that when the decay of the ice-sheet over any area was well advanced, it must have presented, at least during a considerable part of the year, a sur- face of hard, brittle ice ; the conditions being thus more favor- able than at any previous period for the existence of superglacial streams and of crevasses traversing the entire thickness of the ice ; and it is obvious that such crevasses are now consistent with a very slow movement of the ice over very slight inequal- ities of the ground, for the ice has lost, in its loss of weight, not only the power to flow rapidly forward, but also the power to flow around the ledges and mold itself accurately against them without rupture of its mass. Sixth, that with the dimin- ishino- thickness of the ice-sheet there must have come a time when it could no longer overcome the friction of so rouo^h a 159 surface as the Boston Basin presents, and it became essentially stationary, or was subject to slight local movements only ; and presenting still a fairly continuous sheet of solid ice, with surface streams and crevasses, the conditions were especially favorable for the formation of normal glaci-al potholes by stationary moulins. Seventh, the contours and distribution of the modified drift in this region indicate that the ice-sheet became, later, very ragged, being frayed out at the margin and divided into numerous detached masses, so that important surface streams and ynoulins ceased to exist. In conclusion, attention may be called once more to the fact that the glacial potholes prove by their very existence that there has been no appreciable glaciation of the rocks in which they occur since they were formed. In other words, they are a record of a time when the ice-sheet still covered the land, but had become nearly or quite stationary. It may be remarked in this connection, also, that while the till in this region is found to lie almost universally upon strongly glaciated rock surfaces, the deposits of modified drift repose in part upon glaciated, but very largely upon unglaciated and distinctly water-worn sui-faces. It is a common circumstance also to find the protruding and stoss portions of ledges from which the sand and gravel have been removed more or less distinctly glaciated, while all the depi'essed and lee surfaces are beautifully sculptui'ed by water, showing many incipient pot- holes connected with winding channels and separated by smoothly hummocky and undulating surfaces. The glacial striae often end abruptly at the margins of these depressed areas, showing very distinctly the co-operation or simultaneous action of water and ice at a time when the ice had but little erosive power, and proving very clearly, also, that the water was subglacial. MODIFIED DRIFT AND TERRACES. It is generally conceded that the recession and final dis- appearance of the ice-sheet was attended by, and probably 1(30 consequent upon, a marked depression of the land. This depression was, apparently, more than sufficient to restore the preglacial rehations of the hmd and sea ; or at least evidence is not wanting that the shore-line was, for a brief [)eriod, somewhat above its present level. The re-elevation of the land here referred to amounted to 520 feet in the vicinity of Montreal, 225 feet on the coast of Maine, but only to 20 or possibly 30 feet in the Boston Basin ; and the evidences for both the eleva- tion and the slow subsidence which has since been in progress are much less perfectly developed in the Nantasket area than in other parts of the Basin. Partly by subglacial and superglacial streams, but mainly by the great torrents and tiie temporary lakes resulting from the final melting of the ice-sheet, the till or bowlder clay was very largely modified, — that is, washed, assorted, and stratified in the sand plains or deltas, gravel ridges or eskers and kames, and clay beds. The explanation of modified drift in general requires a constant blending or intermingling of glacial, fluvial, and lacustrine conditions; and probably, also, in the Boston Basin, for the clay beds if not the lower deposits of sand, we may pos- tulate marine conditions. Among the various forms of modified drift the eskers or gravel ridges appear usually to be the oldest ; and being the product, chiefly, of superglacial and subglacial streams, they evidently record the actual presence of the ice-sheet as a fairly continuous body of ice and not as the detached masses of ice indicated by the kettles and other features of the sand plains. If anv important or esker-formino- o-lacial streams crossed the Nantasket and Cohasset area, their accumulations have been effaced by subsequent erosion or deposition ; for no distinct eskers have been observed. The sand plains, also, are rather scantily and imperfectly developed. Sand plains, in general, we must regard in part as repre- senting the flood-plains of great rivers flowing from the melting ice-sheet and in part as deltas formed by these rivers in temporary 161 lakes or, possibly, in the case of the lowest plains, in the sea. Most of these lakes were iindoubtedly truly glacial, — that is, enclosed partially, sometimes wholly, by solid walls of ice. The kettles, dimpling the sand plains, it is well understood, are due to the subsequent melting of masses of ice which were buried in the stratified deposits ; and the steep and sharply- defined marginal slopes which the sand plains often present are doubtless in part the natural, free, growing edges or fronts of the deltas, and in part due to the subsequent melting of walls of ice against which either flood-plain or delta deposits had been accumulated, the sand and gravel, as it was thus grad- ually let down, naturally assuming the maximum declivity. The Nantasket Peninsula, being then surrounded by deeper water than at present and even more remote than now from the main land and the mouths of rivers, was almost exempt from these delta and flood-plain deposits. The higher plains are entirely wanting here, and the lowest, which varies usually from fifteen to twenty-five or thirty feet in height, is developed in only a very scanty and fragmentary manner, as on the south side of Point Allerton. In Cohasset and Scituate the plain, having usually a height of from forty to sixty feet, is well devel- oped at intervals. It forms the comparatively level land along the railroad between Hingham and Cohasset Village ; and in the northern part of the village, and especially between Little Harbor and the railroad, it is a very typical plain, this part of the village being designated as the "Plain" on the published maps of the town. It rises quite abruptly 45 feet from the marshes of Little Harbor and, deeply dimpled with beautiful kettles, slopes gently down on the south to the marshes and James River. One of the kettles, on the north side of Beach Street near the Plain, has long been known as the "Punch Bowl"; and several of them hold permanent ponds, the pond near the south end of the Common being an example. When this plain was formed, the ice must have completely filled the basin of Little Harbor ; and the stream supplying the detritus probably OCCAS. FAPEKS, B. S. N. H. IV. 11. 162 came from the west or northwest. It is interesting to observe that this deposit of sand is probably the only barrier between Little Harbor and Cohasset Harbor at this point. The considerable deposits of modified drift in the Beechwood district, south and southeast of Scituate Pond, may be regarded as a fragmentary development of this plain. A much lower plain, which, quite probably, is of marine origin, has a very perfect development in the district south of the Glades. It terminates on the east in Mitchell's Beach and presents a very flat, unbroken surface, from 10 to 15 feet above the sea, over an area nearly half a mile square. It is also scantily developed along the northern edge of the peneplain, in the vicinity of Rockland Street, and at other points ; and we may reasonably suppose that farther inland it underlies the broad, level, and swampy valley of Bound Brook and its tributaries. Beds of clay, or the finer portion of the modified drift, appear to be wholly wanting above sea-level in the Nantasket and Co- hasset area. There can be but little doubt, however, that they are extensively developed under the salt marshes and in the drowned or marine portions of the valleys generally. The slow development of the salt marshes during post-glacial time probably commenced in every important instance with a broad, level expanse of the finest glacial silt or clay covered by a mod- erate depth of water and supporting a luxurious growth of eel grass. In the meshes of this dense jungle or mat of vegetation the detritus brought by the tides is entrapped and slowly sinks to the bottom. The upper surface of the deposit is thus slowly raised to the low-tide level, when the conditions become unfa- vorable for the growth of eel grass and the area enters upon the mud-flat stage. Although the main sweep of the flowing and ebbing tides is now through more or less definite channels or creeks, but little additional silt lodges on the unprotected flat until it has been overrun from the shore outward by higher forms of vegetation, when the silting-up process proceeds un- checked until the high-tide level is reached. A dense, firm 163 mat of peat-forming vegetation then covers the surface, and the marsh is complete. In other parts of the Boston Basin, how- ever, and generally wherever excavations have been made in the salt marshes, the evidence is conclusive that a slow sub- sidence has permitted the formation of several successive beds of peat, separated by beds of clay. Although, as the beaches, marshes, and marine cliffs so plainly indicate, the sea has undoubtedly maintained its present level for a very long time, the rocks of our coast are so intensely hard and resistant that there are very few points where they show any appreciable amount of marine erosion. All along the Nantasket and Cohasset shore it is perfectly obvious that, save where the rocks are very finely jointed, or a dike has yielded to the ceaseless pounding of the waves, the ledges are still essentially intact, showing still the roche moutonnee forms impressed upon them by the ice-sheet, even the glacial striae being, in some cases, well preserved for several feet below the high-tide level. Above the present level of the shore evidences of marine erosion on the hard rocks are, so far as I have observed, wholly wanting. If the sea has stood at higher levels in post-glacial times, the evidence must be sought in the erosion, not of the hard rocks — granite, felsite, melaphyr, etc., but of the non-lithified or drift deposits. Fortunately, the drumlins, which are such a prominent feature of Boston Harbor, present in their firm but yielding material and regular outlines condi- tions exceedingly favorable for making and preserving a record of even a very brief occupation of a higher level by the sea. Any one who notes the extensive erosion of the drumlins by the sea at its present level and the comparative stability of the erosion scarps, cannot doubt that if similar features — terraces and scarps — had ever been developed on the drumlins at higher levels, some indications of them would still be traceable. A general study of the drumlins of the Boston Basin has satisfied me that undoubted horizontal erosion-marks are a common 164 feature up to a height of 100 feet or more above the sea. They exist as strongly-marked and approximately horizontal and longitudinal scarps and terraces, frequently bearing bowlder pavements and showing a general correspondence in height with the sand plains of the region. They have been observed in all parts of the Boston Basin ; but ai'e, perhaps, most favorably exposed for study on several of the drumlins northeast of Boston, including Breed's Hill, Pleasant Hill, Mt. Revere, Mt. Washington, and Powder-Horn Hill. They are also a promi- nent feature of several drumlins on the South Shore, including Baker's and Otis Hills, in Hingham, and have been observed on some of the drumlins of the Nantasket-Cohasset area. These may be briefly described, commencing on the north. On the southwest side of Telegraph Hill, in Hull, a sloping, bowlder-strewn terrace, from 20 to 30 feet above the sea and backed by a steep declivity, extends directly across from shore to shore. The only terraces on Point Allerton Great Hill are such as have probably been formed by the gradual accumulation of soil by rain-wash against stone walls at the lower edges of culti- vated fields. Three such artificial terraces may be readily traced along the southern slope at heights of from 55 to 75 feet. The 20- to 30-foot terrace is well developed on the south and southwest sides of Sagamore Hill, opposite the steamboat wharf, and it may be traced along the southwest side of the northern drumlin of World's End. A strongly marked terrace extends for a shoi^t distance on the northeast side of Planter's Hill, at a height of 55 feet. Occasional indications of terraces have been noted on the southwestern slopes especially of other drumlins in Nantasket and Cohasset, including Strawberry Hill, Bumkin Island, Turkey Hill, Scituate Hill, etc. But the only others requiring particular description are the series on the northeast side of Booth Hill, in Scituate. This great drumlin extends southeast from Bound Brook, near North Scituate, for about one mile. Its northern base is closely skirted by the railroad, and on this side it immediately 165 :||V overlooks the salt marshes trihu- tary to the Gulf. Going up from the marsh we come, at a height of 25 feet, to a distinct and level terrace, which is several hundred feet wide midway of the length of the hill, where the church stands upon it, but narrows toward either end. The road running from the church over the hill rises, at the cemetery, 15 or 20 feet to a well- defined terrace and bowlder pave- ment from 100 to 200 feet wide. In the vicinity of the church the main highway between North Scituate and Scituate Harbor runs on the lower terrace . But farther west, near the school house, it rises to the second terrace. From the school house a private road ascends the hill and brings us at a height of about 30 feet (70 feet above the marsh) to a third terrace, which is at least 300 feet wide and thickly strewn with bowlders. This terrace cannot be clearly traced so far east as the hio-hway across the hill ; and like both the others, it dies out near the west end of the hill. Following the private road about 30 feet higher (100 feet above the marsh) we come to a broad, level plain, which seems at first to form the summit of the hill; but on tracing it eastward 166 across the highway the hill is found to rise abruptly some 20 feet above the plain to the true summit, on which Hatch's Rock (Fig. 23) stands. The plain is here at least 300 feet wide ; and at all jDoints, save where it has been cleared for cultivation, it is strewn with bowlders. From this terrace the view seaward is broad and unobstructed ; and if the sea has ever stood at this level since the drumlin was formed, its storm waves must have broken with such resistless force against the banks of till as to require only a short time to carve this broad platform. Several of the lower drumlins of this region , which show no terraces— Green, Hoop-Pole, and Mann Hills — are so flat-top^jed as almost to suggest that their oi-iginal summits have been worn away by the sea or some agent of horizontal erosion. The lowest terraces on the drumlins consist chiefly and the upper ones to a very limited extent of modified drift, being, properly, narrow, fringing sand plains. But that the drumlins of this region also exhibit many true terraces of erosion, and that these are in many cases of such magnitude as to profoundly modify the normal contours of the hills, there can be no doubt or question. Concerning the origin and real significance of the cut terraces, however, there is still room for the widest differ- ence of opinion. As stated in the general description of the topography (page 7), I at first regarded these terraces as true shore-lines, and for the most part as marine shores, finding in them evidence of an important postglacial elevation of this coast. But a more extended and critical study of these features since the printing of this work began has convinced me that, as already pointed out, they lack the essential characteristics of true shores. They are not only deficient in continuity and uniformity of level ; but it is further impossible to regard them as marine shores, because, with rare exceptions, like Booth Hill, they are on the southwest sides of the drumlins and do not face the sea ; or as the shores of temporary lakes, because they are often, as in the Village of Hull and on the World's 167 End, in very narrow valleys between the drumlins, that is, in the most sheltered spots, where we would least expect erosion by waves. Still another difficulty in the way of regarding them as ordinary shore-marks is the fact that they occur only on the sides of the drumlins, the scai-ps never, so far as observed, changing their directions so as to cut across the ends of the hills. That is, they notch or break the transverse, but never the longitudinal, profile of the drumlins. We may safely conclude, then, that they have not been formed in or by standing water. It has been suggested to me that the terraces may, perhaps, be regarded as original features of the drumlins, owing their formation to the molding action of the moving ice, the drumlins having been grooved or fluted by the ice on a large scale. This explanation would account for their uniform parallelism with the major axes of the drumlins ; and I am inclined to believe that some of the less distinct terraces are really gigantic glacial grooves, formed, of course, while the ice still covered the drumlins, and not by the edge of the ice, since it is probable that when the ice had melted sufficiently to uncover the tops of the drumlins its motion had practically ceased. Starting with the normal type of drumlin, we recognize departures from it in several directions. Thus some are lono- and narrow or rido;e- like, while in others the longitudinal and transverse axes are nearly equal, and still others, though presenting the normal proportions of length and breadth, are unusually flat-topped, like Booth Hill and many of the lower drumlins, including Hoop Pole and Green Hills. These variations are probably original and due in part to the contours of the underlying- ledges and the character of the till and in part to the varying thickness and motion of the ice ; and no reason is apparent why the latter cause, at least, may not give us departures from the regular curvature of the transverse profiles. As a rule, however, the terraces are too deeply incised and much too sharply defined to be explained in this way. It is 168 highly improbable that in the case of a terrace due to glacial molding we should pass abruptly, over a well-defined edge, from the steep erosion slope or scarp behind the terrace to the normal slope of the drumlin above the scarp ; for it is incon- ceivable that the erosive action of a continuous sheet of ice could be so sharply diiFerentiated. The scarps are not simply a little steeper than the normal slopes, but they are in most cases as steep as the material will maintain, as steep and sharply defined, in fact, as the scarps of till on the modern shore which have been protected for a few years from the action of the sea. We are thus, apparently, forced to refer the terraces to the erosive action of running water. It is improbable that strictly subglacial streams would cross the drumlins at all : and if they did, the natural result would be to produce channels rather than terraces. The same would be true of superglacial streams, wherever they cut through the ice into the underlying drumlins ; and such channels, notching their summits, have been observed on some of the drumlins of eastern Massachusetts. The only alternative view that now seems worthy of consideration is that which refers the terraces to the erosive action of lateral streams. When the ice had retreated from the summits of the higher drumlins and, presumably, had ceased to flow, the heat reflected from the ground would naturally melt away the edge of the ice sufficiently to afford pathways for the superglacial streams, and drainage channels for glacial waters impounded north of or behind the drumlins. These lateral streams, having banks of ice on one side and till on the other, would necessarily erode the latter in the manner indicated by the existing terraces. This explanation accounts for the longitudinal direction of the terraces ; for the observed irregularities in height and level ; for the accumulations of modified drift sometimes observed at the eastern ends of tlie terraces ; and especially for the occurrence of the terraces chiefly on the southern slopes of the drumlins, since it is on these slopes that the heat of the sun would be 169 most effective in melting away the ice, the surface of which would naturally slope toward the drumlins. The terraces are thus a joint product of glacial and fluvial conditions, contempo- raneous with the modified drift ; and like that record the rapid waning of the ice-sheet when it had ceased to flow but yet lin- gered in the valleys. Some gullies or vertical erosion channels appear to have been formed on the drumlins at this time ; and probably the many other depressions and hollows observed on their slopes, which are often ill-defined, but commonly some- what saucer-shaped, are the result, chiefly, of local landslips dating from this period when the till was first laid bare and was yet unprotected by vegetation. THE MODEEiSr SHORE. When the sea finally retired from its highest postglacial level, it appears to have subsided quite rapidly to a level some- what below the modern beach. Evidence that the land has re- cently stood higher than now is found at other points in this region chiefly in the form of submerged peat beds and forests, and the former, at least, would probably be found if excava- tions were made in the Nantasket marshes. Although the depressions occupied by the marshes and the winding channels of Weir River Bay, Strait's Pond, etc., are clearly submerged land valleys, as previously explained, their origin — the erosion of the hard rocks — undoubtedly dates from the strong eleva- tion of the land at the beginning of the ice-age. In fact, it is not in land erosion, but in the absence of marine erosion at certain points where it might otherwise be expected to exist, that the evidence sought for is to be found. It is a fact familiar to all that the drumlins of the harbor exhibit marine erosion wherever freely exposed to the sea. This erosion scarp exists, in general, not only on the seaward, but also, in less degree, on the landward, sides of the drumlin islands. Being the work, almost exclusively, of the waves, it varies somewhat with the 170 breadth and depth of the water, and is most marked on the sides of the prevailing winds. It is, however, almost entirely wanting on the eastern ends of Thornbush and Telegraph Hills, Little Hog Island, Bumkin Island, the western part of White Head and Hampton Hill, as well as the eastern shores of World's End and Planters' Hill. At present, of course, these points are bordered by shallow water and, what is still more important, they are in the lee of Nantasket Beach, and are thus protected from the ceaseless beating of the ocean swell. It is perfectly clear, however, that this broad beach has been formed since the existing relations of the land and sea were finally established ; and if the postglacial elevation of the land had ceased when the present level was gained, the sea, sweeping freely through the broad passes between Strawberry Hill and Point Allerton on the north and White Head on the south, and the narrower gaps still further south, would inevitably have worn away the eastern ends of the masses of till which now enjoy the protection of the beach. Considerable and, at some points, ineffaceable results would undoubtedly have been accom- plished before the sea succeeded in throwing up the barrier beach that now checks its own depredations. The absence of sensible erosion at these points is, however, readily explained, if we admit that the comparatively rapid elevation of the land probably continued until it stood somewhat higher than at present ; and that a slow and gradual subsidence has since been in progress. Whenever, during this subsidence, the depres- sions between the drumlins of what is now the Nantasket Peninsula were brought within reach of the waves, barrier beaches were naturally formed across them ; and these barriers have kept pace in their upward growth with the subsidence, so that the land behind Nantasket Beach has been continuously protected from the action of the surf. This view relieves us of the necessity of imagining a cordon of drumlins outside of the present beach which have been completely washed away, although it is not improbable that Harding's Ledge and the 171 Black Rock Islets are the foundations of such vanished drumlins. It is obvious to the most casual observation not only that the present shore is strongly marked in all its various features, but also that its level has been unchanged for a long time. This is seen especially in the essentially finished state of the salt marshes, which, over broad and continuous areas, have clearly attained their maximum elevation — extreme high tide; and in the great breadth of Nantasket Beach. The front of the beach, when it was first established, probably formed a series of rather strongly concave curves connecting tlie drumlins, all of which were then fully exposed to the surf. The over-wash of the beach at this time formed what is now its rather low and marshy inner margin ; while through the additions made by the erosion of the drumlins the line of the. beach has been gradually straightened and advanced seaward until Strawberry Hill and the drumlins south of it, although their sea-cliifs are still sharply defined, are separated and protected from the breakers by a wide belt of sand and shingle. The Point Allerton drumlins are still within reach of the waves. The little hill would probably have been completely swept away by this time but for the massive sea-wall which has been built around it. Its present area is less than two acres, but it probably extended once as far as the beacon and had an esti- mated area of forty-five acres — as large as Boston Common. Between 1847 and 1860 its cliff receded 65 feet, equal to the loss of half an acre. The great hill has lost an estimated area of fifteen acres, but the outward growth of the beach has now nearly stopped the erosion. The finer material washed from the north side of Telegraph Hill has been carried to the west- ward to build Windmill Point, which is a miniature Cape Cod. In like manner the waste of Green Hill, and possibly of other drumlins which once existed in that vicinity, has formed the barrier beaches between Strait's Pond and the Atlantic ; and even the widest of these. Crescent Beach, is still so narrow that 172 it grows chiefly by over-wash during great storms. The same is true of the barrier beaches separating Little Harbor from the open sea, and those on the Scituate shore. The supply of fresh material being now virtually cut ofi^, the outward growth of the beaches must substantially cease or the additions consist of much finer material, such as might be brought from more distant sources. The increasing fineness of the sand on Nan- tasket Beach is apparent from the fact that distinct sand dunes may now be seen along the front of the beach north and south of Strawberry Hill, having a height in some cases of ten feet or more, while they are entirely wanting on the back side of the beach. At the south end of the beach the sand is also very fine, and abundant traces of it may be observed at heights of from 20 to 30 feet on the north side of Atlantic Hill ; while in the valley between Atlantic and Willow Ledge Hills a considerable bed, not of modified drift, but of the same fine white beach sand may be traced as far south as Atlantic Avenue. The accumu- lation of beach sand in this sheltered spot does not appear to be in progress now. But, although it could be used as an argu- ment for the recent elevation of the coast, I am disposed to ascribe it to strong northerly winds in past years. 173 Supplement. During the time that has elapsed since the printing of this work began, some additional observations have been made which, although not involving any material changes in my views concerning the geological structure of the region, it is thought best to indicate briefly, in order to make the exposition of facts as complete as possible. Besides the two small granite quarries on the Cohasset and Scituate shores, referred to on page 17, there is a third quarry on the south side of Scituate Hill which has afforded some o;ood stone for general use. Undoubtedly the most interesting flow of melaphyr in the Nantasket district is the second melaphyr in the Valley Beach section described on page 52. This flow is of particular interest, not alone because it can be traced in one direction to a distinct edge, but also on account of the numerous, irregularly rounded, amygdaloidal masses which it incloses. I have referred to these masses as pseudo-bombs ; but, although well satisfied, as stated, that they are not true volcanic bombs, or projected masses of lava, it is very gratifying to find this view confirmed by the observations of Prof, J. D. Dana on the lavas of the Hawaiian Islands ; and I am grateful to Mr. Geo. P. Merrill for calling my attention to the following statements by Pro- fessor Dana.i The aa streams are remarkable also for the presence of lava- balls of concentric structure that have been wrongly called bombs. These lava-balls ai-e smoothish exteriorly, more or less rounded and bowlder-like, and vary in size from an inch or less to ten feet and more. Some of these lava-balls have, outside, a crust of hard lava, and, inside, fragments of scoria ; others consist of concentric shells, hard and scoriaceous shells alternating with one another. 1 Amer. jouni. sci., 1887, ser. 3, vol. 34, p. 364. 174 One on Hawaii, near Punaluu, was found to have a nucleus of scoria eighteen inches in diameter, and around this successively a stony shell of three inches, a scoriaceous layer of from one to two inches, a stony shell of from four to five inches, and then outside a rough lava shell six inches thick. One of large size, broken open on one side, had had its inside filling of scoria worked out by the natives, and so made into a small cave. A common size on Hawaii is from three to five feet in diameter ; but one enormous lava-ball, in the aa field west of Punaluu, measured 24 by 12 by 9 feet in its extreme dimensions, and contained at least a thou- sand cubic feet. Enough of its hard outer shell was peeled off to ascertain that the second layer was quite vesicular or scoria- ceous, and the next layer inside hard basalt again. These Hawaiian lava-balls lie in the midst of the other blocks of the aa stream, proving that all had a common origin, and that they are not projected bombs, and hence properly not bombs at all. Professor Dana^ also refers in the following terms to the paper on "Fragmentary ejectamenta of volcanoes" by Dr. H. J. Johnston-Lavis'^, who has studied with much care the Vesuvian lavas and eruptions. He (Johnston-Lavis) shows that the "volcanic bombs" of writers on European volcanoes are not bombs any more than those of Mt. Loa ; that they were not projected into the air ; that they occur scattered over lava streams in great numbers when the adjoining country is free from them, and occur within lava streams ; that they vary in size from a walnut to some cubic yards, and yet have often a thin shell and friable nucleus ; that they occur most commonly by far on lava streams whose surface is rough and scoriaceous instead of corded. He regards them as formed of lapilli that fell upon the flowing lava, and, in consequence of its forward motion, became incorporated with it, and may undergo partial fusion, but usually congeal around themselves a coating of the lava in which they are involved. Dr. Johnston-Lavis ' Amer. journ. sci., 1888, ser. 3, vol. 36, p. 103. * Proc. geologists' assoc, London, vol. 9, no. 6. 175 also points out that the ejected blocks of solid lava are wholly different in origin, and readily distinguislied from the so-called volcanic bombs. The red slate referred to on page 55 as outcropping on the edge of the marsh, west of the Rockland House, is probably not a tuff ; but it appears better to regard it as a slaty layer in the conglomerate. A similar layer occurs in the conglomerate south of Willow Ledge Pond ; and, without wishins; to suo-o-est the exact correlation of these very limited deposits of finer sedi- ment, it has appeared best, on the map, to indicate the con- glomerate in the vicinity of the pond as extending westward far enough to embrace the second outcrop of slate. There appears now little reason to doubt that the large out- crop of hai'd, slaty rock on the railroad, west of the Rockland House, is a true tuff; and it should probably, in spite of its finer texture, be correlated with the main bed of tuff at the base of Atlantic Hill. In the description, on page 67, of the faults crossing Cres- cent Hill, it should have been noted in connection with the third fault, which accompanies the small dike, that the dike itself marks a fault plane of later date than the other and compensating with reference to it. The two faults hade in opposite directions, and the only displacement of the strata actually observable is the elevation of the triangular mass of melaphyr between the two faults. It now appears probable, also, that there is one more fault on the southern slope of Crescent Hill than has been described in the text ; and it has been so represented on the map. There appears now no sufficient reason to extend the fault which is supposed to cross the western end of Conglomerate Plateau southward to the granite and the boundary fault, as suggested on page 74 ; and it has not been so drawn. The position of the conglomerate overlying the melaphyr south of Conglomerate Plateau and east of Round Hill appears to be slightly synclinal, probably rising both to the east and west. It now appears improbable that all of the contacts of conglom- 170 ei'fite and mclaphyr in the area between Hull Street and Strait's Pond are duo to faults, as stated on page 75. The more prob- able view is that expressed on the map, viz., that the alternate contacts are faults. The conglomerate of Round Hill and Con- glomerate Plateau, with a oentlo southerly dip, is overlain by the second nielaphyr ; and tiie outcrops of these two beds are repeated several times by east-west displacements downthrovving to the north. Pecent observations have thrown additional light upon the relations of the dikes of the central belt in the vicinity of Hull Street ; and nearly all the dikes in (he section on the railroad can now be traced across Conglomerate Plateau. Nos. 25 and 27 are clearly the same dike : and this can be traced now, as shown on the map, to the intersection of Hull Street and Atlantic Avenue. Just west of Hull Street it appears to cross the course of 29 ; and what appears to be a branch from the latter quite clearly cuts 25 on the north side of Hull Street, and then runs parallel with it. No. ol also branches near Hull Street; but there seems to be no way of dctinitely deciding whether the dike which ultimately reaches the shore of Strait's pond is 29 or 31. No. 24, on East Porphyrite Hill, is clearly a continuation of 38, on Granite Plateau : and there is a clear path for it across Cliff Plateau, although it does not appear to be actually exposed. For the elevations of some of the drumlins in Cohasset 1 am indebted to Mr. H. W. Nichols, a. student in the Geological De[)artmcnt of the Massachusetts Institute of Technology ; and for many of the names on the maps I am under obligations to Col. E. T. Bouve. He has prepared a very interesting chap- ter on the ancient land-marks of Hingham and Cohasset for the forthcoming history of Hingham, and has kindly permitted me to use the proof-sheets. I also gladly embrace the opportunity which this supplement affords to acknowledge the important aid in the prosecution of this work which I have received during the past year from Mr. Thomas A. Watson. He has not only made possible a great 177 ccoiiotny of my time wliilo in tho fl(;l(I ; \)\it. Ijc has proved lii/n- Holfan uriiisiKilly acute and accurate obHCjrver, vvitli a rea(Jy arj(J clear coinprehenHion of the bearin^H of neaHy every new fact ; and njy work \iuh been at Korne jjointH rnateriaJJy Htrcngtiiened by tlie data and conclusions wliir-h hf; lias \'n-c]y placed at my disposal. #aasi0ttitl papers OF I' H E iv. GEOLOGY OF THE BOSTON BASIN WILLIAM 0. CROSBY. VOL. L PATiT II.— HINGHAM. -K» — -. BOSTON : BOSTON SOCIETY OF NATURAL HISTORY 1S94. GEOLOGY BOSTON BASIN WILLIAM 0. CROSBY. IN TWO VOLUMES. VOL. I. PART If.— HINGHAM. BOSTON : BOSTON SOCIETY OF NATUKAL HISTORY. 1894. HINGHAM. INTRODUCTION. The town of Hingham, Mass., forms the south shore of Boston Plarbor for three miles, from Weymouth Back River across Hingham Harbor to Weir River Bay ; and extends inland from six to seven miles, crossing the present border of the Boston Basin, i. e., the southern limit of the sedimentary and volcanic rocks. The town is thus divided into two distinct and very unequal geological areas ; and, as the general map (PI. 1) shows very clearly, the dividmg line is quite irregular. The sedimentary rocks and the interbedded lavas are limited almost wholly to the northwest corner of the town, extending but little south of the railroad and having only a slight areal development east of the harbor. While over the remainder of the town, embracing more than five sixths of the total area, the numerous ledges comprise only granitic rocks (granite, diorite, and felsite), and intersecting dikes of diabase. The granitic area of Hingham is essentially similar to, as well as continuous with, that of Cohasset and Nantasket on the east and Weymouth and Braintree on the west, the entire South Shore district being a unit in this respect. But, as was pointed out on page 2, the sedimentary and volcanic rocks borderinor the granite on the north and forming the immediate shore of the harbor are far less uniform in character and struc- ture, and by their diversity fully warrant or necessitate the division of the South Shore into several distinct areas, which are best described separately. These geological areas agree approximately with the political divisions, the geology of North 180 Hingham contrasting with that of Weymouth on the west and still more with that of the Nantasket areas on the east. The promontory of Kooky Neck, northeast of Planter's Hill, at the mouth of Weir River Bay, is, however, essentially a part of the Nantasket area, and it has been described in that relation, the true or natural boundary between the Hingham and Nantasket areas being marked approximately by the eastern shore of Hingham Harbor. On account of the more abundant drift-deposits, Hingham does not present the almost continuous rock exposures which characterize the Nantasket area. But, fortunately, a different type of structure prevails, and extended outcrops are less essen- tial to the correct interpretation of the geology. Plication, to a large extent, takes the place of faulting, the sedimentary and volcanic rocks being involved in deep and almost isoclinal folds ; and hence, although the actual disturbance and displace- ment of the strata are probably greater in Hingham than in Nantasket, the beds are more continuous and more easily traced in infrequent outcrops. Again, while Nantasket shows repeated alternations of beds of conglomerate with flows of both basic and acid lavas (melaphyr and porphyrite), the porphyrite, so far as known, is wholly wanting in Hingham and the melaphyr is limited to one flow or bed of great thick- ness ; and the princij)al problem of the Nantasket area — the identification of the successive flows of lava — is really not presented at all to the students of Hingham geology. On the other hand, while the sedimentary rocks of the Nantasket area, south of the beach, are almost exclusively conglomerate, the Hingham series embraces many beds of sandstone and brownish slates and a great volume of pure gray slate ; and the special feature of the geology of Hingham, the feature in which it excels not only Nantasket but the entire Boston Basin, is the extended series of alternating beds of conglomerate, sandstone, and slate which it presents in three different sections, and the seemingly clear exhibition of the relations of this conglomerate 181 series to the great slate scries. It is possible, as will a[)pear later, that the Hinghain ledges supplement the Nantjisket ledges, the basal beds of conglomerate, resting upon the granite floor, having a remarkably fine development in the latter, while the former afford continuous exposures of the upper beds of conglomerate and the overlying slate. In my work on the geology of Hingham I have been greatly assisted in various ways by Mr. T. T. Bouve. In fact, we have traversed a large part of the ground together, have com- pared notes at nearly every step, and have discussed together all the interpretations of the facts occurring to either. I thus find myself wholly unable to determine in all cases what part of the work is really my own ; but gratefully acknowledge my indebtedness to Mr. Bouve for ideas as well as material assistance in the field-work. Mr. Bouve has prepared a general account of the geology of Hingham, which forms a part of the history of the town ; and the Committee in charge of this work have kindly permitted the Society to cooperate in the printing of the three special maps of Hingham accompanying this paper. TOPOGRAPHY. Fundamentally, or so far as the hard rocks are concerned, the topography of Hingham is based upon the westward exten- sion of the broad peneplain which we have traced in Cohasset and Nantasket, and which forms the entire south shore of Boston Harbor. In Hingham, as farther east, the peneplain is proved by the generally uniform height of the rocky elevations ; and the evidence is equally clear that the plain itself represents prolonged preglacial erosion ; while the deeply channeled and fragmentary form which the plain now presents must be attributed, as before, mainly to the comparatively rapid erosion attending the strong elevation of the land at the beginning of the great ice-age. Owing chiefly to the large proportion of 182 slate and the fact that the conglomerate and sandstone are less generally indurated through intimate association with the volcanic rocks, the sedimentary rocks of Hingham are less prominent, topographically, than those of Nantasket, The slate, where occurring in large bodies, is very generally eroded nearly to or below the present level of the sea, while the con- glomerate and melaphyr are found mainly below the contour of forty feet, and never rise much if any above seventy feet, attain- iuo- their greatest elevation in the rids-e alonij; South Street in Hingham Village. The granitic area, on the other hand, is, as a whole, somewhat more elevated than in Cohasset, owing prob- ably to its greater distance from the sea ; the average or normal elevation of the ledges south of the railroad ranging, probably, from 60 or 70 feet to nearly or quite 150 feet in the south part of the town. Although the hard rocks thus show, from the lowest slate to the most elevated ledges of granite, a notable range in altitude, there are, properly speaking, no rock hills except such as have resulted from the division of the peneplain and its reduction to a fragnientary condition by glacial erosion. Especially do we note the absence, as in Cohasset and Nantasket, of dominant rock hills or those decidedly overlooking the pene- plain. The characteristic features of an ancient topography — well-defined ridges with culminating summits separating broad level valleys — are wanting here and in Cohasset and through- out this region ; but we find instead the comparatively narrow and abrupt valleys and the broad, plateau-like, interstream surfaces indicative of a region which, having been worn down to its base-level, has found a renewal of its topographic life in a comparatively recent elevation and is only now well started in a new cycle of topographic development. This topography is geologically ancient only in the sense that it starts, not from an elevated sea-bottom, a virgin land-surface, but from an elevated peneplain, itself the last expression of an earlier topography ; and if the present stability of tlie land continues, the features of the modern topography will, after attaining a 183 maximum of ruggedness and diversity, be gradually effaced in a new peneplain at the existing base-level. Although the average; height of" the rock sui-face gradually increases as vv^e recede from the shore, and es{)ecially as we pass from the sedimentary to the granitic areas, the drumlins, which are planted much more thickly near the harbor than farther inland, tend to neutralize this natural erosion gradient. In Hingham, as in Hull and Cohasset, the till, so far as exposed to observation, occurs mainly or almost wholly in the form of drumlins ; and, as the general map so plainly shows, no less than eleven distinct drumlins rise from the low lands bordering the harbor and north of the railroad, including the two double drumlins of World's End and Planter's Hill, east of the harbor. Of the whole number, fully one half are especially typical in form and of nearly the first magnitude ; and since they rise directly from the harbor or the salt marshes, with their graceful slopes unbroken, save by an occasional erosion terrace, they are more impressive than the inland drumlins, bordered and obscured by elevated ledges and sandplains. Crow Point, at the northwest corner of Hingham Harbor, finds its origin in a linear series of three closely connected drumlins ; and it is noteworthy that the other four drundins on the west side of the harbor lie north of or upon the ridge of granite extending from the south end of the harbor to Beal Street. South of the railroad the drumlins are much more scattering. The first to attract attention is the remarkable linear group or chain known as the Turkey Hills, on the boundary between Cohasset and Hingham. This consists (see the map) of one main drumlin of the first magnitude with a low ridge extending southwesterly from it in which we readily recognize three small and approxi- mately equal drumlins. A fifth and closely connected drumlin lies immediately south of the main hill ; and the very perfect detached drumlin concealed in the woods of Turkey Swamp may be referred to this group. Great Hill, south of West Hing- ham, belies its name, since it is one of the smaller drumlins of 184 the town. Its height does not exceed 125 feet, and it rises less than 60 feet above the bordering sandplains. Nutty HilL southwest from Great Hill, is a good example of a drumlin more than half buried in the sandplain. Southwest of the Turkey Hills is an extended area of elevated rocky Avoodland? lying partly in Hingham and partly in Cohasset, a singularly well-preserved section of the ancient peneplain. But, to the west of this tract, in the valley of Weir River, before we come to Prospect Hill, on the east side of the river and just beyond the southern border of the map, several rather inconspicuous drumlin s have been observed. Prospect Hill is well named ; for it is one of the largest drumlins in the Boston Basin, the hio-hest point in Hingham (218 feet) ; and affords a wide prospect in all directions. The general map gives the position, form, size, trend, and, so far as known, the height of every recognized drumlin in the area which it represents.^ The rock surface of Hingham is masked not only by the drumlins, but much more by the modified drift, which is far moi'e abundant than in Hull and Cohasset and is broadly developed in level plains as well as in rounded knolls or kames and winding ridges or eskers. Although the more or less continuous plains of sand and gravel occurring at different and increasing heights from the shore southward tend, as in Cohasset, to emphasize the plateau character of the peneplain, they also give a broadly step-like or terrace form to the topog- raphy in which the hard rocks certainly do not share. Some of these plains gained early recognition, as witness the names 1 The elevations in nortliern Hull, including Peildock's and Little Hog Islands, were taken, by permission, from the unpublished charts of the Harbor Commissioners; while those in southern Hull, including Bumkin ajid Grape Islands, were measured with a hand-level, by the author. The chief elevations in Hingham and Cohasset, including nearly all those conveniently near the sea or salt nuirshes, were very accuratel}- deter- mined by levelling through the kindness of Mr. A. E. Woodward, Mr. Cyrus C. Babb, and Mr. H. W. Nichols, formerly students of civil engineering in the Massachusetts Institute of Technology; and, finallj', the elevations in southern Ilinghani ;ind Cohas- set have been, for the most part, determined barometrically Ity (he author, with the assistance of Mr. Nichols. 185 on the map — Lower Plain, Glad Tidings Plain, and Liberty Plain. Lower Plain , with a normal height of from 50 to 55 feet, has itw best development between High Street and the railroad, west of Weir River, including Hingham Centre and the district about Great Hill. Toward the west and northwest especially, it loses, very largely, the character of a plain, dividing into irregular rounded hummocks and winding ridges or cskers enclosing numerous kettle-holes and small ponds and bogs. In this fragmentary form, only rarely attaining its maximum height, this plain has a particularly interesting development in the Hockley district between West Hingham and Weymouth Back River. The fine series of eskers between Beal's Cove and Stoddard's Neck should be referred to this plain, and also the plain (40-50 feet) so well developed around the southern end of the harbor. Eastward, along the line of the railroad and East Street, it can be traced into Cohasset, where we have already recognized it as the principal jolain of that town. There is, however, in the vicinity of the coast, a distinctly lower plain, with a normal height of from 20 to 30 feet. This is very perfectly developed between Beal Street and Huit's Cove and may be traced at intervals through the northwest part of the town. Glad Tidings Plain extends from the vicinity of High and Free Streets south to the northern margin of Liberty Plain, embracino- Cushins; and Fullin^ Mill Ponds and havino- a normal height of from 65 to 70 feet. Liberty Plain lies almost wholly south of the southern boundary of the map, embracing Accord Pond and extending into the adjoining towns. It rises very abruptly from Glad Tidings Plain to its normal height of 130 or 135 feet. Each plain is developed to some extent in the form of out- liers or islands on the next lower one, and, conversely, kames, eskers, or limited plateaus rising from the surface of either plain may be regarded as representing in height and age the next higher plain. The eskers known as Break-neck Hills, 186 beyond the southern border of the map, on Gushing Street, hold this relation to Liberty Plain, Glad Tidings Plain may thus be traced as far north as Hobart Street, where it forms a very perfect plateau (Pigeon Plain), and it is probably represented by the high ridges or eskers (70-85 feet) between Fort Hill and Beal's Cove. Although the sand plains testify to the postglacial flooding, if not to an actual depression of the land, the salt marshes and the drowned valleys of Weymouth Back River, Hingham Harbor, and Weir R,iver Bay are a sufficient indication that the land formerly stood higher than at present, and that the exist- ing level has been maintained for a very long time. It i§ obvious, then, that so far as the lithified formations are con- cerned, the relief features of Hingham may be summarized as follows : Hingham and the adjoining towns are an area of hard rocks which, in preglacial times, had been slowly worn down nearly to its base-level ; and such topographic ruggedness as was developed in this old peneplain during the strong elevation which ushered in the glacial epoch is pretty well smothered by the marine deposits and the almost continuous mantle of drift. In its drainage system Hingham is almost a unit. With the exception of the northwest corner of the town, which drains directly into the harbor and Weymouth Back River, and the limited basin of Fresh River, also tributary to Weymouth Back River, nearly the entire area is drained by Weir River and its branches ; and this system derives but very little water from beyond the limits of Hingham. It rises in Valley Swamp, in Norwell, and in Accord Pond, which lies in the three towms of Norvvell, Abington, and Hingham. The streams are all small, and, although the total fall is considerable, the drainage is, as a whole, decidedly sluggish, the streams meandering through ))road level meadows and swamps, with little power to clear out their drift-encumbered channels. Strangely enough, Weir River is not now tributary to Hingham Harbor, but when within three fourths of a mile of the head of the harbor and 187 one fourth of :i mile of the Home Meadows, a sak rnarsli wliich is virtually a southward continuation of the harbor, it is abruptly deflected to the eastward by the ridge of modified drift along East Street, a spur from the Lower Plain, and pours its waters over the granite ledges into Weir liiver Bay. The facts concerning this very clear case of diverted drainage will be more fully presented in the section on the surface geology. Not only the streams, but also the areas of obstructed drainage, the ponds, swamps, and marshes, have been traced out with considerable care; and all these features, so far as the scale will permit, are delineated with approximate but not uniform accuracy on the general map. With the exception of Accord Pond, which belongs only in part to Hingham and is beyond the limits of the map, the larger ponds, including Triphammer, Fulling Mill, Gushing, and Foundry Ponds, are all artificial, originating in the construction of dams at favor- able points across Weir River or its tributaries. Round Pond, on the lower part of Fresh River, is also a mill-pond. Accord Pond, the principal source of the water supply for Hingham and Hull, and 135 feet above the sea, owes its existence to the sandplain and connecting esker which form a natural dam directly across the valley in which the pond lies. The swamps, which often represent ponds in course of extinction, still embrace occasional limited sheets of water ; and kettle-ponds, or ponds occupying the deeper hollows in the modified drift, are fairly common in some districts, especially over the broad area of modified drift west of Weir River. Some of these are too small for accurate representation on the map, and the majority ai'e wet-weather ponds only. The swamps and marshes represented on the map require no special description. At the lower end of every fresh-water marsh or swamp there is an obvious obstruction, usually of modified drift, rarely ledges. The feeble and sluo-o-ish character of the streams is seen in the little progress they have made in trenching these drift barriers. An occasional swamp crosses a low water-parting and drains in 188 opposite directions. This is true of the swamp on Rockland Street, of that east of Huit's Cove, and of that between Hersey and Central Streets, the latter being tributary to both Town Brook and Weir River; while many of the smaller swamps, especially, have no visible drainage, occupying more or less irregular kettles or depressions in the sandplains. Some of what Avere originally salt marshes are now fresh, having been reclaimed by the construction of artificial dikes or barriers. These are the "damde meddowes" of the early inhabitants, and the two most important examples are those on Weir River Bay, south of Rocky Neck, and west of Pleasant Hill. The considerable swampy tracts which have been reclaimed by artificial drainage are not represented on the map. The beach deposits, which are such a prominent feature of Hull and Cohasset, are almost entirely wanting on the more sheltered shores of Hingham ; and there is a corresponding absence of erosion of the drumlins and other drift formations. The World's End is joined to Planter's Hill by a short barrier beach, but the most important beach of any kind is that on the north shore, northwest of Pleasant Hill, separating one of the "damde meddowes" just referred to, from the sea. The prin- cipal drowned or submerged valley lying wholly in the town is Hingham Harbor. The silting up of this basin has, over the greater part of its area, reached the eel-grass stage ; and it would probably have become a salt marsh long ago if the scouring action of the tides had been checked by a barrier beach across its mouth. The same is true of Weymouth Back River. The tidal scouring is particularly efficient here on account of the contracted form of the basin. MAPS. This part is illustrated by one general and three special maps. The general map is but a continuation, on the same sheet, of that for Hull and Cohasset, and lience may be 189 described in similar terms. The priMcipal obje(;t of this iiiap is to show the general distribution and relations of the hard rocks, the surface geology, and the topography. This map, as has been previously explained, affords, even in the absence of contour-lines, a general idea of the relief-features. The out- lines of the principal elevations and depressions are shown in the drunilins and the drainage system ; and the remainder of the surface consists either of the sandplains with the kettles and kames, or of ledges, the isolated remnants of the ancient pene- plain. It thus requires only a little imagination to see in the map the actual form of the surface. The delineation of the surface features — the drumlins, streams, swamps, etc. — is far from being uniformly accurate. They have been traced out with the greatest care in the northern part of the town, and in general where the country is most open and accessible. Some of the wooded areas, remote from the streets and destitute of land-marks, have been only very imperfectly explored. This is especially true of the large tract of rocky woodland lying between King Street and Scituate Pond in Cohasset and Union Street and Beechwood River in Hingham, embracing more than four square miles and crossed by only one road — the most complete wilderness in the Boston Basin outside of the Blue Hills. For many of the names of the hills, streams, and other natural features which appear upon this map I am espe- cially indebted to Mr. Edward T. Bouve's most interesting- contribution to the town history on the "ancient landmarks" of Hingham and Cohasset. The special maps, having been printed in co-operation with the Town of Hingham and before the plan of this work was fully decided upon, do not agree in scale and topographic detail with the special map of Nantasket, the most notable difference being the absence of contour-lines and the outlines of the swamps and marshes, and the repre- sentation (in black characters) of the actual outcrops, as well as the theoretical distribution (in colors) of the sedimentary and volcanic rocks. These special maps represent the three 190 sections of special geological interest, exclusive of Rocky Neck ; and, territorially, they are continuous from Crow Point via Huit's Cove, Beal's Cove, and West Hingham to the southern end of the harbor. THE GRANITIC ROCKS OF HINGHAM. On account of the more continuous development of the sand- plains, the outcrops of the granitic rocks over the greater part of Hingham are less frequent than in Cohasset ; and nowhere in Hingham do we find any section of these rocks comparable in clearness and continuity with that along the Cohasset shore. There can be no question, however, that the granitic rocks of Hingham are essentially similar in most respects to those of Cohasset. They embrace in the same general proportions the three principal types — granite, diorite, and felsite. The diorite is, in every case, clearly the oldest rock, its relations' to the granite being the same as in Cohasset. Considerable attention has been given, especially by Mr. Bouve, to tracing the distribution of the diorite, with the object, originally, of repre- senting it by a separate color on the map ; but tliis lias been found wholly impi-acticable. Its relations to the granite are so intimate and intricate that, in the absence of perfect and con- tinuous exposures and a map on an inconveniently large scale, no lines could be drawn which would not include considerable granite or exclude a large pai't of the diorite. We have learned, however, as the result of this attempt to map the diorite, that, while occasional inclusions of diorite (often very small inclusions, it is true), may be observed in almost every good exposure of granite in Hingham, this rock occurs abundantly only in a fairly well-defined east-west belt near the northern edge of the granite. The limits of this belt are roughly indicated by the distribution of the V-shaped characters on the general map.^ Commencing 1 The diorite of Cohasset is siinilavly represented; but tliis feature of the map had not been adopted when tlie Cohasset text Avas printeil. 191 at the Cohasset boundary, in tlie vicinity of" tlic railroad, I he diorite is observed most abundantly in the ledi«,(!s aloiio- Hull Street, Weir River Lane, Kilby and East Streets ; and tiicn, after a partial gap of more than a rnile where the ledges of all kinds are very generally concealed by the Lower Plain, north- west of Fort Hill, near West Hingham, for a short distance, until all the ledges are again blotted out by the modified drift. It should not be supposed, however, that these marks represent, in every case, continuous ledges of diorite, for the most of the ledges are of a mixed character, granite penetrating or inclosing diorite. This irregular belt, it Avill be observed, is in about the same latitude as the principal occurrences of diorite in Cohasset, and on the same east-west line are found the chief outcrops of diorite in Weymouth ; but south of this belt we have observed no important or notable masses of diorite. In lithological character the diorite of Hingham is scarcely to be distinguished from that of Cohasset, showing similar variations, except that it nowhere exhibits a distinct flow-structure. It is usually finely crystalline and dark colored ; but occasionally it is coarser, with the hornblende either very clearly and prominently developed or mainly wanting, giving a light-colored, feldspathic variety which might be readily mistaken for syenite or even granite. Epidote is, as usual, the principal secondary mineral, occurring chiefly as narrow and irregular segregations and veinlets, especially along the joint cracks. Although the granite of Hingham almost certainly embraces the three distinct types — distinct in age and character — observed on the Cohasset shore, few ledges have been observed in which the relations of the two older types — the basic hornblendic o;ranite of medium texture and the coarse acid granite — are clearly exposed. Perhaps the most favorable exposure of this kind is that afforded by Button Island, in Hingham Harbor. The bedrock of the island is wholly granite, with only a thin covering of drift. The granite is chiefly an unusually coarse and distinct example of the second or more acid type, enclosing 192 small fragments of fine-grained diorite. But it also encloses on the south and west shores much laro-er masses of what appears at first sight to be the older and more basic granite of medium texture. The contacts are, however, somewhat ambigu- ous, appearing to favor the view that the finer-grained granite is the newer, and hence to be correlated with the third or microgranitic type. By far the greater part of the granite evidently belongs to the newer variety, the typical, sparingly hornblendic, usually coarsely and distinctly crystalline, gray to pink or red granite of the South Shore district. This is especially true where the granite is not moat intimately associated with the diorite. The hornblendic element, as in Cohasset, is very generally replaced partially, sometimes wholly, by mica (chiefly biotlte) . This is seen very clearly in the abandoned quarry on Long Bridge Lane, which is, with one exception, the only point in Hingham where the granite, or any rock, has been systematically quarried. The feldspar (orthoclase) is commonly gray or pink, more rarely green, at least in part; and, very locally, as in some of the ledges along Thaxter Street, it is porphyritically developed. Undoubtedly, one of the most interesting exposures of granite in Hingham is the small quarry about one third of a mile northeast of Abington Street, in the southwest corner of the town, and nearly two miles beyond the limits of the map. On going in from Abington Street by the quarry road, the ordinary, coarse, biotite granite gives way, with apparent abruptness, to a finely crystalline, homogeneous, light gray variety, which is rich in quartz, and contains but little of an 3^ dark accessory. The other limits of the fine granite were not observed, but it appears to cover a considerable area. The feature of particular interest which it presents is the remarkably perfect parallel joint-structure. In fact, it is the jointing that o-ives the granite its value as a quarry-stone. The parallel jointing is traceable over an area at least 500 feet long north and south, and half as broad ; but it is most perfectly devel- 193 oped in the more northerly and hirger of the two small quarries. The joints of one system are phenomenally perfect, close, and parallel, dividing the granite into almost absolutely plane sheets varying in thickness from half an inch to two feet or more, hut mostly from four to twelve inches. The trend of the joints is approximately N. S. (S. 5°W.), and the hade about vertical (W. 0°-3°.) The granite thus, in a general view, simulates a bedded rock, like sandstone, very closely ; and the surfaces of the sheets are so plane and smooth and the grain so perfect, that blocks suitable for building purposes are obtained with remarkable ease, the joint-structure serving the same useful purpose in this granite as in the Roxbury puddingstone. It is, of course, needless to dwell upon the obvious support which this example lends to the earthquake theory of parallel jointing.^ Irregular dikes and masses of the more finely crystalline and micro-crystalline granite, and of felsite, are frequently observed cutting through the coarser granites and also the diorite, precisely as in Cohasset. Perhaps the most interesting exposure of these older rocks is that on the summit of Fort Hill, where the diorite has been laid bare in grading the street. The diorite is very sharply and clearly intersected by numerous narrow, branching dikes of a finely crystalline pinkish granite. The dikes are broken by slight faults and enclose angular frao^ments of the diorite.^ On the north side of East Street, between Andrew Heights and Kilby Street, the diorite is cut by an irregular dike of a gray felsite from eight to ten inches in width. This proves on analysis to be a basic felsite, agree- ing in composition better with syenite than granite. At many points on Andrew Heights and along Kilby, East, and other ' Proc. B. S. ISr. H., V. 22, p. 72-85. * Since the above paragraph was written, this beautiful and instructive exhibition of the relations of the granite and diorite has been entirely and ruthlessly obliterated by the farther grading of the street ; a fact which every student of Hingham geology will sincerely regret. OCCAS. PAPERS. B. S. N. H. IV. 13. 194 streets as far as the Cohasset boundary, the ordmary coarse granite can, in spite of the lichens, be seen to form irregular dikes in, or to enclose angular fragments and masses of, the diorite. The greenish gray quartz-porphyry obscurely exposed in South Street, east of Hersey Street, in Hingham Village, is probably a dike in the granite , but considerable digging would be required to prove it. In the rear of the first house on Lincoln Street and Fountain Square, north of the Unitarian Church, is a mass, clearly eruptive through the coarse granite, of a compact, flinty, purplish felsite, which has been proved by analysis to agree with the granite in composition. The felsite of Hingham is not wholly intrusive, or in the form of dikes. The gray felsite, on the north side of Beal Street, at the Avestern end of the granite, is quite probably part of a surface flow ; and the beautiful red felsite occurring so plentifully in the form of bowlders in the vicinity of Lincoln and Thaxter Streets is unquestionably effusive. The former varies from a slightly greenish gray or white to a pinkish tint, and encloses many more or less distinct fragments of similar or darker felsite and an occasional fragment of granite. The breccia structure thus resulting is so marked in a portion of the rock that it was at first mistaken for conglomerate ; and the isolated elliptical area on this part of the map marked and colored as conglomerate is really felsite. The exposures of the felsite are not sufficient to show clearly its relations to either the granite, which it seems to overlie, or the melaphyr, which probably once covered it. The red felsite of Thaxter and Lincoln Streets is by far the most attractive of all the older rocks of Hingham. Li fact, it is the most beautiful variety of felsite in eastern Massa- chusetts ; and it is also unique in being the only felsite in the entire south shore district which is certainly effusive. Unfortunately, it occurs in an area which is covered almost continuously by salt marshes and drundins, and there is prac- tically no opportunity to study it 171 situ. This rock, which has attracted attention since tlie first settlement of the town, was described by Prof. Edward Hitchcock, in his final rejjort upon the geology of Massachusetts, under the head of por- phyry, as occurring in ridges a little north of the village. Such ridges do not exist now ; and the statement of this accurate observer should, perhaps, be interpreted as referring to ledges along the line of Lincoln Street which have probably been effaced by the subsequent grading of the street. At the present time, the only actual exposure of the rock in siiu is an obscure outcrop in the roadway at the junction of Crow Point Lane and Downer Avenue. This was formerly a some- what protruding ledge, but it does not now rise above the level of the street, and might easily be overlooked. It is in an area colored as melaphyr on the map, and it is regarded as a boss of the felsite projecting, as the result of erosion, through the sheet of melaphyr. The character for felsite also appears on the map (PI. 9) behind Mr. Bradley's barn, west of Thaxter Street arid south of Lincoln Street. No actual ledge of felsite has been observed here, but a number of excep- tionally large, angular bowlders in the drift, some of which can be seen on the surface, while others were exposed in a temporary excavation, suggest that the ledge is close by. Immediately north and west of this point there are outcrops of melaphyr, and no bowlders of felsite can be seen. ITence the northern boundary of the felsite may be regarded as accurately located at this point. But the boundary between it and the granite on the south is a mere matter of conjecture, although the line on the special map extending from Broad Cove, south of the solitary house on Lincoln Street, across Bradley's Hill and Thaxter Street, is probably an approxima- tion to it. The distribution of the bowlders certainly suggests an east-w^est belt of the red felsite along the northern edge of the granite, and passing under Bradley's Hill. That it extends east under Broad Cove is at least probable ; and that it extends west beneath Squirrel Hill we have some 196 evidence in the ledge of felsite at the western base of this hill and some 500 or 600 feet south of Lincoln Street. At this point, however, the felsite, although still of a reddish or purplish color, is much more compact and homogeneous. Although there is no opportunity to observe its relations to other rocks, the eifusive or volcanic nature of the red felsite forming the numerous bowlders east of Squirrel Hill and the solitary ledge already referred to, on Downer Avenue, is abundantly proved by its structural features. It exhibits throughout a distinct but not conspicuous banding or flow- structure, which is usually rather fine, but sometimes quite coarse and often somewhat contorted, discontinuous, or other- wise irreo-ular. The confused and irreo-ular character of much of the banding is due in part to the enclosure of angular frag- ments of the same or a very similar felsite. The enclosed fragments vary from a small fraction of an inch to several inches in diameter, and are very irregularly distributed, so that while the greater part of the felsite is comparatively free from them, they vary in the remainder from thinly scattering to densely crowded, occasional masses of the rock being packed so full of fragments that the banding is completely obliterated and it closely resembles an ordinary breccia. This fragmental character of the felsite has suggested to several observers, including the present writer, that it is a metamorphic conglomerate. But having observed precisely similar structures (banding and brecciation) in modern obsidian, I have been for a long time thoroughly convinced that this metamorphic theory is untenable, and that the felsite is a true volcanic rock, a devitrified obsidian. Among the arguments against its sedimentary origin are the fxcts that the fragments are all of the same kind of rock ; that they are never assorted or show in any way the action of water ; and that the felsite appears, as has been proved b}^ analyses in other cases, to be chemically essentially intact and homogeneous, still retaining in every part the full [)r()portion of alkalies required for an acid feldspar, which would be very unusual in a 197 clastic rock. The fact that the fvag:ments or so-called pehhles show a gradation in distinctness from those that are very sharply defined to those that are perfectly blended with the enclosing felsite, is only what would be expected when fragtnents of glass (obsidian) are enveloped in melted glass. The only particularly obvious indications of chemical change in the felsite since its eruption are, first, the red color, which may be original, but is probably due in part at least to the peroxidation of iron during devitrification ; and, secondly, the occurrence in the rock of inconspicuous streaks and masses of quartz, either vitreous, chalccdonic, or jaspery, — silica i-eplacing the less stable portions of the glass. Red felsites are not uncommon in eastern Massachusetts ; but the only occurrence at all closely resembling the Hingham felsite is the red felsite near the Neponset River, in Hyde Park. Structurally they are strik- ingly similar, except that the banding of the Hyde Park variety is rather coarser, and it is more generally brecciated ; but the Hingham felsite, with its deeper and brighter color is the more beautiful of the two. THE BEDDED ROCKS OF NORTHERN HINGHAM. GENERAL RELATIONS AND ORIGIN. It is unnecessary to repeat what has been stated under this heading for the Nantasket area ; but the one topic may be regarded as supplementary to the other. The sedimentary rocks of Hingham present, as previously stated, a greater variety than those of Nantasket, embracing besides the conglomerate many intercalated beds of sandstone and slate of both brownish and greenish tints, and the great body of gray slate overlying this conglomerate series. The contemporaneous volcanic rocks or effusive lavas, on the. other hand, are less varied, including no acid or fragmental varieties, but consisting wholly of rather 198 typical melaphyr. Furthermore, the melaphyr appears to be limited to one heavy bed near the base of the conglomerate sei'ies ; and the repeated alternations of sediments and lavas so characteristic of the Nantasket area are wanting in Hingham. The stratigraphic contrast is so great that a satisfactory correlation is impossible with the data now at command. It appears probable, however, as stated in the introduction, that the two areas are complementary, the Hingham series beginning with approximately the same beds with which the Nantasket series ends. This view naturally leads us to regard the melaphyr of Hingham as probably equivalent to the great bed of melaphyr near the top of the Nantasket section ; and certainly the thick beds are far more likely than the thinner ones to extend over considerable areas. Although the great slate series, consisting of gray slate with- out intercalated sandstone and conglomerate, appears to cover a considerable area in the northwestern corner of Hingham, it is well exposed only on the shores of Huit's Cove and Beal's Cove. These exposures, however, are sufficient to show, first, that the slate is certainly conformable with and essentially a continuation of the conglomerate series ; and, secondly, tliat it probably overlies these coarser rocks. But the a])parent absence of fossils in the slate and the entire lack of outcrops connecting the Hingham beds with tlie fossiliferous Cambrian slates of Weymouth and Braintree leave us no certain clue to the geo- logical age of the Hingham strata, except what is afforded by their relations to the granitic rocks and the composition of the conglomerate. Fortunately, however, this evidence, which will be fully set forth in a later section, is sufficiently clear to permit us to say provisionally that, while the granite is certainly newer than tlie Paradoxides beds of Braintree, the conglomerate scries of Hingham, like that of Nantasket, and lience the over lying slate, must be newer than the granite. The history of the Hingham strata may, then, be outlined as follows : subsequently to the deposition and plication of 199 the Paradoxides slates, subsequently to tlic eruption (1ii()ii;j;]i them in succession of the granitic rocks, — diorite, granite, and t'elsite, — and subsequently to the extensive erosion which has so largely swept away these ancient sediments, began the pro- gressive subsidence, accompanied in its earlier stages by volcanic activity, during which were formed the conglomerate and interbedded lavas of the Nantasket area. Only the latest bed of lava poured out in this part of the basin is exposed in Hingham, west of Rocky Neck; but the cessation of the igneous outbursts was folio w^ed by the o-reat con "glomerate series of Hingham, with its mterstratified beds of sandstone and slate ; and the variable character of the strata is a plain indication that the physical conditions were far from uniform during this period. The conglomerate occurs mainly in beds from twenty to eighty feet thick alternating with, usually, thinner beds of gray sandstone and red or gray slate, for a total thickness of nearly one thousand feet. In fact, no other part of the Boston Basin affords such clear and abundant evi- dence of: (1) frequent changes from beach or shallow water deposits (conglomerate and sandstone) to those formed in deep and quiet water (slate), and vice versa; and (2) the oscillations of the earth's crust, upon which these changes usually depend. This intermittent conglomerate series appears to pass somewhat gradually and with perfect conformity up into the great slate series, which consists throughout of a fine, dark gray slate, and has an apparent thickness of at least one thousand feet, without any interstratified beds of coarser material. This is sufficient to prove that the oscillations of level attending the formation of the conglomerate series were followed by a profound and prolonged subsidence ; for during all the time when the slate was being slowly and quietly depos- ited, Hingham must have formed the floor of a comparatively deep ocean. But this tranquility could not last forever ; for the subterranean forces were slowly gathering strength, and the formation of the slate was undoubtedly terminated by the 200 advent of an epoch of severe compression of the earth's crust in this region, in consequence of which the sedimentary rocks were elevated to form dry lard, folded, faulted, and injected by dikes of diabase. During all the long ages since this geographical revolution, Hingham has been mainly, if not continuously, a land area ; and the slate, together with the underlying conglomerate, has sutFered enormous erosion. These rocks have thus been completely removed from large areas of granite which they once covered ; and they are preserved to us now only where they were most deeply folded or faulted down between the granitic masses, and thus protected from erosion. GENERAL STRUCTURE OF NORTHERN HINGHAM. The eastern shore of Hingham Harbor is not only the natural boundary line between the geological districts of Nantasket and northern Hingham, but it probably marks the position, as shown on the general map, of one of the great transverse faults of the South Shore ; and it certainly corresponds, as already explained, to a very important con- trast in geological structure, plication being as characteristic of the Hingliam area as liudting is of the Nantasket area. As the general map so clearly shows, the key to the structure of the volcanic and sedimentary rocks of Hingham is the oblong area of granite and felsite lying north of the railroad and Beal Street, and bearing the three drumlins of Bradley's, Squirrel, and Baker's Hills. The general position of this mass is unquestionably anticlinal. This is most obvious at the western extremity, where the melaphyr and the sedi- mentary strata curve around the granite and dip away from it on both sides. Southward from this point, between Beal Street and Beal's Cove, the structure is monoclinal ; and the ledges afford a nearly continuous section across the entire conss'lomerate series and a considerable thickness of the over- 201 lyini^ slate, the hitter undoubtedly marking the po.sitiijii of a synclinal axis ; but the south side of the syn(;line is prohal)ly cut off by the boundary fault, for we seem to pass abruptly from the slate to the granite. In the vicinity of Hockley Lane a transverse fault appears to separate this normal suc- cession of the strata from an inverted succession which extends thence eastward to Main Street or beyond. The melaphyr is now on the south side, overlying the conglomerate ; and these stratified rocks, although occupying a synclinal position between the granite on the north and south, are, we must suppose, bounded on both sides by important dislocations and terminated on the east by the great faidt along the east side of Hingham Harbor. Northwest from the western extremity of the granite axis, a very steep, narrow, and broken monocline separates the granite from the great trough holding the main body of slate. This faulted monocline is marked by a second band of melaphyr, which broadens toward the northeast, forming the large quad- rangular area of this rock east of Huit's Cove. This is the largest exposure of melaphyr in Hingham, and, although it appears to be bounded on all sides by downthrow faults, the quaquaversal dips of the bordering strata show that, in a lesser degree, it is essentially similar in its structural relations to the granitic area. On the west side, the upper bed of conglom- erate and the slate are seen to dip away from the melaphyr. On the north, the downthrow of the sedimentary rocks is suffi- cient to conceal the conglomerate, and the slate lies with conformable strike against the melaphyr. On the south, the narrow monocline separating this body of melaphyr from the granite broadens somewhat, until it reaches the fault at the northwest end of Sqiiirrel Hill, where it changes, perhaps abruptly, to a broad shallow syncline of melaphyr and con- glomerate on the south, separated by a strike fault from a gentle southerly m.onocline of conglomerate and sandstone on the north. These features probably extend east under Broad 202 Cove and Otis Hill ; and the monocline can be clearly traced still farther in the ledges of Melville Garden and in the three sedimentary islands of the harbor. Westward from the Garden, however, this monocline, of east-west strike and southerly dip, changes gradually but rapidly to a north-south strike and westerly dip, plunging down against the great mass of melaphyr already described. It is obvious from this sketch of the geological structure that Avhile, as previously stated, folds of various types are the dominant form of displacement and give character to the area, the flexures are profoundly modified by longitudinal and transverse faults. The correct interpretation of these main structure lines is evidently essential to the determination of the stratigraphic elements or the normal succession of the strata. Of the four sections accompanying the special maps, three are approximately complete, viz.: (1) the section south of the granite, through the village; (2) the section from Beal Street to Beal's Cove ; and (3) the section from Melville Garden west toward Huit's Cove. They agree in their main features and especially in showing repeated alternations of coarse and fine sediments. But a more careful comparison reveals the fact that they cannot be exactly correlated or synchronized ; and we are obliged to recognize, even in this limited area, important lateral changes in the character or thickness of individual strata ; sandstone at one point being represented by conglomerate or shale at another, and so on. The following table of the strata of Hingham has been compiled from the first and second sections referred to above, and, with the foregoing qualifica- tions, may be regarded as substantially correct. The individual beds are subject to constant variations in thickness ; and since the outcrops are unfavorable to exact measurements, the thick- nesses given are, as the round numbers indicate, sim])ly approximations. Some of the beds attain the maxinnim thickness in the one section and some in the other ; and hence the totals do not correspond to the actual sections, 203 Table of the IIin(/ham Strata. Granitic rocks (diorite, granite, and felsitc). 1. Conglomerate (basal). Thickness uncertaio. 2. Melaphyr 120-240 feet 3. Fine conglomerate and sandstone, alternating . . . 120-200 i. Gray slate 40-60 5. Conglomerate, sandstone, and slate, alternating . . . 100-170 6. Gray and red slate 90-130 7. Conglomerate 30- 50 8. Red slate 20-40 9. Conglomerate 40- 50 10. Red slate • 20-30 11. Conglomerate . . ' 75-100 12. Red slate 50-75 13. Sandstone and conglomerate, alteruatiug 200-300 905-1445 14. Gray slate 500 4- Dikes of diabase intersecting the bedded rocks, owing chiefly, it is jjrobable, to the less continuous outcrops, but partly, no doubt, to the fewer faults, are much less conspicuous in Hing- hani than in Nantasket. They probably agree with the Nantasket dikes in dating from the plication and faulting of the strata. The most important distinction is that between the great masses of coarsely crystalline diabase scores or hundreds of feet in breadth and very irregular in outline, and the ordinary, narrow, wall-like dikes of finely crystalline diabase. The latter, at least, belong chiefly to the east-west systems of Nantasket. No clear intersections have been observed ; and no dikes which could certainly be referred to the youngest or north-south system of Nantasket. LITHOLOGY. The rocks of Hinghani are, lithologically, so similar to those of Nantasket that a very brief treatment of this topic will suffice , the main purpose of these studies being to decipher the 204 structure and physical history of the region and not to under- take a minute investigation of the different kinds of rocks. The granitic rocks (diorite, granite, and felsite) have been described in sufficient detail in the preceding pages. Mr. Merrill has kindly examined thin sections from some of the more typical dikes, and finds them all to be diabase essentially similar to that of the Nantasket dikes. He has also found that the melaphyr is in essential agreement with the common basic variety of Nantasket. It has commonly a dark greenish color, due to chlorite and epidote, but limited portions are often brownish or purplish owing to the local peroxidation of the iron. It varies in texture from almost perfectly compact or even slaty to distinctly and coarsely amygdaloidal. The amygdules con- sist chiefly of quartz and epidote, but include also chlorite, feldspar, and calcite. The quartz is usually crystalline, but sometimes chalcedonic or jaspery. These secondary minerals also occur commonly in veinlets and irreo:ular seo'res'ations : but the brecciation so characteristic of the great bed of melaphyr near the top of the Nantasket series is rarely distinctly observed in Hingham ; although the scoriaceous structure which one naturally looks for in the superficial portions of a flow is plain enough at some points, as in the mass of melaphyr near the junction of Downer Avenue and Crow Point Lane. The Hingham bed, like the Nantasket bed just referred to, is probably composite, a succession of flows, but the only observed facts, besides its great thickness, Avhich clearly point to this conclusion, are the intercalated bed of sandstone and conglom- erate which is exposed south of Beal Street, about midway between the street and the little pond (PI. 8) ; and a small amount of banded slate or possibly tufi^ which a recent excavation has exposed in the melaphyr immediately west of West Hingham Station. The occurrence last mentioned is the only thing resembling volcanic tuff*, or in .any way suggestive of ex])losive eruptions, which has been observed in Hingham. The conglomerate of Hingham is composed for the most 205 part of small and well-rounded pebbles, chiefly of difi^'crcnl. varieties of felsite and granite. It is a typical puddingstoiie, becoming a breccia only where the basal conglomerate rests directly upon the granite. Being less intimately associated with melaphyr than the Nantasket conglomerate, it is, as a whole, less indurated, possessing less, probably, of a distinctly siliceous cement. This is especially noticeable, also, in the arenaceous layers, which in Nantasket have the flinty hardness of red quartzite, but in Hingham are more like normal sand- stone. The well-rounded and assorted pebbles are usually less than an inch and rarely more than two or three inches in diameter. But at one locality, on the eastern shore of Huit's Cove, the conglomerate is exceptionally coarse and irregular in structure, containing rounded pebbles or bowlders of granite, etc., of all sizes up to a yard in diameter. It is interesting to note also that some of the pebbles at this point are an impure gray limestone, and that limited portions of the rock have a distinctly calcai^eous cement. The finer portions of the con- glomerate frequently become gradually but distinctl}^ arenaceous ; and most of the beds show repeated alternations of true conglomerate with coarse, pebbly sandstone, so that it is quite impossible to map the two rocks separately. Although the conglomerate series is so largely arenaceous, there is compara- tively little pure or typical sandstone ; the most important occurrence of this kind being the bed of gray or brownish gray sandstone over a hundred feet thick in the Hersey Street section. The sandstone is usually gi'ay to light brown in color, rarely distinctly ferruginous, and, as stated, rarely a good quartzite. The brown color is probably due more to the admixture of grains of red felsite than to a ferruginous cement. The argillaceous rocks of Hingham are quite varied. The more limited beds of slate included in the conglomerate series are usually of some shade of red, brown, or purple ; but the thicker beds are in large part of a greenish or a greenish gray color, the iron, evidently, being less highly oxidized. It is difficult, in the 206 -conglomerate series, to draw the line between slate and sand- stone, and part of what has been mapped as slate might be otherwise classified. The great slate series above the conglom- erate consists throughout of a homogeneous gray slate, a hard, firm, but not strictly impalpable rock. It is distinctly siliceous in composition, and a fine granular structure is often apparent under a lens. Although usually more or less distinctly banded, the stratification being marked by laminae of alternating colors, the Hinffham slates, whether in the conolomerate series or in the slate series, are rarely shaly or exhibit a lamination cleavage ; and only to a limited extent, as in the vicinity of Huit's Cove, is the true slaty cleavage, transverse to the bedding, well developed. A very perfect cuboidal or rhomboidal joint- structure can. be seen in many exposures. DETAILED STRUCTURE OF NORTHERN HINGHAM. We are now ready for a systematic, ledge-to-ledge study of the bedded rocks of Hingham, following, as at Nantasket, the topographic order of the outcrops. The village area, bordering the railroad, south of the granite axis, is in most respects a convenient and natural starting point for a detailed examina- tion of the sedimentary and volcanic rocks. The section between Bcal Street and Beal's Cove is more complete, and presents a normal rather than an inverted succession of the strata, but the ledges are less continuous, and the beds are much less easily traced along the strike. Moreover, it is a wooded and swampy tract, while the village area is elevated and open, and readily accessible, in spite of the numerous houses and fences. The Village Area. This is the area represented on the first of the special maps (PI. 7), embracing all the ledges between Main Street 207 and Hockley Lane. Topographically it is essentially ouc. con- tinuous ridge, save where it is crossed by Town Urook :im(1 ihc railroad at West Ilinghanj Station. The most contiiuious exposures of this narrow belt are in the vicinity of Hereey Street ; and the dip in this part of the area, especially, is quite constant — S. 70°. The most northei'ly outcrop along this line is the ledge of conglomerate (13)^ about 150 feet west of Hersey Street, in the rear of the second and third houses from the corner, on South Street. A few yards south of this ledge, in the rear end of these lots and extending into the adjoining lot on the south, is an outcrop of dark red slate (12). The slate must cross Hersey Street at the first bend ; and it is exposed repeatedly along the base of the con- glomerate escarpment from 200 to 300 feet east of the street. Going up Hersey Street from the railroad, the first rock actually seen is the conglomerate (11) bordering the red slate just referred to on the south. It commences a hundred feet or so west of the street ; and its northern border forms a con- tinuous escarpment from 15 to 20 feet high, due to the erosion of the red slate and extending about 500 feet east of the street. This escarpment is the northern edge of an area from 600 to 1,000 feet in length (E.— W.) and 500 feet in breadth, over whicli the ledges, as the map shows, are almost continuous, and the thickness of the several beds admits of accurate measurement. The conglomerate just described has a breadth of from 75 to 90 feet ; and it is followed on the south in succession by about 20 feet of red slate (10) ; from 40 to 50 feet of conglomerate (9) ; 25 feet of red slate (8) ; from 35 to 45 feet of conglomerate (7) ; from 115 to 135 feet of red and gray slate (6) ; from 100 to 130 feet of gray sandstone with -some gray slate (5) ; and 40 feet of conglomerate (5). This Hersey Street section is broken by several large and irregular dikes (see PL 7) ; but the breadth of the trap is 1 The numbers in parentheses refer to the general table of Hingham strata on page 202. 208 not included in tlie foregoing measurements of the beds Avbicli it intersects. Farther south on the line of Hersey Street the sandplain conceals everything for nearly 500 feet ; and the first outcrops in that direction are granite and diorite, with other large masses of diabase. Most of the beds which we have crossed on Hersey Street can be traced east by satisfactory outcrops to Lafayette Avenue. The strike gradually changes in this direction, however, from nearly due east-west to east-southeast, as the map shows. Of the most northerly conglomerate (13) there is only one small and rather uncertain exposure nearly GOO feet from Hersey Street. The first red slate (12), after a gap of 700 feet, is well exposed behind Mr. Lane's house on South Street. It forms an abrupt escarpment or cliflffrom 15 to 20 feet high, and has also been found in excavations 50 feet or more north of the cliff. The dip at this point is only 25°, which fully accounts for the increased breadth of the bed as shown on the map. South of this slate we are able to identify in almost continuous ledges the following beds: conglomerate (11); red slate (10) , not clearly exposed, but represented by a blank depressed space of the proper width; conglomerate (9) ; red slate (8) ; conglomerate (7) ; red and gray slate (6) ; and gray sand- stone and slate (5). The more southerly beds along this line have approximately the same dip as on Hersey Street — S. 70° ; but toward the north the dip diminishes to 50° and 25°. The first slate (12) and the first and second conglomerates (11 and 9) are well exposed also in the angle east and south of Lafayette Avenue. The structure here is much more compli- cated and interesting. Immediately in the angle of the avenue the slate is in contact with the conglomerate (11) as usual; but 100 feet farther east it rests in like manner against a parallel mass of granite ; while the conglomerate abuts squarely against the granite and the great dike of diabase from 50 to 70 feet wide which interru})ts the granite at this point. The granite, which has a breadth of from 40 to 50 feet north of the 200 trap, is coarsely crystalline, and there is not the slip^htest indica- tit)n tliat it is eruptive through the sedimentary rocks. No veins or apophyses of granite penetrate the slate, and the normal coarse texture of tlie granite is unchanged near the contact. But the abrupt way in which the diabase ends against the conglomerate and the obvious partial dislocation or lateral shifting of tlie slate are certainly very suggestive of faulting. Again, the slate, although its strike is parallel with the granite, clearly dips against the latter : and in all the outcrops it shows gi'cat disturbance, being strongly contorted and faulted, or, locally, completely crushed. The contact between the granite and slate is probably a strike-fault with the downthrow to the north ; and all the sedimentary beds are probably cut by a transverse fault along the west side of the granite, which, as the arrows on the map indicate, downthrows to the west south of the granite-slate contact and to the east north of it. Along its contact with the slate the granite continues scarcely 200 feet, ending as abruptly as it began. In the direct line of the o^ranite we here find broad outcrops of the slate, which, with a southerly dip of 80°, is exposed almost continuously for its normal breadth ; while the dike of diabase, as before, abuts squarely against the conglomerate. Evidently, the granite and trap are bounded on the east as well as on the west by a fault ; and these two faults are obviously reversed in throw or compensating. To avoid their indefinite extension, they are represented as con- verging northward ; but that is not strictly required by the observed facts. The main point, however, is that the facts appear to justify us in regarding the granite and its included diabase as a somewhat rectangular block which has been ele- vated relatively to the bordering strata, the uplift having been sufficient to carry the sediments normally overlying this mass of granite above the present plane of erosion. Southward this block of granite is continuous with the main area ; and it is simply a displaced portion of the general granitic floor upon which the sediments rest. The evidence for the second fault is OCCAS. TAPERS B. S. N. H. IV. 14. 210 materially sti'engtheiied by the marked change in the strike, from south of east to north of east, as we cross it. In following the sedimentary rocks farther east we find an outcrop of conglomerate in front of the school house on Elm Street, nearly opposite the end of Central Street, and directly in the line of strike of the slate ; while several exposures of the red slate are found in the field north of the school house. These facts clearly suggest the third transverse fault shown on the map. The character of this fault is wholly uncertain. No outcrops of slate or conglomerate have been observed south of Elm Street ; and the map is here largely hypo- thetical. Quite certainly the fault does not downthrow to the west as the arrow indicates ; but it is more probably a hori- zontal thrust-fault, the eastei-n extension of the strata having been shoved bodily to the northward from 100 to 150 feet. The rock showing in the field south of Elm Street and east of Central Street is diorite and probably iyi situ. This indicates a gradual narrowing of the sedimentary belt ; but beyond this point there are no outcrops of any kind for more than half a mile ; and where and how the sedimentary rocks termi- nate we can only conjecture. As previously stated, the gen- eral map is constructed in accordance with the view that they cross Main Street, pass beneath the sandplain occupied by the cemetery, and end in the Home Meadows against the great boundary fault between the Hingham and Nantasket areas, supposed to coincide in position and direction with the east shore of Hingham Harbor. Returning to Hersey Street and tracing this series of strata westward, we find that all the beds which can be traced more than 200 feet from the street are distinctly flexed to the north, the flexure amounting to a horizontal shift of about 100 feet. It is necessary to suppose, of course, that, as shown on the map, the entire series is involved in the displacement ; and there are some indications that the beds are actually com- pressed or pinched on the bend, as if they had experienced a 211 transverse thrust or a tendency t() shear. But, unfortunately, there is at this point a complete hiatus of nearly /}()() feet in the outcrops ; and the full extent and form of the fold can not be made out. That the east-west strike is quickly resumed is shown, however, by the ledges between this and West Hingham Station ; and it is especially satisfactory to find that the very first exposures show the thick bed of gray sandstone, the most unique bed of the entire sedimentary series, and therefore the most to be relied upon in correlation. About 100 feet back of the school house and 400 feet from Soutii Street, an artificial excavation has uncovered the abrupt face of a ledge which shows 75 feet of greenish gray sandstone (5) with distinct slaty partings, dipping S. 70° beneath 40 feet of conglomerate (5) ; while fragments in the soil indicate that the conglomerate is overlain in turn by slate (4). The conglomerate and sandstone are easily traced to the westward several hundred feet, where they form the northern portion of a group of ledges affording the following nearly continuous section, beginning on the south : — granite, in several obscure outcrops ; then, after an interval of a hundred feet or more, with one uncertain exposure of melaphyr, come without an appreciable break, 120 feet of melaphyr (2) ; 120 feet of fine conglomerate and sandstone (3) ; 50 feet of slate, mostly gray, but changing to red on the north (4), the red rock being well exposed behind one of the houses on South Street; 40 feet of conglomerate (5) and nearly 100 feet of gray slaty sandstone (5). The isolated and obscure outcrop of gray slate a few yards from the line of South Street (see map) undoubtedly marks the extension of the great bed (6) of the Hersey Street section ; and on the north side of the street, in the rear of the second house from West Hingham Station, we have a satisfactory exposure of the conglomerate ( 7 ) which borders this slate on the north. The only remaining outcrop in this direction lies some 300 feet farther to the northwest, beyond the railroad and on the north side of the brook. It 212 is a mass of conglomerate about 10 feet across, lying in the meadow, and is possibly only a bowlder. Assuming it to be in situ, or nearly so, it probably represents the most north- erly conglomerate (13) of the Hersey Street section. The section afforded by this group of ledges, it will be observed, overlaps on the south and supplements the section along Hersey Street, the two together affording a nearly com- plete section of the conglomerate series. The melaphyr of this section is the typical variety, greenish and purplish in color, and compact to highly amygdaloidal and scoriaceous in texture. The amygdules consist chiefly of epidote and quartz, and are usually small, or, if larger, rather scattering and not crowded. As a whole the rock is very massive, but portions of the bed show irregular veinlets and segregations of epidote and fer- ruginous quartz ; and the distribution of the amygdaloidal and scoriaceous melaphyr is such as to suggest that this may be a composite flow. The contact between the melaphyr and the arenaceous conglomerate is very clearly exposed for about ten feet on the highest part of the ridge. It is straight, exactly parallel with the strike of the conglomerate, and shows only minute irregularities, variations of an inch or so from the mean line. The melaphyr does not penetrate the conglomerate any more distinctly than the conglomerate penetrates the melaphyr. Some irregular cracks in the lava appear to have been filled with fine sand, which is now highly ferruginous ; but aside from these there are no distinct inclusions of the sedimentary rocks in the melaphyr. On the other hand, a few small pebbles of melaphyr similar to this were observed in the conglomerate, within a few inches of the contact. The conglomerate and sandstone exhibit no special alteration or unusual induration at this point ; and in every way the facts are favorable to the view that the melaphyr is contemporaneous and not intrusive. Between the ledges just described and those next to the west there is a gap of six hundred feet, occupied by the valley of Town Brook and the railroad. But innnediately l)eyond West 213 Hingham Station, the melaj)liyr outcrops proiiiiticntly, toriuing the summit and abrupt southern slope of a hill which is flanked by modified drift on the north. It is essentially similar in character to the melaphyr east of the railroad, except that a larger proportion of the rock is compact or not distinctly amygdaloidal, some of the exposures resembling a massive slate. Irregular segregations of epidote, etc., are common; and, as already stated, the melaphyr encloses at one point a broken or faulted layer a few inches in thickness of a beautifully laminated or banded slate. This is well exposed at the present time in the part of the ledge nearest to the station, where a new street is being graded up the hill from West Street. So far as can be determined, it is near the middle of the bed of melaphyr ; and it is certainly far from any large body of slate. The most natural explanation appears to be that it is a thin layer of tuff, similar to some of the tuff beds of Nantasket, and hence another indication that the bed of melaphyr is composite, consisting of two or more flows. The melaphyr is exposed along the strike (W. by N.) in frequent ledges for about 1,000 feet, or to within 500 feet of Hockley Lane, and is of similar character throughout. On the north side of the first hill, in the angle between the melaphyr and West Street, are several outcrops of conglomerate ; but the contact is not exposed here. North and west from this point the stratified rocks are almost entirely concealed by the undulating modified drift. I have recently discovered, how- ever, that the small ledges of melaphyr on the west slope of the main hill, lying between the two *large groups of ledges and divided by the east-west fence, as shown on the map, are bordered on the north side by fine conglomerate and sandstone precisely similar to the rocks in contact with the melaphyr east of the railroad ; and the contact, which is clearly exposed, is of the same character. It is very evident that the sedimentary rocks were deposited over the melaphyr, for they fill cracks in it and are partly made up of debris derived from it. Near the 214 melaphyr the fragments of that rock are large and angular ; but the conglomerate as a whole is composed chiefly of felsite and granite, and cannot be classed as a tuff. The dip is as usual about S. 70°. Irregular veinlets of iron oxide are seen in both the sandstone and the melaphyr ; and cannot be regarded as evidence that the melaphyr is intrusive. The northern border of the melaphyr, as drawn on the map, is shown by these recent observations to be too far to tlie north at this point ; and it is possible that a small transverse flexure or fault separates this contact ledge from the remaining outcrops of melaphyr in the next field to the west. In this field it is quite noticeable that the melaphyr is more compact toward the granite on the south or the supposed base of the flow, and more amygdaloidal and scoriaceous toward the sedimentary rocks on the north or the supposed top of the flow. Going north from these ledges into the adjoining field we find, after an interval of nearly 75 feet, or 15 feet beyond the fence, 60 feet of fine conglomerate and sandstone with an apparent dip S. 45° ; 50 feet concealed ; and 10 feet of sandstone. Tlicse exposures, which are marked on tlic map, must evidently be referred to the same bed (3) which we have elsewhere found in contact with the melaphyr ; and more recently I have discovered north of these, small and obscure exposures of purple slate (4), dipping S. 70°; and conglomerate (5). These scanty outcrops serve to show that the Village series of strata probably extends without essential change this far to the west. We come now to the interesting group of conglomerate ledges in the field south of the melaphyr and Hockley Lane, and bordered on the east and south by granite (see the map). The relations of the granite and conglomerate are very inti- mate, and could be fully known only by removing all the superficial detritus from this area. Along the south side of the field, especially, and on both sides of the fence, we pass repeatedly and abruptly from the one rock to the other ; and there are probably several or many l)0sscs or knobs of granite 215 projectinii^ throui2^h a thin and approximately horizontal bed of conglomerate ; although the surface relations of tlu; two rocks might be partly explained by faulting. That the granite is not younger than and eruptive through the conglomerate is proved beyond the shadow of a doubt by the nature of the contact and the composition of the conglomerate. The latter is made up very largely, especially where it lies directly upon or against the granite, of the angular and half rounded debris of exactly the same coarsely crystalline red granite ; and the fragments, which range from single grains of quartz and feldspar to masses two feet or more in diameter, are sometimes only imper- fectly separated from the parent ledge, the finer sediment appearing to penetrate cracks in the granite. The relations are essentially the same as on Rocky jSTeck and Granite Plateau of the Nantasket area, where the basal conglomerate rests upon the granite ; and they also agree with the contacts resulting from the rapid deposition of coarse sediments over a disinte- grated land-surface, as recently described by Professor Pumpellyi and Professor Emerson.^ On purely structural grounds, this conglomerate should certainly be correlated with the basal conglomerate of Nantasket ; which would, apparently, separate it widely from all the other Hingham strata. The melaphyr marked on the map as outcropping in the midst of this basal conglomerate, probably overlies it. It is amygda- loidal and otherwise similar to the rest of the Hingham mela- phyr, but may, of course, be the equivalent of any of the earlier basic flows of Nantasket. This completes the detailed examination of the ledges of the Village area. If, as appears necessary, we regard the series as inverted, the general structure and the relations to the granite must be approximately as represented in the section accompanying the map. A long, narrow block of strata is faulted down between walls of granite, the drop on the north > Bull. geol. soc. America, vol. 2, p. 209-224. * Bull. geol. soc. America, vol. 2, p. 451-456. 216 being so much greater than that on the south as, in conjunction with a horizontal or plicating stress, to overturn the beds. Or it may be otherwise described as an inverted syncline deeply folded down in the granite and the northern half carried away by the fault which elevated the granitic axis on the north. The map shows two areas of slate along the northern border of the conglomerate series. These are not based upon any out- crops ; but the topography is favorable to the occurrence of slate here, and they are introduced in accordance with the foregoing explanation of the structure. The same is the case also with the conglomerate shown south of the melaphyr, east of the railroad ; but the conglomerate which is represented in the section as lying upon the granite is the basal conglomerate already described south of the melaphyr, near Hockley Lane. The Beal's Cove Area. This is the area indicated by the second special map (Plate 8) . It is crossed diagonally by Beal Street and embraces all the ledges between Beal's Cove and Lincoln Street. Topographically it belongs chiefly to the broad lower or lowest sandplain, but includes Tucker's Swamp and the western slopes of Baker's and Squirrel Hills. This area has been already described as affording, between the granite on Beal Street and Beal's Cove, the most complete and normal section of the Hingham strata. The western extremity of the granite axis is, fortunately, well defined, as the map shows, by numerous outcrops north of and in the street and the obscure but reliable .outcrops between the houses south of the street. The granite is over- lain at this point, as already described, by several irregular patches of effusive felsite, including the mass wrongly marked as an outlier of conglomerate. One proof that not only this felsite but also that farther east, in the vicinity of Thaxter and Lincoln Streets, are part of a surface flow or truly effusive has not been stated; viz., that it occurs only near the junction of 217 the gnuiitc and tlic bedded rocks, wlucli, it iw so evident, were deposited ii|)ou the granite. In otlier words, we find the felsite ii[)on what we know to liave been tlie ancient snrface of" the granite, from which the sedimentary deposits have recently Ijeen denuded ; and v\^]ierever erosion has cut below this surface the effusive felsite is wholly wanting and we observe only an occasional narrow and irregular dike of intrusive felsite. Along the northwest side of the granite there is a line of cono'lorncrate ledt»:es sufficient to indicate a narrow but continu- ous or nearly continuous band of conglomerate separating the granite from the overlying melaphyr. This conglomerate undoubtedly extends considerabl}^ farther than it is marked, the more northerly outcrops having escaped observation at the time the map was drawn. One of these is on the west side of the winding road known as Hawke's Lane running south from Lincoln Street to an isolated house ; and another, which will be referred to later, is on the south side of Lincoln Street west of the lane. This conglomerate is a true puddingstone, con- sisting of rounded fragments of granite, felsite, and melaphyr : but the contact with the granite is not clearly exposed, and we can only conjecture that the relations of the two rocks are, perhaps, the same as south of Hockley Lane. This supposed basal conglomerate can not be traced around the extremity of the granite axis, and the narrow belt of it marked along the southern margin of the granite is unsupported by a single out- crop, finding its justification simply in the depression between the granite and melaphyr ; while the fact that outliers of melaphyr rest upon the granite north of Beal Street is rather against its existence. In the two sections accompanying the map, it is omitted from the one extending south from Beal Street to Beal's Cove ; although represented, as the facts justify, in the other, running northwest from the granite. South of the granite, the main bed of melaphyr is well exposed in two large groups of ledges. It is the typical green, basic variety, and is abundantly but coarsely and irregularly 218 amygdaloidal, with many irregular veinlets and segregations of impure epidote, etc., portions of the rock being of a distinctly scoriaceous character. A coarse flow-structure is sometimes indicated by tlie arrangement of the amygdules in parallel sheets or layers, which have the normal dip of the section, S. 60° or more. As the map shows, the breadth of the belt varies, perhaps, with the dip, from about 160 to more than 300 feet. In the western group of ledges, a little north of the middle of the belt, a thin layer of conglomerate, passing upward into sandstone, the entire thickness being, possibly, less than ten feet, interrupts the continuity of the melaphyr. This is the normal reddish conglomerate, consisting chiefly of granite and felsite ; and it can not be classed as a tufl" or a fragmental melaphyr. The dip of the conglomerate is S. 60° ; and there is no reason to doubt that it is essentially conformable with the melaphyr above and below. But since neither contact is clearly exposed, the facts are evidently insuflicieut to reveal the true significance of the conglomerate. We can not deter- mine conclusively whether it is a true interbedded conglomerate, or simply an outlier of the very similar overlying conglomerate, displaced by faulting. The latter view would readily explain the unusual breadth of the melaphyr at this point ; and it may be added that, independently of this conglomerate, there is no special indication that the melaphyr is a composite flow. In the eastern group of ledges, in the angle between Beal Street and Portuguese Lane, the contact of the melaphyr and overlying conglomerate is satisfactorily exposed, and is decidedly favorable to the view that the melaphyr is contemporaneous, i. e., that the conglomerate has been deposited upon its surface. The melaphyr is here very coarse and scoriaceous ; and the numerous cracks and inequalities of its surface arc filled with the finer pai't of the conglomerate, in such a way as to jilace the relations of the two rocks beyond question. The conglom- erate contains, however, comparatively little debris that can be referred to this melaphyr ; and this fact, together with the 210 roiin^h cliaracter of the inelaphyr, points to the conclusion that it sidi'ered but little erosion before tlie congloinerute was spread over it, and hence that it is probably a submarine flow. The conglomerate dips south or away from the melaphyr at the normal angle ; and these eastern ledges overlap and su[)ple- nicnt the extensive and prominent outcrops farther west, in the vicinity of the small pond, the entire breadth of the bed being thus satisfactorily exposed. This bed is throughout a normal, massive, well-indurated puddingstone, embracing very little arenaceous material. The pebbles are mainly from one to two or three inches in diameter : but in certain layers the maximum rises to six inches or even a foot. The southerly displacement of the conglomerate in passing from the eastern to the western ledges is very noticeable, and evidently related to the increased breadth of the melaphyr in that dii'ection. Two explanations have been suggested for the latter: first, a diminished dip, which would cause a curvature of the overlying strata, as shown, on the map, even though they retain throughout their normal high inclination ; secondly, a strike-fault, as indicated by the conglomerate enclosed in .the melaphyr, which miglit have the same result as regards the overlying beds, or, ending against transverse faults, might involve an actual dislocation of the sedimentary series. Unfortunately, the entire absence of out- crops along this north-south line makes it impossible to choose between these explanations ; and it may well be that both of them are true. With this introduction, we can pass rapidly over the entire section from the granite southward, since it is, in a large measure, a repetition of the Village section. The strike ranges from S. 65° to S. 80° E. ; and the dip, except perhaps with the melaphyr, varies but little from S. 80°. The thicknesses are chiefly approximations, the beds, either on account of not being sharply defined or from lack of cpntinuity in the outcrops, rarely admitting of exact measurement. The numbers refer to the general table of Hingham strata and also indicate the corre- 220 . lation of this section with the Village section, although the sequence here is normal and not reversed, as follows : — Granite and felsite, with outliers of melaphyr. Conglomerate (1), not observed. Melaphyr (2), 140 feet (minimum). Conglomerate (3), medium to coarse puddingstone, very much coarser and thicker than in the Village section, 175 feet. Slate (4), not exposed, but i-epresented by a depression, 65 (?) feet. Fine conglomerate alternating repeatedly witli sandstone (5) , 170 feet. The broad outcrop south and east of Portuguese Lane shows that this is a complete mixture or blending of coarse sandstone and fine conglomerate ; while in the Village section there are two distinct beds — 40 feet of conglomerate without sandstone and 100 feet of sandstone without conglome- rate. Purplish and gray slate (6), probably 100 feet. Concealed by kames and swamp ; and only one outcrop, east of Portuguese Lane. Conglomerate (7), 35 feet. A single ledge on the west side of the lane, passing south into the solitary outcrop of the next bed. Red slate (8), 40 feet. This crosses the lane just north of the point where it divides ; and a few yards east of the junction is the principal exposure of the next stratum. Fine conglomerate and sandstone (9), 50 feet. Red slate (10), not exposed, but there is room between the nearest outcrops of conglomerates 9 and 11 for 25 or 30 feet of slate. Conglomerate (11) , fine but without much sandstone, about 80 feet ; forms extensive ledges east of the lane. Red slate (12), about 50 feet ; several good outcrops east of the lane. Fine conglomerate and sandstone (13), a broad belt of uncertain width. If correctly mapped, the thickness must be about 300 feet. The more prominent ledges are chiefly fine conglomerate ; but there is undoubtedly a large proportion of sandstone. Outcrops, it will be observed, are almost wholly wanting in the southern half of this belt ; and the southern limit is quite uncertain. Gray slate (14), of un- known thickness. The exposures begin near the north shore of Beal's Cove ; and along the sliorc of the river the outcrop is con- 221 tinuous from the end of Portuguese Lane south to the point, the total exposed thickness being, ])robal)ly, not less than 500 feet. The slate is of a very uniform character, — gray, rather coarse, very thinly and evenly bedded, well jointed, but without marked cleavage. Some layers are finely pitted by weathering and wave-action, indicating, possibly, some lime in the rock ; and there are some indistinct concretions. But although the section appears in eveiy way favorable for the occurrence of fossils, I have searched for them in vain. In the most northerly outcrop of the slate it is distinctly and repeatedly interstratified with layers of hard, gniy, and rather coarse sandstone, from two inches to two feet thick. Since such intercalations of coarse material are wholly wanting in the remaining 500 feet of the slate, this outcrop is regarded as marking the transition from the underlying conglomerate and sandstone (13) to the slate ; and it is the chief fact relied upon in drawing the boundary on the map. It is unnecessar}^ to dwell upon its importance as evidence that the slate conformably overlies the conglomerate series. It may well be, however, that, as in the case of the conglomerate in the Huit's Cove section, to be described in the following pages, this intercalated coarse material is underlain by a considerable thickness — 100 feet or more — of slate. It is impossible to carry the section farther south ; for the east and south shores of Beal's Cove are wholly composed of modified drift which rises from 30 to 80 feet above the water, and extends southeast across the Hockley district to the rail- road, blotting out everything but an occasional ledge of granite. Following the east shore of Weymouth Back River south from Beal's Cove, no ledges of any kind are observed for about 800 feet from the south angle of the cove, and then, when nearly opposite the south side of Whale Island, we reach the beginning of an outcrop of coarse granite which is almost continuous alongc the shore for 500 feet. The granite is divided by three large east-west dikes ; but there is not a trace of slate or any sedimentary rock. Between the nearest 222 outcrops of slate and granite there is thus a gap of fully one fourth of a mile, and concerning the exact position and the character of the contact we know simply nothing. The stratigraphic continuity of the section from the granite north of Beal Street across the melaphyr and the conglomerate series to Beal's Cove is unquestionable , and, although the relations of the Beal's Cove slate to the conglomerate series are not so fully and clearly exposed as could be desired, the perfect agreement in strike and dip and the indications of a gradual transition afforded by the interstratification of sand- stone and slate, appear to justify us, as already stated, in regarding it as a part of the same conformable sequence of strata. If, as this view requires, the slate conformably over- lies the conglomerate series, Beal's Cove must mark the position of a synclinal axis ; and, since the slate must, in this case, be much newer than the granite, the south side of the syncline is probably cut away by a fault with the downthrow to the north, for there is evidently not room enough south of the axis for a repetition of the conglomerate series and melaphyr. Between the granite north of Beal Street, which still holds its normal relation to the bedded rocks deposited upon it, and the granite against which they end, as the result of faulting, on the south, we have, then, a steep monocline (one half of a syncline) and, as previously stated, one complete section of the Hinghara strata. Eastward from Beal's Cove and south of Tucker's Swamp, there are, fortunately, sufficient outcrops to connect the section just described with the ledges in the vicinity of Hockley Lane. The single small outcrop of the true slate (14) east of the cove is much nearer the water than represented, and evidently below the highest bed on the west side of the cove. About 1,000 feet east of the salt marsh bordering the northern arm of the cove, and immediately beyond where the boundary of Tucker's Swamp bends to the north, we come to a con- siderable exposure of the newest conglomerate and sandstone 900 (13), with a large but imperfectly exposed dike of diabase, while farther east we are able to identify slates 12 and 10 and conglomerates 11, 9, and 7 ; and north of the swamp, apparently conglomerates 5 and 3, in very obscure outcrops. Several of these outcrops are not marked on the map, and others are incorrectly marked. It is evident, however, that some of these beds, if they continue, must abut against the end of the raelaphja- of the Village section. This is suggestive of a transverse fault. Farther north the outcrops are so imperfect on both sides of the line that a lack of correspondence or continuity can not be easily proved. Farther south alono- the supposed fault line, however, we find evidence which, so far as it goes, is entirely satisfactory. This is afforded by the slate which is mapped as forming a short northwest and southeast line of outcrops south of the end of Hockley Lane. This slate is the purplish variety, with occasional thin streaks of sandstone. It strikes in the direction named and dips N. E. 70°. In the direction of the strike the slate is exposed very near the granite, and it is evident that it is cut off by the granitic rocks on both the east and south. The most careful search fails to reveal any evidence whatever that the granite is intrusive in the slate ; but the most satisfoctory explanation of their relations is that proposed on the map. This slate has been somewhat doubtfully referred to the great bed (14) ; and, judging from its dip, it belongs on the south side of the synclinal axis, which implies a shallowing of the trough in this direction. The marked southeasterly strike may be regarded as in some way the result of the transverse dis- placement. Having now brought the strata of the Beal's Cove area face to face with those of the Village area, it is readily apparent that they can be correlated only by completely reversing the one series or the other. This is most obvious in the case of the melaphyr, which is on the south in the Village area and on the extreme north in the Beal's Cove area. The entirely 224 normal character of the Beal's Cov^e section leaves no room to doubt that, as previously stated, it is the Village section which requires reversing. This transverse fault is, then, the one important dividing line for the entire district between the Home Meadows and Beal's Cove. West of this line we have a steep but normal monocline, terminated against the granite on the south by faulting ; while east of it a great displacement along the northern border of the bedded rocks, accompanied by a severe plicating strain, has completely inverted and dislocated this end of the series. It is interesting to observe, also, that this change or break occurs opposite the southwest angle of the granite axis on the north, showing that this mass of granite is, as previously pointed out, a dominant or controlling factor in the structure of this region. The foregoing outline of tlie structure of this narrow southern trough will, perhaps, be more readily or fully comprehended if its various phases are pre- sented in succession, as follows : From the Home Meadows to Beal's Cove there was originally, or would have been but for the faulting, one continuous syncline. This is broken transversely by the Hockley Lane fault. The western half remains an open syncline : but the greater part of its southern slope is carried away by a strike-fault, which brings up the underlying granite in that direction. The eastern half, owing to the stronger compression, being in the narrowest part of the granite vise, becomes an inverted isocline with the axial plane dipping to the south ; and its northern side is partly carried away and partly concealed by a strike-fixult, bringing up the ofranite axis wliich these strata once covered. In attempting to trace the various strata of the Beal's Cove section westward, we find them passing at once beneath almost continuous deposits of modified drift, including several high kames. The slate series and the more southern members of the conglomerate series are thus hopelessly cut off. But the nielaphyr and tl)e strata immediately above it emerge sufficiently from the sea of drift so that their relations to the extremity of 225 the granite axis can be obacrvo(^. It is very [)lain that tlio.se beds curve regularly around the granite, following its north- western as faithfully as its southern margin. The broadening of each bed as it rounds the angle is due to the natuial dim- inution of the dip at this point. North of Beal Street the belt of melaphyr is, at first, apparently, scarcely 100 feet wide, l)ut it gradually broadens northward, possibly as the residt of a diminished dip. It is seen in contact with the basal conglom- erate at several points. The most northerly and most satisfactory exposure of the contact is' in a small excavation on the south- west side of Hawke's Lane, where the conglomerate is not marked on the map ; but in every case the appearances are best explained by regarding the melaphyr as contemporaneous rather than intrusive. It fills the inequalities in the surface of the conglomerate ; but does not properly penetrate that rock. Although the great bed of conglomerate (3) overlying the melaphyr is well exposed for the entire distance between Beal Street and Lincoln Street, not a single good contact could be found. Scattered through the woods and swamp, stratigraphically above this conglomerate, are numerous outcrops of conglom- erate, sandstone, and slate which it is difficult to connect satisfactorily in continuous belts. These are somewhat generalized on the map, and the correlation indicated there is probably not entirely correct. It can hardly be doubted, however, that we have here, in normal sequence, the red slate (4), conglomerate and sandstone (5), and red slate (G) ; and then follow in succession conglomerate, slate, conglomerate, melaphyr, and conglomerate, as shown on the map. South of Beal Street, however, and disregarding this melaphyr, it would not be difficult to divide the nearly continuous line of outcrops so as to identify or represent every bed in the Beal's Cove section up to the highest conglomerate (13), The interpreta- tion of this part of the Beal's Cove area is one of the puzzles of Hingham geology. The general structure appears to be OCCAS. PAPERS B. S. N. H. IV. 15 226 synclinal ; and the map and section are constructed in accord- ance with this view. The dip, for the most part, varies but slightly from 90° ; but still, the advocate of the synclinal theory can find some ground for its support in the attitude of the strata. It should be pointed out, however, that the dip of the ledo-e of slate on the north side of Beal Street is W. 85°, and not E. 85°, as marked ; while the ledges southwest of this point and east of the Alms House are vertical. According to the map and section the western strata are a repetition of those on the east side of the belt ; and the syncline, although closely appressed, is shallow, holding only the lower half of the con- glomerate series, up to and including the second slate (6) ; and as it approaches the shore it is merged rapidly with the east- west monocline or rather with the main syncline. Although this interpretation seemed the most satisfactory at the time when the special map and section were drawn, sub- sequent observation and reflection have caused the alternative view to appear more acceptable; viz., that all the beds of the Beal's Cove section pass in regular order around the granite, the structure being monoclinal in both directions from the o-ranite. This later and present view is expressed on the general map, and hence the two maps are not in agreement here. The ledges of slate south of the Alms House and nearest to the shore appear at first sight to stand in the way of the later interpretation. But the strike of these outcrops is really much more northerly than mapped, being directly toward the most westerly ledges east of the Alms House. Certain irregularities in the strike were made the most of to bring the ledges into conformity with the earlier explanation ; and the dip, which is W. 85°, was also regarded as a local irregularity. I am now disposed to refer these ledges to the highest bed of red slate (12), placing them on the northwest side of the great bend, but very near the turning point. And the still higher beds, the massive conglomerate (13) and the gray slate (14), instead of bearing away to the westward, as if to cross the river. 227 are now regarded as bending to the north with the whore and following the lower beds around the bend. This gi-eat bend is, of course, the extremity of the great central and dominant anticline of Hingham, and that it is the extremity is proved by the undeniable fact that the axis here plunges steeply dowr)- ward to the west. Southward from this axis the • beds foi"m a monocline of from 80° to 85°, extending to Beal's Cove ; and on the northwest side a monocline of from 85° to 90° extends out beneath the level sandplain. At the time of my first observations in this locality, fifteen years ago, I was deeply impressed by the fact, that there are, northwest of the granite, two similar and parallel ridges of conglomerate and melaphyr, with an intervening valley com- posed chiefly of sandstone and slate. This strong topographic suggestion of a syncline biased all my later observations, until the })resent writiiig compelled a broader and more rigid examina- tion of the facts, and the absence of any real geological evidence of a synclinal structure became apparent. Whichever view of the structure of the beds on the north- west side of the axis is accepted, important strike-fiiults nnist be introduced to explain the second belt of melaphyr. This melaphyr is essentially similar to that on the east side, against the granite, and one naturally regards it at first as marking a denuded anticline. Its eastern edge, however, is quite cleai'ly transgressive with reference to the bordering strata ; and it is necessary either to introduce a fault here or to regard the melaphyr as a dike. No reliable dips have been observed in the conglomerate on the west side of this melaphyr ; but it may be reasonably correlated with that on the shore of Huit's Cove, which dips westward away from the melaphyr and beneath the slate series. This belt of melaphyr is a direct prolongation of the great body of melaphyr lying east of Huit's Cove, and can not be regarded as intrusive unless we are prepared to ascribe that origin to the entire area, or in fact to all the inelaphyr of Hingham, The fault separating the melaphyr from the slate 228 series on the west conceals for the entire distance all but the highest bed of conglomerate, and this also is cut out or, rather, concealed for a part at least of the distance between the cove and Lincoln Street, the slate lying here directly against the melaphyr. The area crossed by the western portions of Beal and Lincoln Streets is an unbroken sandplain, and the broad expanse of slate represented here is unsupported by a single outcrop. The map finds abundant justification, however, in the topography and in the extensive outcrops of slate on the shores of Huit's Cove, since, if continued, and we have no reason to suppose the contrary, these beds must cross this area. The tongue of melaphyr can not be traced south of a point 500 feet north of Beal Street ; and whether the bounding faults actually meet, as mapped, and, if so, whether the united fault extends farther to the southwest, it is impossible to determine. The general interpretation of the geologic structure here pro- posed makes it unnecessary to suppose that any of the Hingham strata extend far into Weymouth. All of wliich tends to emphasize the importance of Weymouth Back River as a geologic boundary ; and we may reasonably assume that this valley follows a fault comparable in magnitude and structural importance with that along the east side of Hingham Harbor, separating areas which are strongly contrasted in their geologic features. The Lincoln Street and Broad Cove Area. This area is represented by the southern part of the third special map (Plate 9), or all that part southeast of the melaphyr and south of the great dike. It embraces but few ledges, being covered almost continuously by the Squirrel Hill and Bradley Hill drumlins and the low sandplain and broad meadows into which they sink. The monocline northwest of Squirrel Hill can be readily traced to the northeast across Lincoln Street, as the map shows, as far as Huit's Cove Lane. All the beds, so far as 229 observed, are vertical ; and it is obvious that tiiere is room for not more than half of the conglotnerate series between the mclaphyr on the southeast and that on the northwest. The lower beds, from the melaphyr (2) and the great conglomerate (3) to the third conglomerate (7), are clearly continuous. Hence there can be but little doubt that the missing beds belong in the upper half of the section, or that their absen(;e is due to a great fault which has elevated the melaphyr on the northwest. The map represents the melaphyr between the conglomerate series and the granite as broadening rapidly toward the northeast, and this would seem to indicate a marked flattening of the dip. Outcrops are wholly wanting, however, south of Lincoln Street and east of Hawke's Lane ; and it now seems much better that this area should be colored as granite. But even then there would be a decided increase in the breadth of the melaphyr east of the lane. The steep monocline can not be traced beyond Huit's Cove Lane, The first one of the two transverse faults represented here appears to be justified by the offsetting of the strata as seen in the actual ledges ; and it is probably less important than the other, which is based upon the general interruption of the outcrops and especially upon the fact, that to the eastward of this line, so far as the scattering ledges allow us to judge, an entirely different type of structure prevails. The dips, of both the melaphyr and the sedimentary rocks, are everywhere low and indicate broad shallow folds. In the absence of outcrops immediately east of the north end of Huit's Cove Lane, we can not know whether this change takes place abruptly or gradually, by faulting or otherwise ; but the fault line on the map calls attention to the ftict that a change occurs somewhere in that vicinity, and it also makes it easier to interpret the outcrops along the line of Lincoln Street east of Huit's Cove Lane. The melaphyr south of Lincoln Street and west of Thaxter Street, at the northern base of Squirrel Hill, is mostly 230 greenish, sometimes reddish or purplish, abundantly and rather coarsely amygdaloidal in certain parts, with segregations of epidote and ferruginous silica, and in every essential respect similar to that which we have followed all the way around the granite from Portuguese Lane. The dip, so fsir as can be judged by the flow structure — the sheets and layers of amygdules — is nearly horizontal. The occurrence of felsite behind Mr. Bradley's barn, just west of Thaxter Street, as already described, makes it necessary to curve the southern boundary of the melaphyr to the north here. The last exposure of this melaphyr is in Thaxter Street, a few yards south of Lincoln Street; but it is assumed, in the absence of evidence to the contrary, to form a continuous belt bordering the granite as far east as the harbor. The melaphyr in the angle east of Huit's Cove Lane and north of Lincoln Street is of precisely the same character, except that it is, perhaps, more profusely amygdaloidal ; and the layers of amygdules indicate very plainly a gentle southerly dip, suggesting that the conglomerate on Lincoln Street, between Thaxter Street and the Lane, in which the dip has not been made out, lies in a shallow syncline, and that the melaphyr is continuous beneath it. This correlates the conglomerate with the great bed (3), which we have elsewhere found lying directly upon the melaphyr. It is the broad and open character of this fold, as thus indicated, that has led me to extend the conglomerate color so far beyond the observations. The large outcrop of melaphyr west of the junction of Downer Avenue and Crow Point Lane is quite varied in character. It is not conspicuously amygdaloidal, and the lower and northern part of the mass, although somewhat hrecciated and containing irregular, angular segregations of bright red jasper, is mainly quite compact and massive. Above, however, the melapli}^- is very scoriaceous, with numerous jaspery and epidotic segregations and veinlcts and distinct indications of a true conglomerate with well-rounded 231 pebbles lying upon and filling the inequalitieK of its surface. We have exposed here, apparently, the original upper surface of the nearly horizontal bed of melaphyr. Care is required to distinguish the segregations of jasper from the inclusions of red felsite. Several small fragments of felsite similar- to that a hundred yards away, at the junction of Downer Avenue and Crow Point Lane, and one mass a yard or more in diameter of a slightly more granitic character, were observed imbedded in the melaphyr, indicating very plainly that this ancient basic lava broke through the still more ancient acid lava on its way to the surface. The outcrop in the street, as previously explained, probably represents a knob or boss of felsite projecting through the flow of melaphyr where erosion has cut most deeply into the latter. The extension on the general map of these two belts of melaphyr eastward across Broad Cove and Otis Hill to the harbor is, no doubt, somewhat hazardous ; but no reason is apparent for terminating the melaphyr west of the shore, and its extension finds some, if not sufficient, justification in the bowlders of melaphyr observed on Button Island. The only rock observed in place on this islet, as previously stated, is granite ; but the numerous large and angular bowlders of amygdaloidal melaphyr on the north and northwest shores prove that, although this rock does not form part of the island, it must underlie the portion of the harbor immediately to the northward. The melaphyr forming these bowlders is similar to that in the vicinity of Lincoln Street, being of greenish and reddish tints, coarsely and distinctly amygdaloidal with epidote and quartz, and exhibiting numerous irrej^ular seo-reo-ations or veinlets of the same minerals. The great dike represented on the special map as extending west from Downer Avenue across Planter's Fields Lane to the eastern angle of the Huit's Cove melaphyr is probably the best interpretation of the outcrops which it embraces. These are all of precisely the same character, a typical coarsely and 232 uniformly crystalline, massive diabase, the texture being pro- portional to the magnitude of the dike. Another argument for the dike is that the strata on opposite sides of this line of outcrops are, as the map shows, clearly and entirely at variance in dip and strike, and a fault coinciding with the dike is a structural necessity. Whether the dike actually ends against the melaphyr as represented, it is impossible to determine, since it passes in this direction beneath an extensive wet meadow and outcrops are wholly wanting. The coarse, holocrystalline, and homogeneous character of the diabase utterly forbids connecting the dike with the melaphyr as a possible channel of supply for the latter. The melaphyr is certainly older and the dike is just as clearly newer than the bordering sedimentary rocks ; and the smaller dikes running south through the sandstone and conglomerate are probably, as indicated on the map, branches of the mahi dike. The strata between the great dike and the melaphyr of Downer Avenue and Lincoln Street form, apparently, a low monocline of 15° — 30°. We commence on the avenue, north of Planter's Fields Lane, with outcrops of a purplish and gray slate passing south into gray sandstone with fine [)ebbly layers, the whole dipping S. 30°. The sandstone can be traced west across the fields, as mapped, and is seen to change gradually upward into the small-pebbled conglomerate which outcrops so prominently along the entire distance, especially west of the lane. The dip of the conglomerate was not clearly observed ; but it passes on the south, through sandstone to a finely banded slate, 75 feet in breadth, with a southerly dip of only 15°. South of the slate are several slight exposures of sandstone before we come to the conglomerate crossed by the lane, near the melaphyr. This outcrop is more extensive than marked, extending 200 feet southwest of the lane. The correlation of these beds is not easy. Their surface exposures or developments, as compared with other sections, Occas, Papers, Boston Soc.Nat. Hist., YoI.IV; O O O O ^ y oo o oo toncfCoTneroUe. o o o <^ ^ Sandstone. Slate HHwl Dikes ScaleIinch =5oo feet Downer Crow Point and theNb Plate 10 lin^ Langley Id. Sarah Id. ^HBORiNG Islands 233 are, of course, greatly broadened by the relatively low dips ; and there is nothing to show conclusively with which part of the Real's Cove section, for example, they should be identified. No dips, however, have been observed south of the slate ; and it is possible tliat the low and wavering dip of the slate itself should be inter[)reted as marking a synclinal axis, the sand- stone and conglomerate on the south being a repetition of that on the north. The conglomerate could then, perhaps, be correlated with the first great bed (3) and regarded as passing up over the melaphyr, which would thus mark an anticlinal axis ; and we shoidd be able to dispense with the fault between the conglomerate and melaphyr. The Melville Garden and Planter s Fields A.rea. This area embraces the district east of the Huit's Cove melaphyr and north of the great dike and Otis Hill, including a large part of the tract known as Planter's Fields, Melville Garden, Pleasant Hill, the smaller drumlins forming Crow Point, and the adjacent islands of Hingham Harbor — Ragged, Sarah, and Langlee. From the western base of Pleasant Hill eastward, north of Melville Garden, to Downer Landing, the drumlins are continuous, as shown on the general map, and the hard rocks are wholly concealed ; while curving around the south and west sides of this drift area, as previously noticed, is a well-exposed belt of strata, the third and last cjeneral section of the cono-lomerate series. The most perfect outcrops are those afforded by the islands,^ which, it may be noted in passing, are a beautiful illustration of the dependence of relief upon geologic structure. The shores of these islets are almost continuous exposures ; and the atti- tude of the sti'ata is exceedingly constant, the strike being neai-ly due east-west and the dip S. 35°-40°. Langlee Island is an approximately rectangular mass 600 feet long and 400 feet 1 See the uncolored special map (PL 10). 234 wide, with two detached half-tide ledges on the eastei-n end, making the extreme length 800 feet. The two small bays in- denting the western shoi'e correspond to depressions crossing the island parallel with the strike and due, no doubt, to the erosion of the softer strata. Commencing with the lowest or most northerly beds, the island presents the following section : — Coiiiiiomerate, coarse to mediuin, 100 feet, equal to thickness of GO feet. Concealed, probably slate, 35 " " " " " 20 " Conglomerate, medium to fine, 170 " " " " "100 " Streak of sandstone near the northern edge. Concealed, probably slate, 70 " " " " " 40 " Conglomerate, 60 " " " " " 35 " 435 255 Along the north side of each depression there is a dike of uncertain width, the north dike having a northerly hade of 15° ; but it is improbable that the depressions are due wholly or even chiefly to the erosion of the dikes. Langlee Island is separated by about 600 feet of water from Sarah Island, an ol)long mass 275 feet in breadth and 835 feet in extreme length, with a h)ng, lialf-tide ledge about 100 feet from and parallel with the south shore. It is constituted as follows, in ascending order : Sandstone, gray and coarse, 85 feet, equal to thick- ness of 50 feet, with a streak of conglomerate near the middle of the bed. Conglomerate, medium to fine, 190 feet, equal to thickness of 115 feet, with a streak of sandstone near the northern edge. Water, probably concealing sandstone and slate, 75 feet, equal to thickness of 45 feet. Conglomerate, forminof half-tide ledfje. West of Sarah Island, with about 300 feet of water interven- ing, is Ragged Island, 235 feet in breadth and 735 feet in extreme length, with two linear ledges parallel to the southern shore and 100 feet distant. These two islands are thus essentially similar in form ; and they present identical sections, except that in the depression south of Ragged Island a little 235 sandstone is exposed, making it probable tliat, as stated, this break covers a bed of sandstone, or of sandstone and slate. The sandstone forming the north shore encloses from 5 to 10 feet of fine conglomerate, which becomes finer toward the west, changing to sandstone. It encloses masses of banded slate of various colors, some of which are possibly large pebbles ; but the greater number, it is clear, can not be explained in that way, and must be referred to the irregular deposition of fine silt with the coarse. In most cases the stratification planes of the slate coincide exactly with the bedding of the sandstone and conglomerate. The correlation of the two sections is unques- tionable ; but when we carry the line of strike from either island across to the other a lateral displacement is observed, each longitudinal feature on Ragged Island being about 100 feet farther south than its continuation on Sarah Island. This want of alignment seems to be best explained by a transverse fault between the two islands ; with the downthrow to the east, as show^n on the map. If such a fault exists, the vertical displacement must be about 75 feet. Some 300 or 400 feet southwest of Ragged Island, a i)ile of angular blocks of fine conglomerate projects from the mud flat when the tide is out. These masses probably represent an underlying ledge, although it is possible they were derived from the Ragged Island bed. In the .general line of strike of Ragged Island is the high and massive ledge of conglomerate forming the headland and the north shore of Walton's Cove, in Melville Garden. The con- glomerate has a breadth of at least 150 feet in the garden, probably passing beneath the water on the south ; and it may be safely correlated with the main bed on the southern islands. More careful observation, however, shows that there is here, again, a lack of alignment, the southern border of the conglom- erate on Ragged Island coinciding in direction very closely with the northern border of the conglomerate in the garden. This means a horizontal displacement of perhaps 150 feet, and a 236 vertical slip of at least 100 feet, with the downthrow to the east, the transverse fault between Sarah and Ragged Islands being repeated between Ragged Island and the main land. Northwest of the bridge across the cove, the cono-lomerate slopes steeply down to the water in one broad diagonal joint- face. On the north side of the headland at the eastern end of the ledge the conglomerate shows very distinctly a southerly dip of 35°.^ The shore retreats here, tiie headland giving way to a small sandy beach, at the north end of which a grav sand- stone outcrops with the same dip as the conglomerate. It is very evident that the sandstone is really as broad as the beach, that the base of the abrupt northern face of the conglomerate marks the contact of the two rocks, and that this sandstone is the continuation of that formina; the northern shores of Rao-wed and Sarah Islands. Between the buildings in the northern part of the garden is an exposure of conglomerate (see map) which is stratigraphically below the sandstone : and westward on this line, across Downer Avenue and east of Whiton Avenue, it outcrops sufficiently to prove a bed of considerable thickness. In its eastward extension this bed must, of course, pass wholly to the north of the southern islands. The outer [)art of Walton's Cove clearly corresponds to the gap between the islands and the ledges parallel with their southern shores ; but the inner part, influenced no doubt by the diagonal jointing of the conglomerate already referred to, is oblique to the stratifi- cation, and the same bed of conglomerate forms both shores. South of this conglomerate, and in line with the outer part of the cove, there are outcrops, as the map shows, of a purple, banded slate and gray sandstone. The slate is contorted, and is, doubtless, underlain as well as overlain by sandstone. The dips of the slate are, of course, unreliable, but the sandstone shows that the dip observed north of the conglomerate still continues. South of these soft beds, and forming the south shore of the cove, east of the bridge, is a third bed of con- I liy mistiikc the dip is iiuidu 25° on the map. 237 glomerate. This is small pebbled, and is probably the bed forming the ledges south of the ishuids, althougii dislinetly smaller pebbled. South of this eomglomerate there are no outcrops east of Downer Avenue. On the west side of the avenue, we have first the con- glomerate already referred to east of Whiton Avenue, fol- lowed by a blank space wide enough for the sandstone north of Walton's Cove ; and then we come, in the western part of Melville Garden, to a broad exposure of the main conglom- erate. In passing westward the strike changes gradually from east- west to northwest, and there is, apparently, a marked flattening of the dip, intercalated layers of slate and sandstone showing a southwesterly dip of only 5° — 10°. South of this conglomerate are ledges of sandstone followed by a blank space which probably conceals the bed of purple slate already noticed east of the avenue ; and then come, in the western part of the gjirden, near Crescent Avenue, 30 feet in breadth of the sandstone seen south of the purple slate, and after a blank of 50 feet, 100 feet in breadth of massive gray sand- stone with a southwesterly dip of 20°. Following the strike of this heavy bed of sandstone east across the small pond and the avenue, we have, apparently, no alternative but to connect it with the third or most southerly conglomerate in that part of the garden. If this correlation is correct, we have here a remarkable instance of rapid lateral change in the character of the coarser sediments, in view of which we may well hesitate in correlating this section with those south and west of the o-ranite axis. Although this explanation has the merit of simplicity, and has been followed in the construction of the map, it appears best, on the whole, as will be more clearly shown later, to introduce a third transverse or north-south fault here, the fault cutting obliquely across Downer Avenue between the two ponds. The beds of slate are commonly marked by an absence of outcrops, forming smooth open lanes or narrow valleys 238 between the conglomerate ridges ; and this peculiarity is the only fact that can be cited in support of the bed of slate shown on the map south of this sandstone bed. That is, the topography is suggestive of slate, but no outcrops have been observed. Following this blank, comes a belt of fine con- glomerate a good hundred feet in breadth. It outcrops boldly on the southwest border of the garden, northwest of the larger pond ; and can be satisfactorily traced westward along the east side of Grove Avenue to a point 750 feet northwest of Crescent Avenue, where it passes beneath the Pleasant Hill drumlin. The outcrops of this bed, in connection with the preceding, first fully outline the great curve in the strike ; and it will be noticed that while the lower part of this general section is based upon outcrops conforming with the eastern area of the curve and dipping south, the numerous outcrops of the upper half, yet to be described, follow the northern area and dip west. Northwest of the pond. Grove Avenue utilizes onie of the narrow slate lanes, and the slate, which is brown and dips west about 30°, comes to the surface in the northwestern extension of the avenue. The breadth of this slate is scarcely 30 feet ; and then follow in continuous outcrops 30 feet in breadth of gray sandstone, 40 feet of fine conglomerate, 30 feet of brownish slate, 90 feet of conglomerate with streaks of sandstone, and 40 feet of gray sandstone. These beds are clearly traceable along the strike for nearly one third of a mile, as shown on the map ; and the dip, throughout, varies but little from 30°, although constantly changing in direction . We have now reached a slate valley of unusual breadth. At the southeast end, near Grove Avenue, there is an obscure outcrop of brown slate ; and greenish gray slate, changing west to brown or reddish, is well developed on the west side of the artificial pond. But the only complete section is in the northern part of the vaHcy. Here the outcrops, across the strike, are almost continuous, and the breadth of the slate. 239 from the sandstone on the east to tliat on the west, is about 350 feet. The shite is somewhat variable in charaeter ; l>ein"- mainly brownish or chocolate-colored near the upper and lower borders, and gray or greenish gray, banded with dull red, in tlie middle of the belt. It is, throughout, thin-bedded, banded or shaly, and finely jointed. The dip is constantly to the west and southwest, but inconstant in amount. The ano-le "-iven on the map — 60° — is the highest observed, the normal varia- tion being from 30° to 40'^. The slate valley is bounded on the west by 60 feet or more of coarse gray sandstone, about 30 feet of brown slate, and 100 feet of sandstone ; and these beds can be traced south by frequent outcrops to the end of Planter's Fields Lane. The last sandstone is here ovei'lain on the west by a bed of brown slate, which can not be traced far to the north. Then comes a heavy bed of conglomerate, which at the northern end is actually exposed for a hundred feet in breadth and seems to broaden southward, possibly enclosing, as the map shows, another bed of brown slate. At the extreme north this conglom- erate is, apparently, overlain by still another bed of sandstone. That all of these beds are cut off on the south by the great dike scarcely admits of doubt ; but whether we have reached the end of the series toward the west, and how it terminates in that direction, we can only conjecture, as there is an absolute blank of about 400 feet between the last of the sedimentary outcrops and the first appearance of the Huit's Cove mela- phyr, except at the extreme northwest corner of the sedimentary area. lam now satisfied that the small outcrops of sandstone shown on the map near the border of the melaphyr really are melaphyr ; and the boundary line should be carried far enough to the east to include them. The actual gap between the two rocks at this point, measured across the strike, is probably not more than 150 feet. Although there are no reversed dips, a general view of this series of strata sujjo^ests that the main band of slate marks a 240 synclinal axis, the beds on the west appearing to be, so far as they go, a repetition in reverse order of those on the east. This is the explanation which I proposed twelve years ago, in my "Contributions to the geology of eastern Massachusetts," and it is expressed in the section accompanying the special map. This section shows an inverted syncline, on the east side of which the beds, as they recede from the axis, round over to a neai'ly horizontal position, as indicated by the observations in the northern part of Melville Garden, Avest of the avenue ; while on the west they maintain a high inclination until cut off by a fault against the melaphyr. But few dips have been observed west of the slate, and none so high as the section represents. Hence the syncline, if it really exists, is probably more completely inverted than it has been drawn. Then, again, the repetition of the strata, as the map shows, is by no means perfect ; and we note especially that it is impos- sible to find on the west side as many beds of slate as are clearly exposed on the east. Of the three beds which the map shows on the west side, the second is based upon a single outcrop and the thii'd upon none ; while the conglomerate and sandstone prevent the extension of either of these far to tlie north. In the construction of the map, it will be noticed, that the interpretation was preferred in each case which, witliout doing violence to the actual observations, is most favorable to the synclinal theory. It can not be denied, however, that an equally strong or stronger case can be made out for a monoclinal structure ; and the latter appears to me now the more probable view. Although this section, as a whole, bears a general re- semblance to the Vilhige and Bcal's Cove sections, tlie pre- cise correlation of fhe beds is a puzzling problem. The much gentler dips and consequent broader outcrops must be borne in mind ; for it may very well be that what in the other sections is reckoned as a single bed appears here, tlirough tlic expansion of its outcrop, as two or more distinct beds. Great 241 allowance must also be made for the undoubted lateral changes in the character of the strata. In fact, this considera- tion renders the independent correlation of any of the coarser beds extremely hazardous ; and I am fully persuaded that the only reasonably safe clue is afforded by the main belt of slate. This bed, if not folded upon itself, is about 150 feet thick : and if it is the equivalent of anything in the Village section, it must be the bed of red and gray slate near the middle of the section (6) which passes downward (south) into sandstone and upward (north) into conglomerate and has a thickness on Hei'sey Street of 130 feet. Assuming this correlation to be substantially correct, the importance of the lateral changes in the strata becomes apparent when we turn to the coarser and more variable parts of the section, and especially when we attempt to find in beds 3, 4, and 5 beneath the slate in the Village section the extended series of sti^ata north and east of, i. e., below, the slate in the Melville Garden section. This point of view makes it almost necessary to postulate the transverse fault previously referred to as possibly crossing Downer Avenue obliquely between the two ponds. Sup- posing the downthrow to be to the east, as with the faults east and west of Ragged Island, the series of outcrops in the center and eastern part of Melvill-e Garden become the equivalent of those in the extreme western part of the garden and extending from Downer Avenue along the north side of the main slate. Even then a satisfactory detailed correlation is scarcely possible, perhaps on ac- count of imperfect outcrops, and especially are we at a loss to assign a place to the beds of Langlee Island, without the further assumption of a profound east-west or strike-fault, with the downthrow to the north, between Langlee and Sarah Islands, the beds of Langlee Island being, approximately, or in part, a repetition of those of Sarah Island. This hypothetical strike-fault may or may not be supposed to cross the north- south fault between Rao-o-ed and Sarah Islands. In the former OCCAS. PAPERS B. S. N. H. IV. 16. 242 case, Crow Point and the extreme northern part of Hingham Harbor are probably underlain by the conglomerate series. While in the absence of strike-faults this concealed area may be regarded, with much probability, as melaphyr, an oblong and approximately horizontal block with the conglomerate series dipping gently away from it on the south and west. It is important to notice, however, that if the very low dip west of Walton's Cove be regarded as normal, the conglom- erate series would naturally reach across Crow Point and cover the melaphyr ; and since this view is most distinctly indicated by the actual observations, it has been followed in the construction of the map and sections. The eastward extension of the beds of Langlee and Sarah Islands would carry them beneath Pine and Planter's Hills ; and in the absence of evidence to the contrary the conglomerate series is represented on the general map as ending in this direction against the o-reat fault separating the Hingham and Nantasket areas. The Huit's Cove Area. This area embraces all that remains of the third special map (Plate 9), including the large quadrangular body of melaphyr east of Huit's Cove and the sedimentary rocks which border the melaphyr on the north and west and form the immediate shores of the cove. It is a moderately elevated and ledgy tract, with comparatively little swampy ground. The east and southeast borders of the melaphyr, as previously noticed, are not exposed at any point ; and it is possible to determine neither the exact position nor the nature of the contact with the sedimentary rocks. But on the north and west the con- ditions are much more favorable, and the boundary is drawn with substantial accuracy at most points. This melaphyr is, for the most part, a compact to imperfectly crystalline dark o-ray rock. The color varies, according to the condition of tlie iron oxide, from gray or greenish to reddish or purplish tints. 243 In texture it most resembles the melaphyi- of the Village area, but is more slaty in some parts and, apparently, more crystal- line in others, the crystallization being, perhaps, most marked in the least altered dark or gray variety. A finely porphyritic variety resembling porphyrite has also been observed. It is especially contrasted with the melaphyr bordering the granite axis by the comparative absence of the amygdaloidal and scoria- ceous characters. Amygdules are often present, usually in small patches, but are mainly small and indistinct, althouo-h occasionally very large and conspicuous. They consist partly, as elsewhere, of epidote or epidote and quartz, but chiefly of a pure, dark green chlorite ; more rarelv of cleavable calcite. The calcite amygdules have been, superficially, very generally removed through solution, leaving the original steam-holes essentially intact. Chlorite is decidedly the most abundant and characteristic secondary mineral, and epidote, whether dif- fused or segregated is of subordinate importance. The melaphyr area, as a whole, may be fairly described as quite uniform in character ; and no reasons are apparent for drawing geological boundaries through it. No intrusive con- tacts have been observed ; but this point will be referred to again in describing the slate. Neither are there any known facts requiring us to refer the melaphyr to diflPerent periods of eruption ; although it is, perhaps, doubtful whether so large a body should be regarded as belonging to a single flow. In attempting to settle this point, I have searched without success for some development of the flow structure suflficiently marked to show whether the melaphyr still retains in the main, as I suppose it does, an approximately horizontal position. Of course, if still horizontal, the thickness is not necessarily great. The only facts which I have observed that seem to have any bearing upon this point are as follows: (1) At the extreme northeast corner of the melaphyr, not only are the ledges marked as sandstone just east of the boundary, on the map, all melaphyr, but the ledge immediately within the boundary 244 marked as partly sandstone is also all melaphyr ; so that there is no evidence, as once supposed, of sandstone overlying the melaphyr. (2) Directly north of these ledges, about half way between there and the shore, in the area colored as slate, is a ledge which appears on the map as sandstone and slate ; but I am now satisfied that it is really fine-grained granite and felsite overlain by compact, slaty, and chloritic melaphyr. The boundary of the melaphyr at this point should thus be carried a hundred feet farther north as well as east. But the special significance of this outcrop seems to be that it shows us the actual base of the melaphyr, and that the northeast corner of this great block of melaphyr is tilted up a little higher, at least, than any other part. That the contacts between the melaphyr and the sedimentary rocks bounding it on the west and north are lines of profound displacement is unquestionable, unless we are prepared to regard the melaphyr as intrusive in the slate and conglomerate, that is, as forming a vast dike or laccolite ; a view which, it may be stated once more, finds no support whatever in the petrographic characters of the melaphyr, nor in any facts now exposed to our observation. The slate forming the shores of Huit's Cove and extending around the northern end of the melaphyr is undoubtedly, as previously explained, the great slate (14)of the Beal's Cove section, and in its normal position, or stratigraphically, it must be separated from the melaphyr by more than a thousand feet in thickness of the conglomerate series ; yet here it lies directly against the melaphyr or separated from it only by the single bed of conglomerate, which, as the detailed observations will show, is clearly the highest member of the conglomerate series. We thus see that, simply as a measure of erosion, the melaphyr is impressive, since there must have been removed from its surface, not only the eiitire conglomerate series of Hingham, but also the still greater volume of the overlying slate series ; and the bordering displacements, although they have barely sufficed to bring the bottom of the Occas, Papers, Boston Soc.Nat. Hist., Vol. IV, Plate. Cong I om era ie AAA A A A A Slate Mehiphyr Dikes SCALE:IIKCH =330 FEET. '^loA ^°n°/ ^ V7 V rJr x^+y-x*> i^xX>^X><-v: fxxx^^X x>r V^ + X+^'Vx^-i < + XJrX-V+iri i^-X x^k%4 Pti NWe av> j--7-f-iirrrr.jt.R.cut )?X-i-+V>-^A-^7CX % o h ^ (t J < (X o w m h D en h 2 O h < CO X h D O >^ o w > o D g < z O w i; H u, O < O u H D O J and on the north a smooth field sloping gently to the raili-oad, with an abundance of slate debris in the soil. Two to tin-ee hundred feet east of Webb Street on the same direct line, which is here about 500 feet south of Commercial Street, and immediately east of the little brook draining the artificial pond, the granite outci'ops prominently, as before: and imme- diately north of it the slate, with a nearly vertical northerly dip, is exposed for a breadth of 50 feet, and again for about 130 feet at the barn midway between the granite and Commercial Street. The granite ledges continue to the eastward, but the slate is not exposed again on this line. Instead, we find that on the east side of the brook flowing directly north from the artificial pond the ledges of granite and diorite extend north to Commercial Street, and thence northeasterly across the field to the railroad cut on White's Neck. Undoubtedly the slate abuts rather abruptly eastward against a wall of granite. On the map this interruption has been inter- preted as a transverse fault, which would thus be compensating with reference to the Mill Cove fault. But this Fore River 'fault is not so urgently demanded as the Mill Cove fault ; for the actual outcrops are not inconsistent with the view that the great boundary fault is not broken here, but simply takes a diagonal course, as is, in fact, indicated by the oblique way in which the granite approaches and enters the White's Neck section. The Fore Eiver fault is, therefore, proposed with some hesitation and with a desire for uniformity in the geolog- ical interpretation. The topographic indications are entirely consistent with either view. It may be added, however, that, as we shall presently see, the Fore River section, like the White's Neck section, passes northward from black slates to green and red slates, the two sections being practically identi- cal in this respect ; and this important geological fact certainly harmonizes best with the theory of a transverse fault, so that it may be regarded as determining a controlling probability in favor of the map. 400 The ledge between the mouth of Smelt Brook and Quincy Avenue and that projecting from the north bank of Weymouth Fore River directly opposite the brook, are green and highly fissile slate with east-west strike and nearly vertical northerly dip, agreeing closely with the outcrops along the Monatiquot River to the westward. The only remaining outcrop of slate in the Fore River section is that on the north side of Fore River where it bends to the northward, about one third of a mile east of Quincy Avenue. This ledge, whicli is on the land of Mr. T. A. Watson, is about 300 feet long in a northeast and southwest direction and 100 feet wide. It is in the main a very massive greenish gray rock which is in part banded and mottled with reddish brown and purple. The stratification is especially marked by lenticular chloritic and micaceous streaks, which are often partially cavernous, owing to the solution of calcareous matter. The calcite, where undis- solved, is highly crystalline, and presents distinctly prismatic shapes six to twelve inches long and one fourth to one half inch in diameter, lying longitudinally in the dark chloritic streaks. The slate has a high dip to the southeast and is very much broken by transverse rifts and joints. Some blasting has been done on the south side of this ledge ; and Mr. Watson was thus enabled to find, in 1892, the fragment of a large Paradoxides which Mr. Grabau has described. It shows por- tions of nine segments and represents an unusually large indi- vidual. Through the generosity of Mr. Watson this specimen is now the property of the Society. Ledges of fine granite are near by on the west and northwest. In fact, at the southwest end of the ledge the slate and granite are seen nearly in contact, and at the northeast end the actual contact is exposed, but not clearly enough for details. The nortliwest side of the slate ledge thus marks, approximately at least, the contact of the granite and slate, and that it is an igneous contact is plainly indicated by the dikelets or apophyses of fine granite which may be observed in the slate toward either end of the ledge 401 (Fig. 26), as well as by the massive and metamorpliic charac- ter of the slate itself. Westward, on and close north of this line of contact, which must follow closely the junction of the fiat land and marsh Avith the upland, the fine granite is exposed at short intervals nearly all the way to Mr. Watson's house. Here the line is abruptly shifted several hundred feet to the south, to correspond with the outcrops west of Quincy Avenue, along the north side of Allen Street. The salient angle of the gi-anite due to this shift is marked by the bold ledge of granite south of Mr. Watson's house ; and the shift itself is a' very probable fault line, as will be noted later. Fig. 26. — Dikelet of flue granite in tlie Cambrian slate on Weymouth Fore River, Braintree. Scale : 1 inch = 2 feet. The Monatiquot JRiver Area. — This important area em- braces all the Cambrian strata of East Braintree and the Monatiquot valley west of Smelt Brook. The actual outcrops are, however, wholly confined to the district less than a mile in length between Smelt Brook and the millpond of the Jenkins rubber works, as shown on the special map (PI. 19). As OCCAS, PAPEIIS B. S. N. H. 402 before, the slates are dark gray to black on the south side of the valley, but change gradually northward to greenish, banded and mottled with red, and then to brownish red, streaked with green, in the most northerly outcrops ; and the phenomena of special interest are found chiefly along the contacts of the slate and fine granite — an original igneous contact on the north and a fault contact on the south. Except near the igneous contact, the slate, of whatever color, but more especially the black and green, is highly fissile, the cleavage conforming perfectly with the bedding. The fine granite is well exposed on both sides of Quincy Avenue north of the river, and especially in the angle between Quincy Avenue and Allen Street. Going west on Allen Street the granite passes behind the massive, cross-jointed slate on the north side of the street ; but about 800 feet from the avenue it is brought abruptly out to the line of the street again by a transverse fault, as shown on the special map. The fault is parallel with the joint-structure of the slate and hades S. E. 10°. The normal strike of the slate, as revealed by occasional purplish lines in the dark greenish gray, is E.— W. ; and dip S. 80°— 85°; but south of this jutting angle of granite it is deflected to a south of east strike. This salient of the granite is of special interest because its contact with the slate is a typical fiiult on the southeast and typically igneous on the southwest. The igneous contact conforms closely with the bedding of the slate, and the two rocks are firmly welded together, the granite being microcrystalline and almost felsitic near the slate. About 65 feet west of the granite salient, the slate is traversed in a breadth of 6 feet, and very obliquely, by half a dozen or more vertical dikelets of granite varying from one fourth inch to four inches in width and hading S. W. 15°- 20°; and it is probable that by removing the soil these apophyses might be traced back to their source in the main body of granite. They appear, as do the apophyses generally, to be independent of the joint-structure of the slate. The 403 massive, dark greenish gray and brownish slate continues with- out fartlier interruption to Shaw Street, where, immediately north of the slate, the granite outcrops obscurely in the private road. Although the slate bordering the granite along the north side of Allen Street is of a massive and more or less metamorphic character, it becomes gradually but rapidly less metamorphic away from the granite ; and the slate between the street and river, some fifty feet in breadth, is entirely nor- mal — greenish gray, streaked with brown, thin-bedded and shaly, and dipping S. 80° — 90°. Farther west on this line, the fissile slates, chiefly of brown- ish red color, and conformably enclosing two small trap dikes, are well exposed about East Braintree Station and along Front Street; but the nearest outcrops of fine granite, as the special map shows, are several hundred feet distant ; and the contact cannot, therefore, be definitely located. Through the middle of the valley the highly fissile green slates, with streaks and bands of red, are well exposed. At the Jenkins Rubber Mills the outcrop has a breadth of about 200 feet ; strike N. 82° E. , dip N. 87° ; and the slate again encloses conformably a narrow dike of diabase. Along the river, both above and below Com- mercial Street, as far as the railroad and Shaw Street, the outcrops are almost continuous, with a total breadth of fully 300 feet, the strike and dip remaining unchanged. On the north side of the river is a conformable dike one foot or less in width, possibly the same as at the Jenkins Mills, but highly chloritized and largely replaced by vein quartz. It was on the south side of the river, just east of the junction of Commercial and Union Streets, that Mr. Albert Hobart found in the green fissile slate, during the excavation of the ledge for a cistern under his barn, in 1872, the fragment of a Paradoxides which first came to the notice of Mr. Watson in 1896 and was through him contributed to the Museum of the Society. It shows portions of seven thoracic segments ; and like the specimen from Mr. Watson's own ledge, a mile to the 404 northeast, it represents a large individual. Mr. Hobart reports finding in the coarse of the excavation numerous similar frag- ments ; but only this one was preserved. It is identical in lithological character with the adjacent ledges ; but, although I regard the account of its finding as absolutely trustworthy, Mr. Grabau and I were unable by careful search to bring to light any additional paleontological material in this locality. South of the line of Commercial Street we pass to the dark gray and black slates which normally overlie the green slates in the middle of the valley, as these do the red slates on the northern border of the valley. Advancing from the west, the first outcrop occurs on Union Street, southeast of the Jenkins Rubber Mills, showing the thin and fissile black slate of the Weymouth railroad cut striking N. 80° E. and dipping N. 88°. Turning up Liberty Street some three hundred feet, the same black thin slate is found exposed in the gutter for a breadth of one hundred feet, with the strike unchanged and dip vertical. It was also extensively exposed during the construction of the Jonas Perkins school-house on the east side of Liberty Street ; and in the fields to the southeast, as shown on the map (PI. 19) , the outcrops extend nearly six hundred feet south of Commer- cial Street. The chief outcrop on this line, however, is that half a mile farther east, at the junction of Commercial Street and Quincy Avenue. This shows, on the east side of the street, from the south: — (1) Fine pink granite south of Smelt Brook: (2) concealed, 75 feet; (3) trap dike, inclosing a large body of slate and granite, 225 feet ; (4) slate north of the trap, nearly to the railroad, 120 feet. The slate is dark gray to black, and indistinguishable from tliat in the railroad cut east of Smelt Brook, except that near the great dike it is rather more mas- sive. The strike is N. 70°-80° E. and the dip N. 85°. The contact witli the large body of compact grecnisli trap is a})prox- imately conformable with the bedding, but distinctly igneous and not due to faulting, the two rocks being firndy welded 405 together and the trap becoming appreciably more compact and tlie slate less fissile near the contact. In the slate, fifty feet north of the great dike, are two vertical, east-west dikes of gray compact trap six to twelve inches wide and separated by a foot of slate. The inclusion in the great dike consists of 80 to 50 or 60 feet of slate (soft, gray and visibly stratified) bordered on the south by a gray, slaty, bastard sort of granite. Under the blacksmith shop on the edge of the marsh is trap, which seems to divide the inclosed mass. The whole appearance is as if this great dike 225 feet wide had come up along the contact of the granite and slate and had inclosed a large block of the contact, the inclusion showing rather more slate tlian granite. Westward from Quincy Avenue, along the south side of Commercial Street, the same conditions continue. The great dike is exposed continuously for nearly a thousand feet, with an average breadth of at least three hundred feet. Ledges of fine granite are quite near it on the south, and the included mass or partition of slate and granite, with several detached masses of granite, can be traced as far as the first lane running south from Commercial Street, about six hundred feet west of Quincy Avenue ; so that in this part of its course the dike is virtually double. Farther west on this line the rocks are largely concealed by an embryo drumlin ; but on the west side of the lane midway between Quincy Avenue and Liberty Street, the trap and granite are well exposed, showing that the dike is still at least two hundred feet wide, but apparently without included slate, although inclosing fragments of granite. This outcrop lies to the south of the main line of the dike, suggesting a transverse fault with a horizontal displacement of about two hundred feet, or approximately equal to the breadth of the dike ; and it is interesting to note that this fault corresponds closely in position and direction of throw with the fault breaking the northern contact, on Allen Street. Farther west on this line the country is heavily drift-clad ; the great dike does not outcrop again ; and it cannot be confidently traced by its erratics ; 406 although it is altogether probable that it continues at least as far to the westward as the main line of the Old Colony Railroad. Returning to the section on Quincy Avenue, and proceeding thence eastward, we find on the east side of Smelt Brook, in the trend of the great dike and the slate, a continuous and abrupt wall of fine granite, extending north to the west end of the railroad cut at Weymouth Station. Smelt Brook appears, thus, to mark the position of a transverse fault, with a lateral shift of four or five hundred feet. The topographic indica- tions are very favorable to its extension northward across the river, passing in the rear of Mr. Watson's house, where the northern wall of o-ranite is ioo: O u -J o < O o U H D O 415 silicates (epidote, etc.) which give it exceptioiiiil hardness. Selected specimens gave, on analysis by Dr. W. H. Walker : — Insoluble residue 27.80 CaCO^ 71.50 MgCOg > by difference 0.70 FeCO.3 > 100.00 Like the calcareous concretions described above, the lime- stone bands are purest in the middle and merge rather gradu- ally outw^ard into the slate. One of the most puzzling features of this group of ledges is a band of massive greenish slate, twenty to thirty feet in width, crossing the formation in a generally north-south direction and interrupting the continuity of the red beds along the strike. As indicated on the map, it commences on the shore west of the second bed of limestone and extends directly north across the third limestone band; and then, apparently, it is bent abruptly to the northeast, reappearing beyond a gap in the ledges sixty feet to the eastward. Still farther to the north- east, across the marsh, the same or a similar band reappears and branches toward Pearl Street. This rock is so massive as to suggest a dike of fine-grained trap ; but there are no walls, and in both color and structure it blends both east and west gradually and perfectly with the red slates. The bedding lines and calcareous segregations are wholly wanting ; but the epidote, which occurs both in a diffused form, accentuating the green color of the rock, and as veinlets, and is undoubtedly a secondary constituent, is a plain indication that the massive slate was once calcareous. I can see no escape from the con- clusion that the slate along this line was, at an early pei'iod and probably before its complete consolidation or the segre- gation of the calcareous matter, exposed to some influence, such as faulting or compression and shearing movements, which has, as it were, shaken it down, obliterated its original structures, and reduced it to a structureless mass ; giving 416 rise to what is virtually a slate dike of the same type as the quartzite dikes described by Professor J. D. Dana^ in Dutchess County, New York. In the practical absence of fossils, the geological interest of these Lower Cambrian outcrops culminates in the alteration of the calcareous portions through the metamorphic influence of the granite. The only conspicuous secondary mineral is ejoidote, and it is very obvious that its development has been confined chiefly to the impure (aluminous and ferruginous) peripheral portions of both the claystones and the limestone bands, so that each calcareous body of whatever kind has normally an epidotic shell or outer envelope, which for the clay- stones, is commonly from a quarter to half an inch in thickness, and for the limestone strata from an inch to five or six inches. In the purer central portions, alike of the claystones and lime- stone strata, the aluminous impurity or clay was, apparently, insufficient for the formation of epidote ; and, except in the case of the thicker limestone layers, the residual calcareous matter has very generally been dissolved out by meteoric waters, leaving within the epidotic shells only the insoluble silica, iron oxide, etc. I have, however, observed concretions in which the calcareous center or core is still undissolved ; and in a few instances at least this core is a white crystalline lime- stone, like the centers of the limestone strata, suggesting that the formation of the epidote may have been, through the removal of the silica, alumina and iron oxide, a purifying pro- cess for the calcareous residue. We may, perhaps, suppose that the formation of epidote began toward the exterior of the mass, where the various constituents occurred most nearly in the right proportions ; and then that the shell thus established thickened both externally and internally until the supply of certain constituents — lime for the outside and alumina, silica, etc., for the inside — was exhausted. A central resithie of calcium carbonate (calcite) appears the more probable in view 1 Aiiier. .journ, sci., vol. 10;{, ])]•. 181-182 ; Bull. Essex Inst., vol. 27, j). 145. 417 of the fcict tliat while u pure epidotc rcquh-es only 2o per cent, of lime, claystones may be more than half calcium carbonate, equivalent to at least 30 per cent, of lime ; and the fui'thcr fact that the epidotic shells are really very impure, containiMi;' in intimate admixture with the epidote a large proportion of free silica. Mr. White has noted the impurity of the epidotic shells ; and he also mentions peculiar phenomena of rifting in these walls, as though they had been regions of special shrinkage and distortion, and states that the same tendency is observed in thin sections, in the formation of minute rifts. This shrinkage is certainly very probable, since the partial conversion of a highly hydrated and carbonated clay stone to a nearly anhydrous silicate of superior hardness and density must be attended by condensation. The probability of a resultant central residue of calcium carbonate might have been tested by an analysis of the central portion of a claystone which had not suffered the epidotic alteration ; but it has appeared ])ractically impossible to obtain such a specimen which had not had some of its calcium carbonate dissolved out by meteoric waters. Besides the epidote, the only other distinct silicate recognized among the alteration products is chlorite, and this occurs very sparingly, a fact which harmonizes well with the small proportion of magnesia among the original carbonates, as shown by the analysis of the limestone. The metamorphism described in the preceding paragraphs is strictly confined to the calcareous portions of the formation, and gradually diminishes in intensity northward, so that in the most northerly outcrops the secondary minerals are wholly wanting. Attention is thus directed to the belt of granite bordering the sedimentary rocks on the souths as the probable source of the metamorphic influences ; the only other -igneous mass in sight being the small trap dike cutting the slate at the southern end of the promontory, and this evidently was formed after the epidote and has exerted no appreciable metamorphic influence upon the inclosing sediments. OCCAS. PAPERS B. S. N. H. IV. 26. 418 Eastward from the shore, these red slates with the calcare- ous segregations showing increasing epidotic alteration to the southward, outcrop almost continuously for half a mile. On and near Pearl Street they are well exposed both north and south of the brook and for a total breadth slightly greater than on the wooded promontory. Peai-1 Street turns to the east here, following the northern edge of tlie granite ; and between the granite and the brook, north of and in, but not south of, the street, the slate is much in evidence. On and near North Street^ from the corner of Pearl Street northward, and in the field east of the street, the slate outcrops are almost continuous for a breadth of 400 feet. The most easterly outcrops are fully five hundred feet east of the street, and the dip is practically vertical throughout. The epidotic shells representing original calcareous segregations ai'e a conspicuous feature to the south- ward, and in this direction, also, the color is, as usual, dark red streaked and blotched with green ; but to the northward the color proportions are gradually changed, until we have mainly green slate streaked with red. Farther north on North Street, after an interval of a hundred feet, the slate outcrops obscurely in the field about a hundred feet west of the street and the same distance south of the southern margin of the sand plain ; and just in line with the south side of Neck Street. The out- crop shows at least thirty feet in breadth of greenish gray, fis- sile and non-calcareous slate, with east- west strike and vertical diD. This is the most northerly outcrop in the Mill Cove area ; and the total exposed breadth of the Lower Cambrian strata is quite certainly not less than eight hundred feet. Modi- fied drift conceals everything from the North Street ledges northward to the shore and eastward to Back River. The granite is not exposed east of North Street ; but it out- crops boldly at the corner of Nortii and Pearl Streets, near the slate, and forms a broad ridge extending west and south- west, parallel with Pearl Street, to the shore. The most southerly outcrop of granite in this- area is on tlie south side 419 of the high sandy inclge or esker on which the cemetery standB, and several hundred feet west of Norton Street, at the point where it cuts through the ridge and turns to the southeast. This outcrop is of medium coarse texture — distinctly coarser than any of the outcrops north of the cemetery I'idge ; and as we approach the slate the granite, which is everywhere very dark-colored below the zone of surface oxidation, becomes gradually finer-grained, and is essentially microcrystalline along the northern margin of the belt, where it faces the slate. Westward, the granite may be regarded as ending against the north-south Mill Cpve fault, which farther south forms the common boundary of the King Oak Hill and White's Neck areas of the southern boundary belt. Unfortunately, the contact of the granite and slate is nowhere clearly exposed ; and on Norton Street the nearest outcrops of the two rocks are about 150 feet distant. But there can be no doubt that it is an original igneous contact. The increasing fineness of the granite and raetamorphism of the slate toward their common boundary require this explana- tion. We have, however, still more conclusive proof of the invasion of the slate by the granite along this line. The granite ledges end on the west in Burying Island, a low emi- nence projecting into Mill Cove some three hundred feet south of the wooded promontory first described, and connected with the mainland by the salt max'sh. On the north side of the island the fine dark OTanite or granodiorite cuts and incloses isolated masses of a dark gray and extremely tough and mas- sive slate. The slate is decidedly metamorphic, showing con- siderable secondary mica (sericite), and is firmly welded to the ffranite. On account of its dark and massive character it is easily mistaken for trap : but it is also faintly stratified in a vertical east-west direction. It is clearly cut and injected b}^ the granodiorite ; and beautiful veinlets of pegmatite (red feld- spar and quartz) from a line to one fourth of an inch wide •extend from the granodiorite through the slate. The princi- 420 pal mass of slate is three to six feet wide, and is exposed for a length (east-west) of over forty feet, but it may be much longer. One foot north of this main body of slate is the smaller mass shown in Fio-. 27. Fig. 27. — Showing tlie relations of the massive gray slate to the grano- diorite on Burying Island, Mill Cove. Scale : 1 inch = 1 foot. The relations of this massive gray slate to the red calcareous and epidotic slate are not quite clear ; but that they belong to the same conformable series is indicated by the ledge on Norton Street immediately south of Pearl Street. This outcrop shows, on the north, twenty-five feet or more of the typical red slates dipping south as usual at a high angle — about 85°; and this is closely followed on tlie south by a dark gray, mas- sive and semi-metamorphic slate similar to that on Burying Island, but containing less sericite. The contact of the two slates is not clearly exposed, but appears to be quite conforma- ble. The bedding cannot be certainly made out in the gray slate, except in partings two to four feet apart. These are parallel to the bedding of the purple slate. The exposed thick- ness of the gray slate is 22 feet ; and the granite is probably quite near it on the south. Up to the time when the preceding paragraphs were written, paleontological proof of tlie Lower Cambrian age of the Mill Cove slates was, as stated, practically wanting: altliough 421 Foerste ^ had noted in the limestone bands obscure indications of Hyolithes ; and the vvritei- had found many years ago, in the red shite, a cross section of a trilobite, wliich, unfortunately, was lost before it could be described. But all this has been changed by the recent (December, 1899) brilliant discovery by Mr. Henry T. Burr,^ in the dark reddish slates in the southern part of the field, of an important Lower Cambrian fauna. The precise locality is the slate ledge on the north side of Pearl Street, about midway between Norton and North Streets, on the land of Mr. Richard Ash. The fossiliferous bed, which has been traced for only about thirty feet on the strike, is about one hundred feet north of the granite, and, apparently, lies wholly south of or stratigraphically above the calcareous poi'- tion of the series. By diligent search, aided by blasting, the expense of which was generously assumed by Mr. Watson, ten species have been identified up to the present time, all of them typical Lower Cambrian forms. These have been briefly described by Mr. Burr and more fully, with illustrations, by Mr. Grabau in his monograph on the Cambrian faunas of the Boston Basin published herewith. Altogether, this exceed- ingly opportune discovery is unquestionably the most important contribution to the paleontology of the Boston Basin since the discovery of Paradoxides by Professor W. B. Rogers. Concerning the northern limits of the Mill Cove slates we have no definite knowledge. They give no indication north- ward of another body of granite, nor do they exhibit the marked shaly structure suggestive of proximity to an important longitudinal fault. We have noted a tendency to a change northward from red calcareous to green non-calcareous slates, which suggests a passage in that direction from the Lower Cambrian to the Middle Cambrian. This would reverse the topographic order of succession observed in the eastern part of the southern boundary belt, where we pass southivard from iProc. Boston soc. nat. hist., vol. 24, 1889, pp. 261-263. - American geologist, vol. 25, Jan., 1900, pp. 41-50. 422 Lower to Middle Cambrian, and would put the two boundary belts in anticlinal relation to each other — a slightly fan- shaped or constricted anticline, since the prevailing dips are northerly in the southern belt and southerly in the northern belt. The composition of the North Weymouth sand-plain gives us no definite or trustworthy clue to the character of the under- lying rocks. It may be stated, however, that there is no indi- cation whatever of the occurrence of the fine granite or any plutonic rock north of the Mill Cove slates ; but that the drift is full of slate debris, — mostly black and fissile and partly gray or greenish, with ■ little or no red slate. The vague testimony of the drift is, therefore, confirmatory of that of the ledges, and we may fairly conclude that the Lower Cambrian slates are probably followed northward by the Mid- dle Cambrian. On Neck Street, nearly opposite the northern end of Green Street, I observed, in the sidewalk, some large and angular fragments of coarse arkose sandstone identical with that which, farther west, in Milton, immediately borders the Complex and the Cambrian strata on the north. Assuming this mate- rial to be approximately in sitzi and to have the same strati- graphic significance here as in Milton, Avould double the breadth of the Mill Cove section and establish its northern boundary a little south of Bridge Street, in Old Spain. In the absence of any more positive or trustworthy evidence, the map has been drawn to accord with this vicAv. The Washington Sti'eet Area. — Crossing Weymouth Fore River to Quincy Point, we find again a very perfect development of the Quincy plain, extending west along Wash- ington Street, between the shore of Town River Bay on the north and the ledg-es of fine o:ranite on the south. The ffran- ite is exposed northward to a fairly definite east-west line mid- way between South Street and Washington Street. The structural value of this line is apparent to the westward, in 423 the vicinity of Union Street, where it is seen to mark tlie boundary between the granite and the Cambrian strata ; and we may safely assume that tlie broad level plain north of the granite and east of Union Street, which is unbroken by a single outcrop of any kind, is underlain continuously by sedimentary rocks. This Washington Street area, which may be regarded as extending west to the Old Colony Railroad, embraces no actual outcrops excejit those already referred to in the vicinity of Union Street. The most northerly outcrop is on the south side of Washington Street just east of Pond Street, where the smoothly glaciated slate ledge has been uncovered for twenty feet in laying the electric car track. The slate is of a massive, gray, semi-metamorphic character, with a nearly vertical dip and east-west strike. Turning south through Union Street, scattering outcrops may be observed on either side as far as Baxter Street, where we pass onto the fine granite, giving a total breadth of slate, from the outcrop on Washington Street to the ffranite, of nearlv eio-ht hundred feet. The strike is N. 80° E., and the dip S. or towards the granite, 70° -80°. Eastward from Union Street, along Newcomb Place, the slate, closely bordered on the south by the granite, is exposed for about five hundred feet ; and the granite ledges cross Newcomb Place between six and seven hundred feet east of Union Street. The exact contact of the slate and granite is not exposed at any point ; but it evidently conforms approxi- mately with the strike of the slate and is undoubtedly igneous. The granite is very fine-grained near the slate, and in part microcrystalline. On the main line of contact no apophyses of granite or inclusions of slate have been observed ; but on the west side of Union Street, between three and four hundred feet south of Baxter Street and one hundred feet north of South Walnut Street, on the line of the coarse and fine gran- ite, is an outcrop which throws clear light upon the relations of the granite and slate. It shows from south to riorth : (1) 424 coarse granite ; (2) about twenty feet concealed ; (3) a band of hai'd siliceous slate, fifteen to twenty feet wide ; and (4) the typical fine granite fifteen feet, and probably extending much further. An undoubted igneous contact between the slate and fine granite is clearly exposed ; and this is evidently a large mass of slate enclosed in the fine granite near its junction with the coarse granite, several hundred feet south of the main body of slate. The attitude of the slate is normal — east-v\^est strike and high southerly dip. In all these Union Street outcrops the slate is dark gray, with usually a reddish tone. In structure it is identical with the slate at Mill Cove, — elongated epidotic nodules, often hollow, forming linear groups and continuous lines or bands in the bedding planes. In the outcrop nearest Washington Street, on the west side of Union Street, the slate encloses a layer of white crystalline limestone, six to eight inches thick, and several thinner layers, down to half an inch. The lime- stone is indistinguishable from that at Mill Cove. Elsewhere the presence of lime is indicated by the epidotic lenses and shells ; and igneous metamorphism is shown also, near the granite, by the hard and flinty character of the slate. Between Edwards and Baxter Streets, west of Union Street, according to information obtained from a cellar and well digger, the slate is covered by only from four to six feet of drift. Extending the boundary between the granite and slate east- ward from Newcomb Place, without change of direction, brings us to Weymouth Fore River at a point, a little south of Washington Street ; and this is so far to the north of the Mill Cove section, with which, at least in a general Avay, the Washington Street beds must be correlated, that it has appeared best to recognize a transverse fault between the two areas ; and this, in turn, is conveniently regarded as a north- ward extension of the Fore Kiver fault ah-cady postulated to explain a break in the southern boundary belt. The A.dams Street Area. — Goinij westward from Union 425 Street on the line of the granite-slate contact, across Town Brook and the raih'oad, brings us to the southeast end of President's Hill, — a long, narrow, l)ut very typical and shapely dnunlin. lletween Town Brook and the rail)'()ad is a very perfect development of the Quincy plain, on which the oldest and densest part of the city stands. So far as I can learn by diligent inquiry, no excavations for sewers, wells, and the like have ever reached bed-rock in this area ; but follow- ing the topographic indication we may fairly assume that the bed-rock is slate north of the line of the granite-slate contact. The President's Hill drumlin, on the other hand, is quite cer- tainly built upon ledges of granite, as may, indeed, be clearly seen along its southwestern base. In the valley between Presi- dent's Hill and Cranch Hill, its larger companion drumlin, the normal granite is exposed northward nearly to the northern end of the hill. Some five hundred feet farther north, across the northern end of the hill, the massive gray Cambrian slate is well exposed with the normal east-west strike. The metamor- j)hic character of the slate and the facts observed elsewhere along this boundary, both to the east and to the west, indi- cate that the gap between the ledges of granite and slate is occupied, mainly at least, by fine granite ; and the actual contact of the slate and granite is probably near the most southerly outcrops of slate. It thus becomes reasonably cer- tain that the granite-slate contact is, in the vicinity of the Old Colony Railroad, shifted to the north approximately half a mile ; and since there is in the strike of the slates no sugges- tion of a flexure, it appears necessary to postulate .a north- south fault, which thus falls in line with and may, perhaps, be connected with the very similar displacement of the southern boundary belt on the line of the railroad in North Braintree. Tlie Adams Street area embraces that part of the northern boundary belt between the main line of the Old Colony Rail- road and the West Quincy Branch ; but the outcrops of sedi- mentary rocks are wholly confined to the eastern half of the 426 area, closely bordering the part of Adams Street east and south of Furnace Brook. We find here two quite distinct groups of Cambrian outcrops, — one north of the drumlins (President's and Cranch Hills) and the other on the northern slope of Mt. Ararat, the high northern end of North Common Hill. The first group afibrds the best general section, and the second reveals very perfectly the relations of the sedimen- tary rocks and the granite. In the first or eastern group the outcrops of Cambrian slate begin, as previously noted, at the north end of President's Hill, in the angle between Adams and Goffe Streets. The most southerly exposure is on the west side of GofFe Street, just, south of Glendale Road, and some six hundred feet south of Adams Street. From this point the ledges, separated onl)^ by short intervals, extend in a west and northwest direction almost to Adams Street ; and north of Adams Street, west of the abrupt bend, the slate, essentially continuous with the ledges south of the street, is exposed for a breadth of 420 feet, the total breadth of the Cambrian outcrops being, apparently, nearly one thousand feet. The slate is throughout a dark gray variety, without a trace of red or brown, and with no distinct calcareous segregations, to suggest that it may possibly be Lower Cambrian. It resembles the slate on Washington Street near Pond Street, but not that on Union Street ; and in the absence of all more decisive evidence may be referred provisionally to the Middle Cambrian. Southward, or toward the granite, it is hard, mas- sive, semi-metaniorphic and obscurely bedded ; but toward Adams Street the bedding is very plain, showing a fine and even banding. The strike is N. 65°- 75° E. and the dip steadily to the south, 80°- 85°. The metamorphic character of the slate certainly increases toward the granite, suggesting an igneous contact ; and one doubtful apophysis of granite was observed, although the nearest undoubted outcrop of granite is about four hundred feet distant. 427 Northward, in the vicinity of Furnace Brook, the slate is abruptly succeeded by a heavy bed of normal conglomerate or pudding stone, composed of water-worn pebbles, one to several inches in diameter, of quartzite, felsite, normal and fine granite and slate. The conglomerate has a total breadth, on both sides of the brook, of 450 feet; and it is succeeded northward, wdth apparent conformity, by the distinctly banded, vertical, dark gray slates of the newer or Carboniferous series, with an exposed thickness of 300 feet, and to be, with the conglom- erate, more fully described in a later section. To the westward, the conglomerate iDasses wholly south of Furnace Brook, and, still bordered by the newer slate on the north, can be traced across the northern base of North Com- mon Hill, nearly to the junction of Adams and Common Streets. It is here largely composed of rather angular frag- ments of fine granite in a very firm and flinty paste, and appears to be closely bordered on the south by fine granite, the thousand feet of Cambrian slate intervening half a mile to the eastward being, apparently, wholly wanting here. Going north from Adams Street, midway betw^een Common and Whitwell Streets, we have fine granite outcropping obscurely for three hundred to four hundred feet, followed by two hundred to over four hundred feet of conglomerate, narrowing rapidly w^estward, and about a hundred feet of slate, disappearing northward beneath the swampy margin of the brook. The slate strikes N. 60°- 70° E., with ver- tical dip. The most uncertain factor in the section is the contact between the conglomerate and granite. It is practi- cally impossible to locate this satisfactorily, partly because of insufficient outcrops, and partly because ,the conglomerate is largely composed of granite debris in an almost crystalline paste, but mainly because the granite itself has, near the contact, a more or less marked fragmental structure, having been locally reduced to a crush breccia which is almost indistinguish- able in these weathered and lichen-covered outcrops from the 428 true cons^lomerate. These facts ^ — the cuttino; out of a o-reat body of Cambrian slate, the brecciation of the granite, the rapid narrowing of the conglomerate, and the exceptional induration of the conglomerate, as if by hydrothermal action, — seem to be most readily and satisfactorily explained by an important longitudinal fault between the granite and conglom- erate, with the down-throw, of course, to the north, and increasing to the westward. This is the northern boundary fault of the Blue Hills Complex : and eastward it may be regarded as passing between the Cambrian slates and the conglomerate. Where the fault crosses the northern slope of North Com- mon Hill, the Cambrian beds were not lifted wholly above the present plane of erosion, but numerous remnants of them are still left sticking in the granite, — roots of a once continuous body of slate. The brushy and ledgy slope between Adams Street and the quarries at the top of the hill (Mt. Ararat), and between Common Street on the west and Whitwell Street on the east, is in this respect of very special interest, affording in the most lavish profusion the essential facts concerning the original relations of the Cambi'ian sediments and the granite, which elsewhere are exposed so meagerly or wholly concealed. The special outcrop map of this area (PL 21) has been worked out with great care ; and it shows as accurately as the scale per- mits the position and outline of every ledge and the structural relations of the granite and slate. A glance at this map suf- fices to show that this area is, on a small scale, an ideal com- j)lex, the granite, which is in the main of the normal variety and of medium coarse texture, liaving eaten its way upward with extreme irregularity into the vertical Cambrian beds. The slate is throughout extremely massive, obscurely strati- fied, and semi-metamorphic ; but the bedding can be made out in almost every outcrop. The strike is N. 80° E., and the dip S. 80°- 85°. The perfect and constant orientation of the slate, even where occurring in quite small and isolated masses. 0«as. PjpEf^ S.i5lon Soc. Nat. Hist.. V.il. IV. OUTCROP MAP SHOWING THE UKLATIONS OF THI£ liRANITE AND SLATt ON NORTH (JjMMON HILL, QUINCY. 429 tells strongly in favor of the view proposed here that tlie exist- ing relation of the granite and slate is not due to the mechani- cal intrusion of the former into the latter, but rather to a slow and gradual melting and absorption of the slate by the granite. Mr. White, in describing this slate, which he regards as analo- gous to the German " hornfels " or hornstone, says the feldspar grains are often elongated in lath-shaped forms, intermino-led with large flakes of partially altered and frayed out biotite, some brown fibrous alteration product, and grains of magne- tite. Single large-sized fragments of milky feldspars, par- tially colored by iron, are present throughout. The mica is 1 mcli = 6 feet. 1 inch = 6 feet. 1 inch = 2.5 feet. 1 inch = 5 feet. Figs, slate in 28-32.— Showing in detail the relations of the granite and Cambrian the ledges on the northern slope of North Common Hill, Quiucy. 430 at times quite a conspicuous feature of the slate, and the other secondary minerals include, besides the lath-shaped feldspars, red garnets in very sharp and perfect dodecahedrons up to one eighth of an inch in diameter ; and tourmaline, as dark brown crystals in cavities throughout the hornfels. Although the lines of stratification are often distinctly lenticular in form, this slate, uniformly of a dark gray color, is never appreciably 1 inch = 5 feet. # ■A aW. v5^^ \ r^ #:'*'" •; ■tev"-'' ■■• 1 incli = = 10 feet, 1 inch =8 feet. Figs. 33-35. — Showing in detail the relations of the granite and Camhriau slate in the ledges on the northern slope of North Common Hill, Quincy. calcareous or epidotic, and the best conclusion seems to be, as before, that it is a Middle Cambrian sediment profoundly altered by igneous influences. The ledges of greatest struct- in-al interest are shown in greater detail in figures 28 to 35, which require no additional explanation. According to the interpretation of the general structure of 431 the Adams Street area accepted by the writer, we .should find, if tlie rocks were exposed continuously from North Common Hill to President's Plill and Goffe Street, that eastward the slate patches become rather gradually larger and more con- tinuous, until the granite disappears beneath a continuous cover of slate, the contact being everywhere highly irregular, with the bedding planes of the slate ending abruptly downward against it. It is a necessary corollary of this view that the slate north of President's Hill, notwithstanding its continuity and nearly vertical dip, has, probably, no great depth or vei'- tical thickness near its southern margin, the depth of the granite-slate contact, we may suppose, increasing northward until both rocks are cut off by the boundary fault and the conglomerate. 2^he Quincy -Miltoyi A.rea. — Between North Common Hill and the Granite Branch of the Old Colony Railroad, about three fourths of a mile farther west, the Cambrian strata and the granite-Cambrian contact are entirely con- cealed by the drift deposits and the swampy tracts in the vicinity of Furnace Brook. But immediately west of the railroad, between Furnace Brook and Grove Street, and in the direct line of the contact belt on North Common Hill, a precisely similar complex of granite and slate is exposed in numerous outcrops, as shown in PI. 22. On the southwest, as this special map shows, a prominent wooded drumlin partially conceals the belt ; but ledges of granite and slate, with frequent exposures of the contact, occur at short intervals along the valley of Furnace Brook to Pleasant Street in Milton, and thence westward along Pleasant Street nearly to Randolph Avenue, a total distance of one and a half miles. Thei^e is no evidence of important transverse faulting along this line, and the character of the slate and its relations to the granite are so uniform throughout that this may properly be considered as one homogeneous area. The slate in the group of ledges i-epresented in PL 22 is iden- 432 tical in character with that in the North Common Hill ledges. It is, as a rule, dark gray, massive, hard, and obscurely bed- ded ; usually micaceous or semi-metamorphic ; and it even incloses at one point small red garnets indistinguishable from those on North Common Hill. The dip is unchanged — S. 80°- 85° and sometimes vertical : but the strike is more north- westerly, varying usually between N. 70° and 80° W., the extreme range being N. 60° and 90° W. As before, these I^ig. 3G.— Groiip of ledges in the Quincy-Milton area, between Grove Street and Euruace Brook, showing in detail the relations of the granite and Cambrian slate. Scale : 1 inch =10 feet. 433 masses of slate isolated in the granite are remarkably constant in the orientation of the bedding planes ; and the contacts usually conform approximately with the bedding, showing that the structure of the slate has been sufficiently marked to influence the intrusion of the granite. The details of some of the more interesting ledges are shown in figures 36 to 38 ; and one of these (36) shows an apparent slight departure from the. parallel orientation of the slate. As the outcrop map (PL 22) shows, the granite is normal,. Figp. S7 aid £8. — Showing in detail the relations of the granite and Cam- brian slate in the ledges between Grove Street and Furnace Brook, in the Quincy-Milton area. Scale : 1 inch = 4 feet. except in the more northern ledges, along the line of Furnace Brook, where are considerable masses of fine granite, holding precisely the same relation to the slate. The dike-like mass of fine granite on the north side of the brook, near the west end of the map, lying between slate on the south and normal granite on the north, is twenty to thirty feet broad. The contact with the normal granite can be traced, but it is not a sharp line, the two granites appearing to be blended through OCCAS. I'APEBS B. S. N. H. IV. 28. 434 an inch or two. The extreme nortliern ang^le of this lecl^e of normal granite shows a little slate, and sixty feet farther north, on the north side of the road, the slate is seen passing under the south ano-le of another ledo-e of o-ranite ; while the lar^e ledge of granite west of this, touching the margin of the map, is bounded on the southeast by slate which is exposed for a breadth of fifteen feet. It is evident that these exposures give us the margins only of a body of slate at least one hundred feet wide. Where the brook turns to the northward, near the south Avest corner of the map, the granite crosses the brook and is exposed for over two hundred feet along its southern bank with a breadth of fifty feet ; and the granite is closely bordered on the south for the entire length of the ledge by the cotitinuation, with an exposed breadth of twenty feet and a southerlj^ dip of 80° to 90°, of the slate seen south of the fine granite where the road crosses the brook ; but the fine granite is wanting in the more westerly outci'op; The actual contact of slate and granite, which is at this point about two hundred and fifty feet north of the town line, is not exposed in this ledge ; but it evidently conforms approximately with the bedding of the slate. About five hundred feet farther west the slate, unchanged in character, and with its usual strike and dip, outcrops again in several low, broken, moss -covered ledges scattered over the swampy floor of the valley, on either side of the brook, with a total exposed breadth of 340 feet. Near the northern edge of this group of outcrops the slate is evi- dently much cut by coarse granite with, apparently, some laro;e dikes of granite. A part of one of the small dikes is shown in figure 39 ; and the orientation of the slate has evi- dently been disturbed in this instance, the strike being N. 60° E. Near the southern border of the slate, also, I observed several small irregular granite dikes (one six inches wide), but none clear enough for sketching. A large trap dike appears to run in a general east-west course through tlie mid- dle of the group of ledges south of the brook. 435 Some three liiindred feet west of these outcrops is the private road of the Cuuniiighani estate, and the abrupt eastern end of the beautiful esker on which the house stands. The swampy floor of the valley, probably still underlain by slate, has a breadth north of the esker of nearly five hundred feet. On Fig. 39. — Showing in detail the relations of the granite and Cambrian slate in the valley of Furnace Brook, on the Cunningham estate, Milton. Scale : 1 inch =4 feet. the northern edge of the swamp, probably in situ, is an obscure exposure of granite ; and a little farther west is the private road passing the large barn between the swamp and Edge Hill Road. Immediately north of the swamp this road crosses the west side of a prominent ledge which shows, north- ward, 10 to 15 feet of granite followed by 120 feet of slate dipping S. 80°— 85°, and indistinguishable from the slates pre- viously described. The contact is seen only at one point ; but it must be irregular, as the granite clearly crosses the strike of the slate. The northei"n edge of the outcrop is 350 feet from Edge Hill Street, along the private road. Crossing the ledge diagonally (Fig. 40) is an irregular dike of rather coarse granite, 4 to 10 feet wide, with no appreciable gradation in the granite near the clearly exposed contact. From this point the course of Furnace Brook is southwest- ward, toward the Blue Hills : but the greater part of the swamp turns to the northward and crosses Edge Hill Road, whence the drainage is noi'thward into Unkety Brook and the Nepon- set. Westward, followino- the o-eneral trend of the contact belt, the outcrops are interrupted for half a mile ; but half that 436 distance to the northwest, across the small tributary of Unkety Brook just mentioned, brings us to a Cambrian outcrop which Fig. 40. — Showing dikes of granite cutting the Cambrian slate between Furnace Brook and Edge Hill Road, Milton. Scale : 1 inch= 65 feet. is approximately nine hundred feet north of Edge Hill Road, one hundred and forty feet southeast of Pleasant Street, and five hundred and fifty feet east of the bend in Pleasant Street, where it turns due south. The outcrop is 75 feet long, 15 feet wide, trends east- west, and consists of a massive gray slate of quite uniform character. The slate is somewhat shaly (imperfect cleavage), but otherwise the stratification cannot be made out. The cleavage strikes E.— W., and dips N. 75° approximately. No granite or other igneous rock is exposed here. Between two hundred and three hundred feet east of this ledge, and about one hundred and fifty feet further north, is a small outcrop of the arkose sandstone of the Carboniferous series, which outcrops again on the east side of the little north- ward flowino- brook : and some two luindred feet northeast of 437 this l.'ist exposure, where the brook erosses the private road, near Pleasant Street, is purple and green slate of the newer (Carboniferous) series, dipping; north, with an exposed breadth of about one hundred feet. These are the only outcrops I have been able to find north of the Cambrian-Granite belt, between the Granite Branch Railroad and Pleasant Street ; but bowlders of arkose and conglomerate are thickly scattered over the entire tract, both north and south of Edge Hill Road. The boundary fault undoubtedly lies between the northern- most Cambrian outcrops and the arkose, and it is, therefore, quite definitely determined in the vicinity of Pleasant Street. Turning west on Pleasant Street at its junction with Edge Pleasant Street Fig. 41. — Relations of granite and Cambrian slate on Pleasant Street, Milton. Scale: 1 inch =100 feet. Hill Road, we come, after about nine hundred feet, or just beyond the first bend, to a small ledge, on the- north side of the street, of the typical, dark gray, massive Cambrian slate. Between the second and third bends in Pleasant Street, and opposite the two short streets running off from Pleasant Street on the south, is the group of ledges shown in figure 41. It is apparent at a glance that the complex of granite and slate still continues, and is nowhere more typically 438 developed than here. The ledges at the noi'theast corner of the enclosure are a perfect tangle of granite and slate ; and in the large ledge east of the enclosure are two distinct dikes of granite cutting the slate obliquely, one two to four feet wide, and the other three to five inches. The slate in these ledo;es is of the usual character, and dips S. 80°- 90°. Some three hundred to four hundred feet farther west a dike of trap (diabase) fully 40 feet wide, and trending N. 60° to 70° W., cuts both the slate and granite, the normal granite being exposed for a width of 20 feet on the north side of the dike. It formerly cropped very boldly, but has been levelled by blasting during the recent widening of Pleasant Street. The same improvement has shown that the complex of slate and granite underlies the street for a considerable distance : and both rocks are now well exposed, but not in contact, in the bank along the south side of the street, the massive gray slate showing for about 135 feet. Some 350 feet south of Pleasant Street, just before we reach the school-house at the corner of Gun Hill Road, is a small quarry, in the northern edge of which the granite incloses a mass of slate 6 to 8 feet wide and 45 feet long east- west, with sharply defined contacts, the coarse, normal granite showing absolutely no gradation in texture or composition against the slate. This is the last Cambrian exposure along the north side of the Blue Hills. Westward on this line for a distance of one and a half miles both the granite and slate are wholly con- cealed by the drift deposits. The outcrops of arkose on Gun Hill Road, Randolph Avenue, and Reedsdale Road, to be described in the section on the Carboniferous strata, enable us, however, to trace the approximate northern margin of the Complex half way across this gap. On the west side of Asylum or Town Farm Lane, which runs south from Canton Avenue east of Pine Tree Brook, and one eighth to one fourth mile south of Canton Avenue, is plenty of granite, and some unmistakable ledges with large 439 blocks of conglomerate resting upon them. West of Pine Tree Brook, the fine granite characteristic of the extreme margin of the complex is exposed on the south side of Canton Avenue, just west of Robbins Street, where a wood road runs south ; and two hundred to three hundred feet farther west, the fine red o;ranite shows under the sidewalk on the north side of Canton Avenue, apparently overlain by arkose ; while 450 feet north of the avenue, on the west side of Robbins Street, the arkose crops prominently. About five hundred feet soutliwest of these ledges of fine granite the coarse normal granite rises boldly and forms an extensive, elevated, ledgy tract, closely bordering Canton Avenue on the south for more than half a mile. Beyond this the granite, also, is hopelessly concealed beneath the drift ; and the boundary of the Complex would be entirely lost but for the fact that slight traces of the arkose, on Milton Street, north of Blue Hill Avenue, probably crossing Brush Hill Avenue midway between Blue Hill Avenue and Paul's Bridge, indicate that it still pursues its westward course. Taking a general view of this boundary from the Old Colony Kailroad to the Neponset, its course is seen to be nearly due west as far as Pleasant Street, in Milton, and then it begins to curve to the southward, gently at first, and more rapidly beyond Robbins Street. Although the relations of the gran- ite and Cambrian slate are throughout, so far as exposed, exceedingly intricate, — a typical complex, — the boundary of the Complex itself appears to be a remarkably simple and nearly straight line ; a fact which harmonizes well with the theory that it is a fault line. The facts sustaining this theory will be more fully discussed in the section on the Carboniferous series. THE INTERMEDIATE AEEAS OE CAMBRIAN STRATA. Besides the main down-foldings of the Cambrian strata along east-west lines, which, becoming thus lines of weakness and 440 consequent faulting, have determined the general outlines or boundaries of the Blue Hills Complex as differentiated from the rest of the great batholite of Eastern Massachusetts and isolated by the bordering Carboniferous strata still connecting the tracts depressed by faulting, there are several intermediate and minor belts conforming in trend with the boundary belts, and like them attended by more or less important longitudinal displace- ments. These isolated Cambrian remnants differentiate the orographic block determined by the boundary faults, and make of it a true complex of plutonic and sedimentary rocks. Al- though it is convenient to speak of all these belts as downfold- ings or deep, closed synclines of the Cambrian strata, they are really of less definite structural value, being simply the portions of the once continuous cover of closely folded sediments pene- trating most deeply into the batholite. In other words, they represent the rugosities or most prominent inequalities (com- parable with the teeth of a comb) of the under surface of the batholite cover ; and they are residuary with reference to both surface erosion, and the original corrosion and solvent action of the plutonic magma. This absoi'ption of the covering sediments by the magma must have been highly differential in a chemical sense, and the fact that the roots of the Cambrian series now belong chiefly, apparently, to the Middle Cam- brian, is probably not due wholly to the inferior and hence more exposed position of the Lower Cambrian, but partly to the fact that its mixture of calcareous and siliceous sediments rendered it more fusible. The TIayward Oreeh and Eldridge Hill Area. — This area embraces the mouth of Haywai-d Creek, on the line between Quincy and Braintree, and, extending westward across Howard Street and over Eldridge Hill, it tapers to a blunt point on the western brow of the hill, the exposed length of the belt being approximately half a mile, and its maximum bi'eadth, east of Howard Street, about one thousand feet. The northern boundary is straight (a very probable fault line) ; 441 it trends east-west, parallel with the strike of the slate ; and the slate is separated from the fine granite along this line by a prominent dike. The southern boundary cuts obliquely across the slate, with a south-southeast trend, and is throughout a typically igneous contact. This limited area is, more than any other, classic ground for the Cambrian strata of the Boston Basin, since it was in the little quarry at the mouth of Hay ward Creek, on the southern or Braintree bank, that, as noted in detail by Mr. Grabau in the section on the paleontology of the Cambrian terranes, fossils were first found in situ, and brought to the notice of the scientific world by Professor William B. Rogers in 1856, and during the last forty years this locality has prob- ably received more attention from students of geology than all the other Cambrian outcrops of the Boston Basin taken together. It was in this area, also, and within a few yards of the Paradoxides Quarry, that Dr. M. E. Wadswoi'th,^ in 1881, brought to light, for the first time, incontestable proof that the granite is younger than and intrusive in the undoubted Cambrian slates. The Trilobite or Paradoxides Quarry (PI. 23) is about three hundred feet from the southern margin of the area, and shows to orood advantaoje the massive structure characteristic of the Cambrian slate near an io-neous contact with the o-ranite. The slate is here a gray or greenish gray, compact and flinty variety of remarkably uniform character. Minute grains of pyrite are jDi'etty generally diflfused, with, very rarely, more conspicuous developments of cubic crystals. In the quarry the slate is apparently only coarsely jointed ; but the weathered surface reveals much finer division by this means. The strati- fication is very obscure, and can be determined with certainty only by means of the trilobite remains. These may be supposed to lie in the plane of the bedding ; and according to this indi- cation the strike, as determined by the writer twenty years 1 Proc. Boston soc. nat. hist., vol. 21, pp. 274-278. 442 ago, when the quarry was being worked and examples of Paradoxides were clearly exposed in its walls, is E.— W., and the dip S. 80°— 85°. The higher dip accords best with the other outcrops of this area, as on-Eldridge Hill and in adja- cent areas. In fact, so extreme was the plication of the Cam- brian sediments of this region that an undoubted dip below sixty degrees has nowhere come under my observation among- the sedimentary areas of the Blue Hills Complex. The maxi- mum breadth of slate south of the creek is 350 feet. On the edge of the salt marsh, some three hundred feet south and about one hundred feet east of the Paradoxides quarry, is the obscure outcrop where Dr. Wadsworth noted the contact of the slate and granite. He says (/. c, p. 275), — ' ' The slate near the junction is greatly indurated and changed in color, while the granite is compact and has lost its distinc- tive characters, being transformed into a spherulitic quartz- porphyry. The modification of the granite, the induration and alteration of the argillite, the complete welding of the two formations, and the irregularity of the line of junction, prove that the granite is an eruptive rock, whose period of eruption is of later date than the Primordial argillite." Westward from this ledge as far as Hay ward Creek, the contact is effectually concealed, with no outcrops very near on either side ; but the topographic indications are that the line runs first N. 78° W., then 67° W., and finally N. 58° W., to Hayward Creek, where, on the south bank, eight hundred feet northeast of Quincy Avenue, the contact is nearly, but not quite, exposed a second time. From this point the contact of the slate and fine granite evidently crosses Hayward Creek in a northerly direction ; and it is readily located on the northwest side of How:ard Street, seven hundred feet from Quincy Avenue ; but, again, it is not clearly exposed for observation. Wadsworth states, however, that he found the contact by digging, and that it presents the same characters as before. Continuing in the same general direction, the contact can be traced by infre- w h < m ^' w w cc o Q PC < < >^ ffi <: D a [d h s o S h 443 qucnt outcroj)s of" slate and granite to the top of Eldridgc Hill ; and thence, by frequent and satisfactory exposures of the actual junction, westward across the level summit about five hundred feet to the somewhat abrupt termination of the slate on the western brow of the hill. Near the west end the contact is exposed continuously for ten to fifteen feet. It is, throughout, a good, typical igneous contact. The two rocks are firmly welded, the slate is hard, massive, and very obscurely bedded, and the granite, along the immediate contact, is almost felsitic. Several hundred feet east of the western ter- mination of the slate, a distinct dike of fine granite from two to three inches wide runs into the slate ; and a specimen col- lected for the Museum shows a dikelet of micro-granite from one eighth to one half inch wide traversing the massive green- ish gray slate. At the west end of the belt, the granite ledges advance northward and clearly cut off the slate. The breadth of the slate is here about 90 feet to its straight and simple northern margin, which can be traced in a due east-west direction by abundant and more or less continuous exposures to Howard Street, at the corner of Howard Avenue. Along this line the slate and fine granite, commencing at the west end, are sepa- rated by a vertical dike of fine-grained greenish diabase 20 to 40 feet wide (Fig. 42) . About five hundred feet west of How- ard Street the dike seems to leave the contact and run very obliquely northward into the granite, although there is a smaller dike wholly in the slate near this point. On Howard Street the slate and granite, as I had an opportunity to observe several years ago in an excavation for a water main, are in close contact, with no trap intervening. The trap is un- doubtedly newer than the granite ; and the main dike incloses, on the northeastern corner of Eldridge Hill, a mass of the fine granite some three feet in diameter. The outcrops of slate in this acutely triangular area cross- ing the summit and eastern slope of Eldridge Hill are broad ,444 and numerous ; and in the vicinity of Howard Street we have a nearly continuous section across the belt for a breadth of over five hundred feet. The slate is throughout of a massive, obscurely bedded, hard and flinty character. At some points it is semi-metamorphic in character, being perceptibly crystal- line or micaceous. It varies in color from greenish gray to a dark but distinct red or purplish red. The colors are more or less alternating or interstratified ; but the gray is found chiefly to the south and the red tints chiefly to the north. The strike is throughout approximately east-west, pai-allel with the straight northern margin of the belt ; and that this margin is a fault is quite plainly indicated by its rectilinear character and also by the occurrence along it of the large trap dike, which may be regarded as occupying a fault fissure (Fig. 42). Certainly the dike, which is so clearly newer than both the slate and granite, would be most unlikely to follow so faith- fully an igneous contact, where the slate and granite are inva- riably firmly welded and all tendency in the slate to cleave with the bedding is completely obliterated. The dip, at all points where it can be made out, is approximately vertical, varying perhaps five degrees both north and south of the vertical. East of Howard Street the slate outcrops obscurely in Howard Avenue for nearly three hundred feet ; but beyond this both the slate and granite are lost beneath the level sand plain of Quincy Neck. South of Hayward Creek, the granite-slate contact follows, perhaps fortuitously, a depression ; and this topo- graphic expression of the contact suggests the possibility that it may follow the shore of Weymouth Fore River, which trends east-southeast toward White's Neck in Weymouth for more than half a mile before turning due south toward Weymouth Landing. The shore skirts the base of Wyman Hill quite closely, all the way from Hayward Creek, southeast, south and southwest, to the Monatiquot River ; and the hill, except for the drumlin which crowns it, is undoubtedly wholly composed of fine granite, 445 which is exposed at several points nciivly or quite to the water's edge. Soutlieastward from Hayward Creek, the fine granite outcrops more or less obscurely at several points on and near the shore, whicii is thickly paved with angular fragments of the granite : but tlie slate seemed at first to be wholly wanting. Mr. Watson, however, has called my attention to an obscure but indubitable occurrence of the slate in situ at the north- eastern base of the hill. The angle of the shore where it turns southward is due to a small drift-clad hill of fine granite ; but immediately west of this we pass northward from the granite to a much broken half-tide ledge of slate. Several hundred feet west of this the granite crops again on the same line, but the slate is not exposed. These ledges of slate and granite are on the general line of the contact, as traced from the top of Eldridge Hill to the mouth of Hayward Creek ; and enable us, with confidence, to extend the Hayward Creek and Eldridge Hill Cambrian area eastward under a considerable breadth of the Weymouth Fore River. About two thousand feet of fine granite separate this slate ledge on the northern bend of the shore from the ledo;e on the southern bend, where the shore turns westward again, and where Mr. Watson found the fragment of a large Paradoxides in the massive gray slate. This ledge shows a northeasterly strike ; and the slate belts north and south of Wyman Hill are thus seen to converge eastward. There is, independently of this consideration (see p. 396), a strong probability that the bed-rock of White's Neck is wholly Cambrian ; and hence we might, perhaps, fairly conclude that the original igneous con- tact of the slate and granite follows the western shore of the river, at least approximately, around the eastern base of Wyman Hill. It might be better, however, in view of the pro- nounced east-west trend of the Cambrian belts throughout the Blue Hills Complex, to let the granite terminate under the river, against the north-south fault which we have already introduced to explain the apparent dislocation of the White's Neck and Fore River areas of the southern boundary belt. 446 This Fore River fault is complementary with reference to the Mill Cove fault, separating the King Oak Hill and White's Neck areas, against which we have terminated the fine granite of North Weymouth. The White's Neck fault block is on the downthrow side of both displacements ; and it has thus carried the granite below the present plane of erosion. This view o-ives exceptional value to the part of the great southern bound- ary fault bounding the White's Neck area on the south ; and thus, perhaps, may be explained the exceptional alteration of both the slate and granite along this line and the marked hydrothermal action evidenced by the extensive development of vein quartz in the slate adjacent to the fault. The slate immediately bordering the fine granite of Wyman's Hill, on both the north and the south, is shown by unquestion- able paleontologic evidence to be Middle Cambrian ; and that is the most probable age of the dark gray massive slate on Norton Street in North Weymouth, and inclosed in the granite on Burying Island (see p. 419). Hence it is reasonably certain that the northern end of White's Neck is also Middle Cam- brian ; and since the southern end of the Neck is quite clearly Middle Cambrian, an anticlinal structure is suggested. The strike of the Lower Cambrian beds of the Mill Cove section would carry them somewhat to the north of the Hay- ward's Creek and Eldridge Hill section ; and the calcareous layers and segregations of the red slate of both the Mill Cove and Washington Street sections are wholly wanting in the red slates of the more northern outcrops on Eldridge Hill. I con- clude, therefore, that the red slate of Eldridge Hill marks the passage upward from the calcai'eous red slate of the Lower Cambrian to the non-calcareous greenish gray slate carrying the Middle Cambrian fiiuna ; and this is in accord with the conclusion of Aug. F. Foerste ' that the Mill Cove beds nor- mally underlie the Paradoxides beds. The general relations of this area to the granite are shown diagrammatically in figure 42. 1 rroc. Boston soc. nat. hist., vol. 24, p. 2G2. 447 The Ruggles Creeh Area. — This belt of slate forms the floor of the straight and narrow east-west valley of Ruggles Creek, extending inland a mile, from the tidal portion of the creek, with a maximum breadth of less than five hundred feet. The outline is severely simple and regular, and the north-south profile presents the usual structural contrast, the slate meeting the bordering fine granite in an original igneous contact on the south and a secondary fault contact on the north (Fig. 42). This displacement is remarkably straight, unbroken for the entire length of the valley, and developed topographically, through the more rapid erosion of the slate, Fig. 42. — Diagrammatic section from Hayward Creek nortliward across Eldridge Hill and Ruggles Creek, Quincy. as a smooth and abrupt wall of granite, unquestionably the most ideal fault scarp in the Boston Basin (PI. 24). Going west along the north side of the valley, the granite ledge is continuous, and the fault scarp, commonly ten to twenty feet high, with a steep slope to the south, is well devel- oped west of Howard Street. Immediately west of Chubbuck Street the slate begins to appear in the grassy, sloping floor of the valley below the fault scarp, and outcrops almost continu- ously for half a mile. The slate is mainly of a dark red or reddish gray color, with occasional greenish stripes, resembling the slate of Eldridge Hill ; and, as in that, the bedding is, on the whole, very obscure. The strike is E.-W., parallel with the fault and the valley, and the dip is S. 70°- 80°. The 448 cleavage of the slate is quite marked at most points, trending with the strike and hading steeply to the north. About one fourth of a mile west of ISTorth Street is a small road-metal quarry in the fine granite, showing the contact with the slate directly in the line of the escarpment. Fig. 43 shows the contact as it was clearly exposed in 1891. The slate lies flat against a smooth, straight wall of granite hading S. about 20° ; and for the first six to twelve inches the slate is very much sheared and comminuted. There are also several subordinate shear planes parallel to the fault in the first Fig. 43. — Fault contact of the fine granite and Cambrian slate, on the north side of Ruggles Creek, Quincy. yard or two of the slate, and a very distinct one in the granite nearly three feet from the contact. The slate is gray and mas- sive, probably by leaching near the contact, but becomes fissile and red at a little distance. Farther west the slate is not clearly seen on the north side of the brook, the fault scarp being closely bordered by a wet meadow. At the west end of the slate belt, just before reacliing New Road, crossing obliquely from South Street to Quincy Avenue, the ledges of fine granite advance a little to the south, cutting ofi" some of the slate. The slate ledges come to an abrupt and definite end about one hundred feet east of New Road, and afford a practically continuous section across the belt at its narrow western extremity. Parallel with New Road the slate measures 200 feet from the fine granite on the nortli to the normal granite which here meets it on the soutli ; Occas. Papers, BoslOTi Soc.Nat.HivSt.,Vol.Iv: Plate VA V"^^^V-J^ t?'%^^ .r THE HELIOTYPE PRINTING CO BUTTON FAULT-SCARP, RUGGLES CREEK, QUINCY. 449 and continuing in the same direction, we cross 300 feet of the coarse, reddish, normal granite before reaching Quincy Avenue, The normal granite outcrops west of New Road and again to the northward on Glencoe Street, where it meets and incloses the fine granite, as described on page 353. In the direct line of the slate there are no outcrops for a long distance ; but the rapid narrowing of the slate at the last, and the relations of the fine and normal granites north and south of the slate, make it very improbable that the slate extends much, if any, beyond New Road, the normal granite probably being continuous across its path. The contact of the slate and normal granite on the south is not clearly exposed, a blank of five feet separating the nearest outcrops. The granite shows no appreciable gradation in texture toward the slate ; but that this is a true igneous contact is indicated by the hard and massive character of the slate, and more especially by a distinct dike of granite one and one half inches wide which Mr. Watson and I found in the slate near the contact. A north-south dike of diabase, about eight feet wide, cuts across the slate a hundred to a hundred and fifty feet east of New Road, but it cannot be traced south into the normal granite, nor north into the fine granite, seeming to be confined to the slate, and suggesting the possibility that it belongs to the pre-granitic series of dikes. Eastward from New Road the contact of the slate and nor- mal granite runs obliquely toward Quincy Avenue, and the fine granite begins to appear in its normal place, between the slate and normal granite, with no appreciable gradation be- tween the two granites. The fine granite widens eastward, and thus the rather sharp contact of the normal and fine gran- ites is kept near Quincy Avenue, as the map shows, all the way to Hayward Creek. East of the little brook crossing Quincy Avenue some five hundred feet east of New Road, the slate comes within about one hundred and twenty-five feet of the avenue, and the intervening space is all fine granite. A OCCAS. PAPERS B. S. N. II. IV. 29. 450 little farther east the normal granite appears on the north side of the avenue, and the slate is 150 feet from the avenue, with fine granite between. Beyond this the contact of the slate and fine granite is jogged one hundred and fifty to two hundred feet by two transverse and complementary faults, as shown on the map, the second fault coinciding with the second small brook crossing' the avenue. A good breadth of slate is exposed on the south side of the valley for about two thousand feet east of New Road, or to a point just east of the small quarry where the slate outcrops begin on the noi'th side of the brook. East of this point the outcrops are wholly wanting on the south side, as west of it they are wanting on the north side. South of the brook the slate and fine granite are seen very near each other at several points, but the actual contact is nowhere clearly exposed. That it is an igneous contact may be safely inferred from the hard and massive character of the slate, the bedding being so obscure as to make strike and dip determinations quite unreliable or impossible. In the vicinity of South Street the fine granite forming the immediate south wall of the valley is traversed by a large east- west dike of diabase, very similar to the dike bounding the Eldridge Hill slate on the north. Immediately east of Howard Street Ruggles Creek expands into an estuary, and the bordering ledges, alike of slate and o-ranite, subside beneath the sand plains of Quincy Neck and Quincy Point. But that the slate belt continues essentially unchanged at least as far as Weymouth Fore River, is indi- cated by an obscure but undoubted ledge of fine granite in the frino-e of salt marsh on the north side of the estuary, a quarter of a mile east of Howard Street and exactly in line with tlie east-west fiiult scarp west of Howard Street. Beyond this we can only speculate ; but it may be noted in passing that the belt of fine granite separating the Ruggles Creek and Union Street Cambrian areas probably narrows 451 eastward. This is indicated by the distribution of the out- crops, and by the gradual narrowing eastward and dying out of the axis of normal granite dividinar the fine granite on the west. This plainly suggests the convergence and probable union under Weymouth Fore River of these two sedimentary belts. Topographically, both Quincy Neck and Quincy Point, east of Howard Street, should be relegated wholly to the sedi- mentary rocks ; and the fact that these low and level tracts must be regarded as largely underlain by granite, is one of the puz- zles of the region. The contact of normal and fine granites south of Ruggles Creek belongs to the modified or sub-normal type, as explained on page 356, a displacement, with the downthrow to the north, subsequent to the solidification of the fine granite and prior to the solidification of the normal granite. In the vicinity of New Road, the original westward extension of the Ruggles Creek slate belt and the fine o-ranite borderino- it have thus been lifted above the present plane of erosion, and the contact of the slate and normal granite is readily accounted for. The irregularity and evidently igneous char- acter of both the granite-granite and granite-slate contacts south of Ruggles Creek, make the view that these are ordi- nary displacements quite untenable ; while the modified or sub-normal theory of the granite-granite contact affords an easy and natural interpretation of all the observed facts. The granite-slate contact on the north side of Ruo^o-les Creek is an unmistakable fault ; but the displacement is not, apparently, very great, since at no point has it brought the normal granite up to the level of the present surface, and it seems to die out rapidly as it approaches New Road. Evidence of true faulting wholly subsequent to the develop- ment of the granitic series is not entirely wanting on the line of Quincy Avenue, although it is confined to the normal gran- ite and cannot be correlated with either the granite-granite or srranite-slate contacts. Along the south side of the avenue. 452 west of Howard Street, the normal granite forms a precipitous escarpment 30 to 50 feet high for nearly a thousand feet. This is, conceivably, only a line of master-jointing ; but in appear- ance, certainly, it is an ideal fault scarp. Furthermore, Mr. Watson has found that some of the granite on the line of this topographic break is comminuted and highly oxidized, and that the Eldridge Hill dike is quite certainly jogged or displaced to the northward in crossing this line. That the Quincy Avenue escarpment corresponds to an actual displacement or fault is most probable ; but the facts do not warrant a positive conclu- sion as to the magnitude of the movement. West of the Haggles Creek area the bed-rocks are concealed over a considerable area east of the Old Colony Railroad by the sand plain (Quincy plain) in the valley of Town Brook. Continuous sewer excavations in School and Water Streets and Quincy Avenue, twelve feet deep, have failed to reach the bed-rock at any point. The topographic indications are favorable to a concealed slate area here ; and this view seemed to be confirmed by a study of the drift on the northwest slope of Payne's Hill. Many fragments of the massive, gray Cam- brian slate were observed in the drift ; and at some points the slate forms fully one third of the drift bowlders. But the slate is so similar to that of the Adams Street area, which also is on the line of glacial movement towai-d Payne's Hill and but a mile distant, that it has appeared best, on the whole, to regard the Town Brook slate area as not })roved, especially as there is no indication of the fine granite border which a normal occurr-ence of slate would demand. The Pine Hill Area. — This is the smallest and most completely isolated Cambrian area yet discovered in the Blue Hills Complex. My attention was first called to it by Mr. Watson ; and it has proved ujion investigation to be a very com[)lete area, almost ideal in its relations to the plutonic rocks, and contrasted with all the areas previously described by occurring in that j)art of the complex where quartz por- p!vyry is the (loiiiinant feature of the eontact zone and by the absence of fault boundaries. It is in a brushy wihlerness in the northeastern part of the Pine Ilill plateau, near the rail- road between Braintree and West Quincy, and in a direct line between the junction of Willard and West Streets on the West side of the plateau and Liberty Street on the east side. The smallness of the area has somewhat interfered with its accurate delineation on the general map, while the brushy and imprac- ticable character of the surface and the lack of landmarks have discouraged me from attempting a special map of the Pine Hill plateau. Briefly described, this Cambrian area is a body of slate of compact and simple outline, fifty to possibly a hundred feet wide, and extending from near the northeastern edge of the plateau in a west-northwest direction for more than five hun- dred feet. The rock is wholly massive gray slate, like that of Hayward Creek and the northern boundary belt west of the Old Colony Railroad, and may provisionally be called Middle Cambrian. Although very firm and massive, as would be expected of so small an island in a sea of plutonic rocks, it is yet distinctly bedded at several points, the strike being parallel with the belt — N. 70°- 80° W. — and the dip about vertical. Much of the slate is very siliceous, and it embraces some good slaty quartzite. The contact of the slate and bordering basic porphyry is exposed at several points and is typically igneous, the two rocks being firmly welded together, while the porphyry encloses fragments of the slate, and at the west end of the belt distinct dikes of the porphyry cut the slate. The slate is encircled by successive zones of the plutonic rocks. These are most distinct on the south side, where the normal gradation, as described on page 365, is developed with almost ideal regu- larity. We have first the basic porphyry just referred to, in one of its most perfect and continuous developments. This, with a total breadth of one hundred to two hundred feet, grad- 454 ually changes outward through acid or quartz porphyry to fine granite, which has a similar breadth and blends outward with the widely extended normal granite. The fine granite encloses numerous large segregations of the basic porphyry ; and these extend outward into the normal granite, becoming gradually fiswer and smaller. About one hundred and fifty feet north of this main body of slate, near its eastern end, is a smaller mass of slate parallel with the main belt in dip and strike and about fifty feet long in the direction of the strike. This also o-ives S'ood io-neous con- tacts, and is cut by several dikes thi^ee to six inches wide. The slate also incloses some trap which does not appear to cut the granite and may belong to the pre-granitic series. As would be expected with so small a mass, the zonal arrangement of the inclosing plutonics is less distinct than for the main belt ; although with the latter it holds only for the sides and not for the ends of the belt. From the west end of the main belt of slate the basic por- phyry extends northwest with a breadth of several hundred feet for at least one fourth of a mile to the east end of the small street running east from the junction of Willard and West Streets. This band of basic porphyry is bordered on both sides by the continuation of the fine granite zones, fading outward into the normal granite. In other words, the zonal arrange- ment of the plutonics extends northwest a quarter of a mile, and possibly farther under the drift, beyond the slate in which the zonal structure has its origin. As previously explained (page 365) , this must mean a former extension of the slate itself far beyond its present limits. Erosion has cut here below the bottom of the slate ; and the slate that remains is a mere rem- nant of the original belt. The change of trend is real, though perhaps less abrupt than the map shows it. At the east end of the slate the conditions appear to be essentially similar, but the exposures are less continuous. The basic porpliyry is continuous around the end of the slate, but 455 is itself soon cut off by a great dike of trap which here forms the margin of the phiteau, overlooking the railroad. About one fourth of a mile southeast of the slate, on tiie railroad, opposite Liberty Street, is a magnificently glaciated ledge of medium to normal granite crowded with lai'ge segregations of basic porphyry. It is an exceptionally good exposure of the plutonics south of the slate in their eastward extension, and testifies to the former extension of the slate at least thus far to the eastward, and in this instance without appreciable change in the trend of the belt. The Pine Tree Brook Area. — This is much the largest of the insular Cambrian areas and the one lying most com- pletely within the area of the Blue Hills proper. It forms a highly irregular east-west belt at least two miles and not more than two and a half miles long. It is crossed midway in its length by Randolph Avenue, extending about a mile east of the avenue, and rather more than a mile west of it, ending in the latter direction under the swamp west of Harland Street. It varies in breadth from about five hundred feet at the east end to fifteen hundred feet near the avenue, and possibly half a mile near Harland Street. Yielding more readily than the granite and quartz porphyry to the agents of erosion, it is developed topographically as a longitudinal valley dividing the Blue Hills into two ranges and occupied for almost its entire length by Pine Tree Brook, which finally escapes from the hills through a wide notch in the northern range at the northwestern corner of the valley. In the upper part of the valley, east of Randolph Avenue, the outcrops are frequent and extensive, so that the main structural features are readily and satisfactorily determined (see special map, PI. 26). But west of Randolph Avenue the outcrops are few ; and the lower third of the valley is a continuous swamp, unrelieved by a single exposure of the underlying rocks. That the sedimen- tary rocks continue to some point west of Harland Street, is indicated by the topography and proved by the abundant occur- rence of slate in the drift on the south side of the valley. 456 The only sedimentary rock of which we have any trace in the valley is the dull gray to black slate similar to that of the Pine Hill area, Hayward Creek, and the northern boundary belt, and presumably of Middle Cambrian age. When not too closely involved with the plutonic rocks, as in the central portions of the belt near Randolph Avenue, the slate is usually distinctly bedded, and even, in part, more or less fissile: but near the granite and porphyry, and especially in the narrow eastern end of the belt, the slate is often very obscurely bedded, hard and brittle. However, it exhibits only slightly the micaceous character so strongly developed in the northern boundary belt : and no garnets or other secondary minerals have been noted. Its relations to the plutonics are so intimate that, apparently, w^e must assume that this contrast in meta- morphic character is due to a difference in composition. On the noi'th side of the valley, the outcrops of slate begin about midway between Harland Street and Randolph Avenue ; and from this point to the avenue, and for fifteen hundred feet east of the avenue, the strike is N. 80°— 90° E., and the dip S. 70°- 90°. Farther east, as the special map shows, the con- tinuity of the slate is interrupted by an irregular belt of fine granite and porphyry three hundred to five hundred feet wide : and beyond this the strike is southeasterly for about eight hundred feet and then S. 30° E. for five hundred feet. The dip for these two sections is reversed, being northeasterly 70°- 90°. We have now reached the hill of basic porphyry known as Little Dome, about two thirds of a mile from Randolph Avenue. The slate here regains, with apparent abruptness, its normal strike — N. 80^— 90° E., and passes east- ward along the north side of Little Dome, with a possible breadth of six hundred feet. North of Great Dome the breadth of the slate diminishes from five hundred feet to three hundred feet and less, as it approaches its rather abrupt termination against an oblique (N. W.— S. E.) wall of fine granite a short dis- tance south of the Glencoe granite quarry. As indicated on Occas. Papers, Boston Soc, Nat. Hist., Vol. IV. OUTCROP MAP OF PINE T Plate 26. . ^ + tL. ^j,#?^^ .-^^f""^'^' ''''''^'Kt\'':^^x\\ Ji ♦//=Roc •-A -A E BROOK AREA. 457 the map, the slate probably extends beyond its last outcrop to the intersection of tlie belts of fine granite on the south and on the northeast, at a point between Great Dome and Sawcut Notch. In the vicinity of liandolph Avenue the slate is traversed, pai;allel with the strike and, mainly at least, with the dip, by the quite remarkable series of pre-granitic diabase dikes described on page 388. These vary in width up to forty feet ; but the great dike along the southern border of the slate must be over three hundred feet wide, if all the exposures are to be referred to one dike ; although two dikes one hundred and fifty to two hundred feet wide would also satisfy the observations. The conformity of the dikes with the strike of the slate is especially well shown in their trends changing as the strike changes. The dikes are strictly limited to the slate. In no case do they extend into the granite ; but, as the map shows in several instances, they abut with undiminished breadth against the granite. Even the great dike, the slate walls of which are nowhere exposed, is definitely and absolutely cut off by the granite in the same general longitude as the smaller dikes. The dikes are of very uniform character (see page 388), and more numerous than represented on the map, almost every good exposure of slate in the central part of the belt showing one or more dikes. In its more westerly outcrops, nearly half a mile Avest of Randolph Avenue, the slate is still accompanied by the dikes ; but toward the eastern end of the belt the dikes are apparently wholly wanting, although this may be due in part to the insufficiency of the outcrops. The dikes are undoubtedly older than and cut off by the granite and por- phyry ; and in several instances fragments of similar diabase, one to two feet in diameter, were found inclosed in the granite. Also, in one ledge (Fig. 44) in the central part of the group of dike outcrops, Mr. Watson and I found a small but clear dike of o-ranite traversino- a large dike of diabase. The dikelet is six to seven inches wide at the laro'e end and, with a curvino^ 458 diagonal course, tapers to nothing in a distance of about twenty-five feet. One of the simpler of the many intrusions Granite DiKe Fig. 44. — Small dike of granite intersecting a large dike of diabase in the Cambrian slate of the Pine Tree Brook area. of granite in the slate is also shown in this figure. Unlike the diabase intrusions they do not conform with the strike of the slate, except occasionally by mere chance (Fig. 45). As previously described (page 348), and as the map plainly shows, the granite bordering and intimately associated with the slate is the fine granite ; but it changes rapidly outward or away from the slate to normal granite. At some points, and notably near Randolph Avenue, and in Pine Rock and Little Dome farther east, the coarsely and profusely porphyritic basic phorphyry takes the place of the fine granite as the immediate contact zone ; and throughout the field the fine granite is more or less spotted with patches and segregations of the basic por- phyry. Almost every ledge of fine granite shows some por- phyry, and many ledges ai^e crowded with the porpliyry inclu- 459 sions, which are numbered by thousands, grading into the continuous bodies of porphyry. In the outcrop of basic por- phyry nearest to Randolph Avenue, is a distinct dike of fine granite from three to four inches wide, and I observed another of about the same size in the porphyry between Pine Rock and Little Dome. These contact zone phenomena are equally v/ell developed along the north and south sides and across the eastern end of the slate ; and, as with the Pine Hill Area, and unlike all the other Cambrian areas where the facts are exposed, evidence of a fault contact along any margin of the slate is entirely wanting. On the other hand, the proofs of the intrusive relation of the plutonic rocks to the slate are, in every direction, manifold and convincing. They include, first, the zonal arrangement of the plutonics themselves ; second, the highly irregular general out- lines of the slate so entirely at vai'iance with its own structure ; third, the irregular masses and dikes of fine granite and basic porphyry dividing the slate in every direction, and mainly across the strike ; fourth, the numerous dikelets and apophy- ses of these plutonics penetrating the slate and ramifying through it; and fifth, the fragments of slate inclosed in the plutonics. The belt of granite and porphyry, three hundred to five hundred feet wide, dividing the slate about one third of a mile east of Randolph Avenue, tapers southward, and may be likened to a great plutonic wedge splitting the slate ; or, coming as it does at the point where the first marked change in the strike occurs, it may be called the keystone of the slate arch. Doubtless, if the outcrops were more continuous, this area would present an ideal complex. One of the more regu- lar dikelets of granite is shown in figure 45. It cuts a ledge of slate near the centre of the widest part of the belt. It is ten to twelve inches wide, and is exposed for fifty feet, although probably much longer. At the west end it is parallel to the strike, but eastward it curves to the north across the strike. The dike is nearly vertical, while the slate dips S. 70°, 460 appi^oximately. The actual contact of the slate and plutonics, wherever exposed, is typically igneous, the two rocks being firmly welded. Eig. 45. — ■ A dike of line granite cutting the Cambrian slate of tlie Pine Tree Broolc area. Scale: 1 inch= 15 feet. Around the eastern end of the slate the actual contact with the main body of granite does not seem to be anywliere exposed, although nearly so at many points, especially across the northern slope of Great Dome, where the slate reaches nearly to the summit of the liill. Nevertheless, the relation of the two rocks is not uncertain, for the slate is traversed in every direction by dikes of fine granite and basic porphyry. The dikes are most numerous in the low ledges north of Great Dome ; but several very distinct ones were observed near the top of the liill. They vary in size from a fraction of an inch to a yard or more, are usually quite irregular and branching, and tlie larger dikes sometimes inclose fragments of the slate. The dark gray slate is hard, brittle, and very massive. In fact, the bedding is so obscure that I could not be quite sure of the dip and strike at any point. Although the breadth of the slate is considerable (three hundred to five hundred feet), it is evidently shallow, every part of the area being near the underlying or surrounding granite, which lias injected the downward-opening cracks of the slate. This shallowness of the vertical strata is one of the most general and important facts of the Cambrian outcrops of the Blue Hills Complex, and points more distinctly, perhaps, than any other to the gen- eral conclusion that a great volume of the formation was absorbed by the plutonic magma. 461 THE BLUE HILLS COMPLEX IN CARBONIFER- OUS TIMES. The exact geological dates of" the different series of igneous rocks formed during the Cambro-Carboniferous interval, and their relations to erosion, must remain, apparently, in the absence of contemporaneous sediments, a matter of conjecture, unless we incur the risk of correhation with the similar rocks of the Maritime Provinces. But the beginning of Carboniferous time gives the student of our local geology once more a well- determined hoi'izon within his own field, — a new base-line or point of departure for the study of the later periods of the geological history of this region. Strictly speaking, the Sub- carboniferous formation, including the Mountain Limestone of the Mississippi Valley, the Vespertine series of the Appalach- ian region, and the Lower Carboniferous limestones, shales and gypsum of the Maritime Provinces, is, so far as known, unrepi'esented in New England ; and the great conglomerate series underlying the coal measures of this region may, appar- ently, be best referred to the Millstone Grit epoch. The great thickness of the Carboniferous conglomerates, alike in the Appalachian region and in New England and the Maritime Provinces, indicates that this was a period of subsidence, when the stable marine conditions, with flat shores and stagnant lagoons, of Subcarboniferous time gave way to an encroach- ment of the sea upon the land, attended, however, by marked and oft-repeated oscillations of level. Each advance of the sea spreads a layer of conglomerate over the invaded area ; and during the subsequent recession of the shore-line this newly formed and unlithified stratum is, in large part, borne seaward : and so, as tiie process is repeated time after time, there is gradually built uj), delta-like, a great formation of conglomerate. In some such way as this, we may suppose, 462 the Carboniferous conglomerate was spread over Eastern Massachusetts. The conglomerate is composed very largely, in the Boston Basin almost wholly, of the Cambrian quartzite, the different varieties of felsite, and the granites. The abundance of felsite debi'is in the conglomerate is particularly noteworthy ; and, since felsite is by no means the most abundant rock in Eastern Massachusetts at the present time, it is a legitimate inference that, as before noted, it covered much wider areas in pre-Car- boniferous times. The validity of this point is especially obvious when we note the extraordinarily large proportion of felsite peb- bles in the conglomerate in some districts, such as Nantasket, where ledges of felsite are almost wholly Avanting. All this harmonizes well with the effusive and superficial origin of the felsites ; since, so far as we know, they were the newest rocks in this region at the beginning of Carboniferous time. The granites, on the other hand, although covering hundreds of square miles in the vicinity of the Carboniferous basins, have contributed relatively little material to the composition of the conglomerate. Moreover, the granitic pebbles are largely of the finer-grained (contact-zone and intrusive) varieties ; the coarse (normal) varieties, which largely prevail in our granite ledges to-day, being of somewhat exceptional occurrence, save very locally where Carboniferous erosion cut most deeply ; the exposure of the coarse or normal granites requiring, in general, the removal of, first, the felsites, and, second, the finer granites of the contact zone, with, possibly, some slate. In fact the bio- titic normal granite and its contact zone of basic granite and dio- rite, the most truly abyssal rocks of this region, are also the most scantily represented in the conglomerate. It may be, liowever, that, as I have elsewhere suggested, ^ the paucity of the diorite detritus in the conglomerate is due in part to the greater sus- ceptibility of this rock to chemical decay. Next to the felsites, in the order of abundance in the conglomerate of the Boston 1 Occas. papers Koston soc. iiat. hist., vdl. 3, p. 28. 463 Basin, come the Cambrian quartzites, so extensively developed in the country north and west of Boston. These are the most resistant and truly imperishable rocks of the region ; and they probably formed salient features of the early Carboniferous landscape ; in fact, the topographic prominence of the quartz- ites in still earlier times may have prevented their being entirely covered by the outflows of felsite. It is interesting in this con- nection to observe that, as I have pointed out,^ the quartzite pebbles are most abundant and largest in those portions of the conglomerate series lying nearest to the existing areas of quartzite. The controverted question as to whether the Cam- brian slates are represented in the Carboniferous conglomerate must now be answered in the affirmative ; and in explanation of their rather scanty occurrence in this way it may be noted that such soft rocks are little likely, in the presence of the much harder felsites, granites, and quartzite, to form pebbles ; and that the slates, like the granite^, were probably buried beneath the flows of felsite. The oscillations of level during the formation of the Carbonifer- ous conglomerate are marked in all the basins, but especially in the Narragansett and Norfolk basins, by frequent and somewhat extensive beds of sandstone and shale, usually of reddish or greenish colors. The green tints, it is well understood, may be referred to the reduction of the red colorinaf assent — ferric ox- ide — by organic matter ; but it is believed that the red color (and through that the green) has climatic significance. In other words, the colors of these sediments lend important sup- port to the view proposed by Mr. T. T. Bouve^ that during long ages preceding the Carboniferous subsidence the rocks of this region (felsites, granites, etc.) were subject to chemical as well as mechanical decay, under conditions similar to those which now obtain in the Southern States and in the Tropics, but even more favorable to the production of important results. All the iLoc. cit. 2Proc. Boston soc. nat. hist., vol. 23, pp. 29-36. 464 rocks, except, perhaps, the quartzites, were rotted to a great depth, possibly hundreds of feet, and the residual earths were strongly colored with ferric oxide. When the sea finally in- vaded this area it found an immense amount of detritus ready to be washed away ; and the deposition of the Carboniferous strata went forward at a comparatively rapid rate. Where the sea and tributary streams acted directly upon the still undecom- posed rocks, the pebbles were fasliioned to build the conglomer- ate beds ; but the alternating beds of sandstone and shale ai*e in large part chemical detritus which was transported, assorted and deposited by the water without farther essential change ; and the proof of this is found, as indicated, in the prevailing red color. One of the earliest and most important applications of this princi- ple was made by RusselP in explaining the abundant red sediments of the Triassic formation. He was able to show that not only is the argillaceous deti-itus strongly colored, but that it closely invests the grains and fragments of quartz and other materials, so that the red color is very generally imparted to the coarser as well as the finer sediments. In this connection, it is inter- esting to note, also, that the red beds occur chiefly in the lower part of the Carboniferous series of Eastern Massachusetts, and in general where we may suppose that the material has not been transported far from its source. Some of the seemingly exceptional cases are found on examination to owe their red color largely to an entirely distinct cause, viz., the mechanical comminution of intrinsically red rocks, such as the red felsite and the red Lower Cambrian slates. At this period (early Carboniferous), it is improbable that the existing Carboniferous basins were even outlined ; but the Carboniferous sea, subject to the oscillations of level already noted, spread its sediment far and wide over the entire region. Indications are not wanting, however, that the subsidence of the sea-floor was most marked in the latitude of Boston, one of these indications being the almost comj)lete absence of ter- ' 15nll. no. 52, U. S. geol. survey, Wasliinnton, 1889. 465 restrlal plants (Calamites, etc.) in the conglomerate of the Boston Basin ; and a decided differentiation soon manifested itself in the breaking out within the present limits of the Bos- ton Basin of volcanic activity. The eruptions appear to have been submarine, and numerous sheets of neutral and basic lavas (porphyrite and melaphyre, chiefly the latter), with occasional beds of tuff, were interstratified with the conglomer- ates and sandstones. In this way, through the cooperation of aqueous and igneous agencies and the alternation over the same area of beds of conglomerate, sandstone, shale, and con- temporaneous lavas, was built up the great conglomerate series of the Boston Basin, which has a measured thickness in the Nantasket and Hingham sections of more than a thousand feet, and which differs essentially from the synchronous series of the neighboring Norfolk Basin only in the volcanic rocks. The contemporaneous lavas are well developed in the Neponset Valley, on Hough's Neck, in Hingham, and in the Nantasket district; but south of the Blue Hills, in the Norfolk Basin, they are wholly wanting. Therefore although we can hardly suppose that the Blue Hills were in existence at the time when these strata were forming, their future position was thus early defined, in the southern margin of the Carboniferous volcanic area ; and no better concise definition of the Boston Basin at this period can be framed than that it was an area of volcanic activity on the floor of the Carboniferous sea. Lest the vol- canic rocks should be regarded as a serious bar to the correla- tion of the Boston and Norfolk Basins, it may be noted that a similar localization of volcanic phenomena occurs in the un- doubted Carboniferous beds of the Maritime Provinces. The vast development of Carboniferous strata in New Brunswick, Nova Scotia, and Cape Breton is singularly free from effusive igneous rocks, except in a few districts. Dawson says of the Pictou district : ' ' The lowest Carboniferous rocks seen here are conglomerates interstratified with beds of amygdaloidal trap, which have flowed over their surfaces as lava currents, just as OCCAS. PAPERS B. S. N. H. IV. 30. 466 the trap of the Bay of Funcly has flowed over the red sand- stone. Several of these ancient lava streams alternate with beds of conglomerate ; and while their lower parts have by their heat slightly altered the underlying bed, their upper parts, cooled and acted on by the waves, have contributed fragments to the overlying conglomerate. Over these conglomerates is a great series of reddish and gray sandstones and shales." ^ The oscillations of level during the formation of the Carbonif- erous conglomerate series were of a massive character for the region at large, but attended by deformation of the crust in the Boston Basin ; and so it happened that, while the Carbon- iferous land was approximately base-levelled and the main supply of coarse sediments thus cut oft', the extensive flat areas near sea-level and favorable for the formation of coal, partly constructive and partly destructive in origin, were established ■over the Norfolk and Narragansett Basins ; and the Boston Basin became, after the volcanic fires had died out, a trough too deep for the formation of coal-beds, and too remote from ?iny elevated land to receive coarse sediments. Thus we pass gradually and with entire conformity from the conglomerate series to the coal measures in the area of the shallow southern basins and to the synchronous newer slate series of the deeper Boston Basin, the thickness of which can hardly be less than a thousand feet. Where the Blue Hills now stand, or in the region immediately to the southward, the two sets of conditions — deep estuary or bay and marine flats and marshes — must have merged ; and we may aftirm with a good degree of confidence that the true coal measures never extended over the Boston Basin. The Boston and Norfolk Basins, separated by the Blue Hills Complex, three miles in breadth, are overlapping troughs, the one shallowing: westward and formino- the entire northern border of the Complex, and the other shallowing eastward to its termination in the vicinity of Great Pond in Braintree, 1 Acadian Geology, p. 31G. 467 mi(lway on tlie soutliern border. Although it cannot be doubted that the two basins were once continuous across this common barrier, the fact that not a single outlier or slightest vestige of the Carboniferous strata has been discovered within the area of the Complex indicates very clearly that these newer strata have never been, over this area, folded or faulted down so as to become in any sense a part of the Complex ; but they formed, rather, a broad, simple, and, doubtless, much broken, anticlinal cover of the Complex, which erosion has completely removed. The Carboniferous strata do not, however, form a complete frame for the Complex, erosion having, for five miles, from Weymouth Back River to Braintree Great Pond, cut below the shallow eastern end of the Norfolk Basin. THE CARBONIFEROUS STRATA ALONG THE SOUTHERN BORDER OE THE BLUE HILLS COMPLEX. Since the Carboniferous strata south of the Blue Hills belono;' to the Norfolk Basin, and since they encroach but slightly upon the area of the map of the Boston Basin (Plate 13), their chief claims to consideration here are, first, their struc- tural relations to the Blue Hills Complex and the southern boundary fault, and, second, their lithological relations to the corresponding strata of the Boston Basin. It is proposed, therefore, to give particular attention only to that part of the Norfolk Basin east of the Neponset River ; and not to under- take a systematic or exhaustive study of even this limited area, for which, indeed, the infrequent outcrops oiFer little encouragement . The Norfolk Basin is approximately two miles wide where crossed by the Boston and Providence Railroad, one and a half miles in the longitude of Ponkapog Pond, and one mile on the line of Randolph Avenue. Beyond this it appears to narrow more rapidly ; and it is quite certainly not more than half a 468 mile wide in the vicinity of Braintree Great Pond. The southern margin is heavily and broadly drift-covered, and its accurate definition is impossible with the facts now available ; but the northern boundary, fortunately for our present pur- pose, may be readily traced by contiguous outcrops of con- glomerate and quartz porphyry, besides being, at most points, strongly expressed in the topography. The most complete and satisfactory section of this or any part of the Norfolk Basin is afforded by the ledges on and near the Boston and Providence Railroad. It begins on the north in the ledges of very coarse or giant conglomerate near Back Street south of Willow Street in Canton ; and in the broad rocky ridge between Back Street and Ponkapog Brook the section is almost continuous southward for three fourths of a mile to Pecunit Street. The coarse conglomerate, which is undoubt- edly here, as elsewhere throughout the Basin, the basal mem- ber of the series, is almost ideally developed in broad ledges and great bowlders, consisting largely of pebbles from six inches to a foot and even two feet in diameter. Southward the coarse conglomerate repeatedly alternates with and gradu- ally merges into the more normal type, and at last the con- glomerate series, which has a total breadth of about a quarter of a mile, gives way in like gradual manner to the red sand- stone series, which, with many layers and partings of red shale and occasional bands of fine conglomerate, is exposed for half a mile farther. The dip is steadily to the south and steep — 70°— 90° ; it can, however, rarely be observed in the coarse conglomerate. Ripple-marked surfaces in the sandstone show that the strata are not inverted. The outcrops east of Back Street probably overlap some- what, southward, the more northerly cut on the railroad, which embraces, with some interruptions and an approximately verti- cal dip, about four hundred feet of strata. Red sandstone and shale predominate througliout, but with an occasional green layer and an increasing proportion of fine conglomerate south- ward. Near the middle of the section are three bands of im- 469 pure limestone and calcareous slate one to four feet wide and a dike (diabase?) eighteen inches wide. The rocks all show great disturbance — crushing, shearing, etc. Some thirty- eiglit hundred feet south of this is the second cut, about two hundred feet long, of essentially similar composition, — alter- nating sandstones, shales and conglomerates, — except that the prevailing tone is gray instead of red, conglomerate is relatively more abundant, and the dip is S. 70°. As before, the rocks are much crushed and sheared. About a hundred feet farther brings us to a ledge on the east side of the track, showing sixty feet of gray sandstone and shale with the same dip as the last. It was in the greenish gray shales of these highest beds that Woodworth ^ found the compressed stems of Calamites and Sig- illaria, thus confirming and supplementing the proof of the Car- boniferous age of the Norfolk Basin series discovered by Pro- fessor George H. Barton and the writer at Rockdale, nine miles to the southwest.^ More recently Miss C. M. Endicott has found other forms of Carboniferous vegetation in these Canton ledges. A blank of nearly half a mile (2400 feet) separates this exposure from the granite ledges near Railroad Street north of Canton Junction. Neither here nor elsewhere is there the slightest indication of a repetition of the coarse basal conglomerate on the southern margin of the basin ; but all the facts support the view that the general structure of this eastern extremity of the basin is mono- clinal. Although faulting or isoclinal folding may have re- peated the beds to some extent, and thereby increased the ap- parent thickness of the section, the great basal conglomerate, the red beds and the gray beds must be regarded as essentially distinct members of a terrane having an aggregate thickness of at least five thousand feet, and possibly nearly ten thousand feet. .An important displacement along the southern margin of the basin, with the down-throw to the north, is a necessary concomitant of the monoclinal structure ; and this displacement, 1 Amer. journ. sci., vol. 148, pp. 145-148. = A.mei'. .iourn. sci., vol. 120, pp. 416-420. 470 hopelessly concealed though it be by the drift, is best regarded as a continuation to the west and southwest of the great south- ern boundary fault of the Blue Hills Complex. Eastward from the Neponset Valley, the gray beds, through the narrowing of the basin by this obliquely longitudinal fault, and in perfect accordance with the monoclinal structure, are cut out of the section : and east of Ponkapog the red beds, in diminishing proportion, occupy exclusively the southern half of the basin. The ledges of red sandstone and shale on Farm Street, south of Ponkapog Pond, exhibit in part a low dip to the north, but this may be readily explained by their proximity to the boundary fault, the di'ag of which has locally reversed the dip. In contrast with the faulted and narrowing^ southern mar^'in of the basin, the northern margin, in which we now have a more special interest, consists throughout, from the Neponset to Braintree Great Pond, of the giant conglomerate, grading southward throug-h normal conoiomerate to the red sandstone series, with persistent high southerly dips. East of Ponkapog Brook the giant conglomerate passes beneath the high-level sand plain of the Neponset Valley, which has here, south of great Blue Hill, one of its most perfect developments, and reappears two miles to the eastward, in the vicinity of Hoosickwhissick Pond, protruding from the level surface of the plain. South- east of the pond it is well developed, the jjrincipal outcrops be- ing represented on the topographic map (PL 14). It is in part very coarse, while partings of sandstone show that the dip is approximately vertical. The most northerly outcrops are still about five hundred feet south of the most southerly exposures of the quartz porphyry, which forms the entire southern slope of the Blue Hills from Green Street eastward. But the inter- vening ground is quite certainly underlain by the quartz por- phyry ; for west of the pond and in line with its southern shore is a ledge of quartz porpltyry, and other ledges of this rock reach the same latitude half a mile or so east of the pond. Crossing Monatiquot Stream by the Ponkapog Trail, we find 471 in the woods north and northeast of Ponkapog Pond many large and extremely picturesque bowlders of conglomerate, some of which appear to be residuar}^ masses or approximately in situ ; and farther east on this line are some undoubted ledges of con- glomerate, the exposed breadth of the conglomerate belt being somewhat greater than near the Neponset. Following the southern base of the Blue Hills eastward, the ledges and prominent bowlders and groups of bowlders so faithfully indicated on the topographic map (PL 14) are found to be all quartz porphyry for nearly a mile, and the next exposure of the conglomerate ioi situ is found in the high ledge rising steeply from the northern bank of Monatiquot Stream at the point where it crowds most closely against the hills, between six hundred and a thousand feet west of the junction of Randolph Avenue and High Street. The coarse- ness of the conglomerate is impressive. Rounded bowlders of the typical quartz porphyry of the Blue Hills, and identical in character with the rock in the neighboring ledges, from one to two feet in diameter, make up the mass of the rock, while the largest are three and even four feet in diameter ; and nearly all are well rounded or water- worn, being thus readily dis- tinguished, wdiere weathered out, from the drift bowlders. Besides the quartz porphyry, there are in the conglomerate many bowlders of the fine granite into which the quartz por- phyry natui'ally grades, and a surprising abundance of dark Cambrian slate, mainly but not wholly in small fragments. The most western exposure, probably in situ, shows, in layers of sandstone, a southerly dip of 85°- 90°. Immediately north of these ledges the rocks are concealed by till, and the contact with the quartz porphyry is not exposed. But the quartz porphyry ledges between the conglomerate and Ran- dolph Avenue are so nearly in line with the northern edge of the conglomerate as to indicate that we have exposed here the very base of the conglomerate series. The conglomerate is clearly exposed again on this line in High Street at its 472 junction with Randolph Avenue, and in the angle between the two streets. Eight in the ano;le between Hio;h Street and the avenue, and resting on the conglomerate, is a bowlder, about five feet in diameter, of what appears at first sight to be the quartz porphyry conglomerate, distinguished only by being composed of greenish gray masses six inches to a foot in diameter, in a brownish red paste. But a closer examination shows that pebbles and paste are alike, diifering only in color ; and a comparison with the true conglomerate, as exposed near by in High Street, makes it clear that this bowlder is simply a mass of quartz porphyry mottled by diffei-ential oxidation, through the agency of meteoric waters following the joint cracks of the rock. On re-examination of the ledges of quartz porphyry immediately north of the conglomerate, they are found to exhibit the same red and green mottling ; but farther north this character is wholly wanting; and it may, therefore, be regarded, provisionally, as a contact phase of the quartz porphyry. Between the high conglomerate ledge and Monatiquot Stream are many bowlders, and weathered-out bowlder-like pebbles, of the cono'lomerate. South of the stream the cono-lomerate is well developed in ledges and bowlders for a breadth of about one fourth of a mile. It becomes rapidly finer southward, the main body being of rather normal character, with considerable beds of red shale and sandstone interstratified. The red sand- stone series begins near the point, west of High Street, where an attempt was once made to quarry it, and with many alter- nations and partings of red shale, and bands of fine conglomer- ate, extends southward along the line of Randolph Avenue nearly if not C[uite to West Street. Both the conglomerate and the sandstone are occasionally, and very locally, some- what calcareous, as indicated by a cavernous mode of weathering. Following the basal conglomerate eastward from Randolph Avenue, we find only till with bowlders of quartz; porphyry for a thousand feet. Beyond this, for a thousand feet more, is a steep slope, with ledges of quartz porphyry above, and bowl- ders of the same extending down to the swamp and river. But in all of this two thousand feet I could discover no trace of the conglomerate or the mottled porphyry. Beyond this still is an irregularly rounded hill, some five hundred feet in diameter and from 25 to 30 feet high, the southern margin of which is marked Streamside Ledge on the topographic map (PL 14) ; and about seven hundred feet due east of this is a very similar but smaller hill. These little hills, rising like islands in the swamp between the river and the straight southern margin of the Blue Hills, are united by a very perfect little sand plain and separated from the high land to the north by a depression v^^hich is occupied by a chain of little kettle ponds, as shown on the map (PI. 14). These elevations are entirely composed of the giant conglomer- ate, just like that west of Randolph Avenue ; but the exposures are far more extensive and satisfactory. The conglomerate is chiefly made up of well-rounded bowlders of quartz porphyry and fine granite porphyry, from six inches to three and four feet in diameter. No normal granite or fluidal felsite was ob- served. There is, however, a large amount of Cambrian slate of all the different varieties recog'nized in the reoion of the Blue Hills Complex — black, gray, greenish, reddish, etc., and some fine brown sandstone or quartzite of probably Cambrian age ; but no banded slate, such as might represent the newer or Car- boniferous slates of the Boston Basin. Pebbles of the green quartz porphyry are not uncommon in the conglomerate ; but no masses of the mottled i^ed and green rock were observed, al- though it is probable that they occur ; and I could find no peb- bles of melaphyre or trap. Neither could I discover any clue to the stratification of the conglomerate, or any variation or gra- dation in the character of the conglomerate from north to south ; but it is just as coarse and free from visible stratification on the south side of these little hills as on the north side, suggesting 474 that possibly the bedding may be nearly horizontal. The Cam- brian slate debris, as we should naturally expect, is mostly rather fine, frag-ments more than six inches lono; beino- uncom- mon ; but its abundance goes far to prove that the Blue Hills were formerly covered with slate. Resting on the east side of Streamside Ledge are several bowlders of the mottled quartz porphyry ; and north of this hill and the row of kettle ponds is a fine group of ledges of this rock. It begins where the wood road running north (PL 14) passes through the east-west stone wall, about two hundred feet north of the most northerly exposure of the conglomerate, and extends north fully two hundred and possibly nearly three hundred feet. It is identical in character with the bowlder and ledges near Randolph Avenue, and certainly presents a most striking; o-eneral resemblance to a cono-lomerate. But that it must be a greenish quartz porphyry deeply oxidized along the joints, and not a true conglomerate, is proved by : — (1) The remarkable uniformity of material — the apparent pebbles being all of the normal quartz porphyry, with no granite or slate, or anything foreign to the porphyry, and identical, except in color, with the apparent paste or matrix. The only variation noted consists of two angular inclusions of red jasper or fine quartzite, from one to two inches long. (2) The great uniformity in size of the pseudo-pebbles. They correspond closely in size and arrangement with the joint blocks of the normal porphyry, with no small masses two inches in diameter and less filling the interstices. (3) The green masses or pseudo-pebbles rarely touch each other as do the bowlders in the true conglomerate. The mottled porphyry is probably a surface and contact phase, representing, we may suppose, a surface altera- tion of the porphyry dating from the Carboniferous age. The quartz porphyry was probably all greenish originally. The bowlders in the conglomerate are normally oxidized, at least externally, as is also the quartz porphyry in ledges all 475 over the Blue Hills to a greater depth than any artificial exca- vations, except where it has recently (geologically) been uncovered by the erosion of the conglomerate. From these ledges eastward are only ledges and bowlders of normal quartz porphyry for fifteen hundred feet, to where the road cuts the wall six hundred feet northeast of the eastern conglomerate hill. Here is a high ledge of normal quartz porphyry and immediately south of it, on the road, two masses (possibly bowlders) of the mottled rock, with a little fine red slaty material between the greenish nuclei, making the whole seem very much like a true conglomerate with a slaty paste. I concluded, however, that it marks the exact contact of the conglomerate and quartz porphyry, with the paste of the former penetrating the latter a little way where erosion had cut out the oxidized and more or less decomposed material, as it seems disposed to do. On the same surface is a bowlder 18 inches in diameter of the normal quartz porphyry half imbedded among the green nuclei. From what precedes we may con- clude that befox'e the formation of the conglomerate the quartz porphyry was deeply oxidized, — wholly oxidized at the sur- face, and only along joint cracks deeper down. The super- ficial oxidized portion was worn away to provide material for the conglomerate, down to the mottled zone, which was cov- ered by the conglomerate, and protected from further oxidation. Farther east, following the northern edge of the swamp (see PI. 14), are many bowlders of quartz porphyry, but no ledges. About fifteen hundred feet due east of the second little hill of coarse conglomerate, however, is a third exposure of this rock, in the form of a ridge about five hundred feet long east-west, and connected with the northern edge of the swamp by a low, wooded isthmus of drift. This ledge is simi- lar to the preceding, and contains many rounded bowlders of quartz porphyry one to two feet in diameter. A layer of red shale shows a southerly dip of about 60° ; and this is confirmed 476 by a band of normal conglomerate south of the main mass and almost isolated by the swamp. Continuing in the same east- west line obliquely across the swamp and Monatiquot Stream, we find the next and last exposure of the giant conglomerate on the southeast side of West Street, and some two hundred feet north of Great Pond. The dip cannot be made out here. North of the ledge are rounded bowlders of quartz porphyry, probably representing, in part at least, a disintegrated mass of the conglomerate. So far as I can discover, this is the most easterly outcrop of the Norfolk Basin ; and it is of interest in this connection to note that Mr. Fuller's study * of the distribu- tion of the Carboniferous rocks in the drift points very plainly to the same general conclusion, viz., that the Carboniferous strata of the Norfolk Basin terminate definitely and with appar- ent abruptness in the vicinity of Braintree Great Pond. The quartz porphyry floor of the Norfolk Basin also comes to an end in this longitude, or, if it extends east of Cedar Swamp, is covered, as previously suggested, by the compact and fluidal (eflfusive) forms of aporhyolite. The extensive and continu- ous swampy tract in the direct course of the Norfolk Basin east of Great Pond would ordinarily be regarded as a suflficient explanation of the absence of outcrops, if not of erratics ; and it is a satisfaction, therefore, to find that the testimony of the quartz porphyry and the drift is, to some extent, supplemented and confirmed by the stratigraphy of the Carboniferous beds. On the northwest side of Great Pond, in the angle between the pond and Monatiquot Stream, we have, first, large bowl- ders of the coarse conoiomerate, followed southward bv bowlders and apparent ledges of greenish and red, normal and fine cong^lomerate and o-rit. After about two hundred feet of these coarse sediments, we come to bowlders and some good ledges of the red sandstone series, which is quite satisfactorily exposed for a breadth of eight hundred feet, with many alter- nations of sandstone and bright red shale, before finally disap- 1 Proc. Boston soc. nat. hist., vol. 28, pp. 251-264. 477 pearing beneath the drift. A strongly marked cleavage dips N. 80°— 85°, but the true dip is prevailingly south about sixty degrees. At several points (probable synclinal axes), how- ever, the southerly dip completely subsides, and we note instead a westerly dip of five to ten degrees. This westward pitch is, apparently, best explained as due to an endwise tilting of the formation ; and for that we may find a simple and sufficient cause in the drag of an important transverse fault with the down-throw to the west. This interpretation of the structure satisfies and reconciles all the latest essential facts, including the abrupt termination of the Carboniferous strata ; and it has, accordingly, been observed in the construction of the map. According to this view, the original eastern extension and gradual termination of the Norfolk Basin was lifted above the present plane of erosion. In a general view of this part of the Norfolk Basin, no facts stand out more clearly than that the giant conglomerate is the true basal member of the Carboniferous series, that it was deposited over the partially denuded surface of the Blue Hills Complex, that this original floor of the Carboniferous sea is preserved in the contact of the giant conglomerate and the quartz porphyry against which it rests, and that above this floor the general structure of the Norfolk Basin is monoclinal through a great thickness of overlying strata, including both the red and gray sandstone series ; and, finally, the conclusion cannot be avoided that these thousands of feet of coarse sedi- ments once arched over the Blue Hills, areas from which they have recently been denuded being indicated to some extent by the diflerential oxidation or mottling of the quartz porphyry, a surface alteration which must, apparently, date from Carbon- iferous times. One instance of this mottling not previously noted occurs on the west side of Green Street, southwest of Little Blue Hill and nearly half a mile north of the ledges of coarse conglomerate on Back Street in Canton, If this occur- rence is rightly interpreted, it goes far to prove that above the 478 present plane of erosion the Carboniferous strata arched rather srentlv over the now denuded axis of the Blue Hills. At the southern margin of the hills they plunge steeply down to a nearly vertical monocline, recalling most strikingly the struc- ture of the Front Range of Colorado, in the vicinity of Manitou and elsewhere ; and it is clear in each case that to this strati- graphic break is due the strongly accentuated outlines of the relief. THE CARBONIFEROUS STRATA ALONG THE NORTHERN BORDER OF THE BLUE HILLS COMPLEX. The heavy drift deposits of North Quincy and Milton form a natural boundary between the Blue Hills Complex and the lower Neponset Valley, and offer little encouragement for the investigation of the geological conditions along one of the most critical lines of the region and a comparison of the northern and southern margins of the Complex. In fact the meager data are not easily correlated ; and general conclusions are, of necessity, unduly speculative. The outcrops occur in three distinct and widely separated groups or areas which appear to have little in common, and may conveniently be described separately. The Milton A.rea. — It is intended to include here all the outcrops adjacent to the noi'thern margin of the Complex, in Milton. They are widely scattered along a line three miles in length, which is crossed midway by Randolph Avenue ; and the chief exposures are on or near this thoroughfare. Going north on Randolph Avenue from the corner of Reeds- dale Road about 875 feet across a rather level till-covered area without outcrops, we come to the great dike of diabase pre- viously noted (page 438) as cutting the Cambrian slate and ffr-anite on Pleasant Street east of Gun Hill Road. On Ran- 479 dolph Avenue the exposed bresidth of the dike is 125 feet ; and it here probably marks the boundary between the complex of granite and Cambrian strata on the south and the newer and presumably Carboniferous strata on the north. About 75 feet north of the dike, across a well-marked, canal- like depression, we come to a rock cutting which affords an almost continuous section for nearly five hundred feet northward down the slope toward Unkety Brook, terminating at a point about 750 feet south of the intersection of Randolph Avenue and Gun Hill Road. The entire leno-th of this section is throuo;h a coarse, gritty, white sandstone or quartzite with frequent thin layers or partings of purplish brown slate. Viewed more closely, the sandstone is seen to be, in the main at least, of distinctly arkose character. The feldspathic element is now, however, generally kaolinized ; but in spite of this some of the rock is quite granitic in its aspect, and there can be no doubt that the granitic rocks of the Blue Hills were exposed to erosion during the accumulation of this sediment. The slaty partings show the dip to be constantly to the north, but varying in amount from forty to seventy degrees. Irregular veinlets of quartz are a very characteristic feature of the arkose. The great dike probably coincides, in the vicinity of Ran- dolph Avenue, with the northern boundary fault of the Blue Hills Complex ; and the northward dip of the arkose beds is, perhaps, due to the drag of the fault. The depres- sion separating the dike and the arkose at the surface is a quite remarkable topographic feature. It is steep-walled, 15 to 25 feet deep, 50 feet wide at bottom, and about 100 feet at the top in the narrowest places. Eastward it deepens somewhat, and is quite straight, following the strike of the formations. It holds its form well as far as Gun Hill Road, and there widens out and loses its distinctive character. The arkose is obscurely exposed on Gun Hill Road, north of the entrance to the cemetery ; and the dike, bearing to the east- southeast, enters the granite, passing about five hundred feet 480 north of the schoolhouse at the corner of Gun Hill Road and Pleasant Street. The granite extends fully seven hundred feet north of this corner, or nearly to the bottom of the slope. The boundary fault appears thus, regardless of the deflection of the dike, to hold to its normal east-west course. One fourth of a mile east of Gun Hill Road, in the southeastern corner of the cemetery, the arkose is well exposed again, forming the steep slope on the north side of the brook. It is practically in situ for several hundred feet farther, and is readily traced by bowlders to the point where the brook turns abruptly to the north. Westward from Randolph Avenue the depression between the dike and arkose curves to the southwest slightly and dies out before reaching Reedsdale Road. The dike continues, and is exposed for a breadth of fifty feet on Reedsdale Road, closely bordered by granite in obscure outcrops and bowl- ders on the south, while from two hundred to four hundred feet south of the trap the granite outcrops more freely. About two hundred feet north of the trap, on the west side of the road, the arkose is obscurely exposed, with, apparently, a high dip to the north ; and before the recent grading of the bank the total breadth of the outcrop was nearly three hundred feet. ■ Farther west, across Highland Street, the mantle of drift is unbroken for one and a half miles. The country north and south of Canton Avenue, between Milton Centre and Pine Tree Brook, is strewn with numerous bowlders of con- glomerate, some of which are of considerable size ; and it appears in the highest degree probable that a belt of conglomer- ate is concealed here, not far to the north of the arkose series. The arkose itself appears again on Robbins Street, 450 feet north of the fine red granite on Canton Avenue. The arkose is exposed for a breadth of about seventy feet, but without the slaty partings. The dip is N. W. 70°. Between two hundred and three hundred feet west of Robbins Street, on the north 481 side of Canton Avenue, the arkose appears to overlie the fine granite under the sidewalk ; which would make the total breadth of the arkose about the same as on Randolph Avenue. Farther west on this line indubitable outcrops are wholly wanting. But on Milton Street, about five hundred feet north of Blue Hill Avenue, and one and a fourth miles west of Robbing Street, the arkose occurs in the drift to such an extent as to be practically equivalent to an outcrop. Eastward from Randolph Avenue and Gun Hill Road and the valley of Unkety Brook, the arkose series is next exposed between Pleasant Street and Edge Hill Road, immediately north of the line of the Cambrian slate, as described on page 436, where it is also stated that north of the arkose about a hundred feet of purple and green slate are exposed, dipping north at a hio;h ansle. The rather meager and widely separated outcrops of the arkose series are reinforced by the drift to such an extent as to prove a continuous belt of arkose sandstone extending from Brush Hill Road south of Paul's Bridge to the Granite Branch Railway, a distance of over four miles ; and it is highly prob- able that the arkose belt is paralleled on the north by a belt of conglomerate, with an intervening belt of purple and green slate. The undoubted persistence of the unique and easily recognized arkose leaves little room to doubt that the other rocks are equally constant. The Furnace Brooh Area. — Between the Granite Branch Railway and North Common Hill in Quincy, outcrops are wholly wanting north of the granite : and the modified character of the drift — a level sand plain — makes its indications of the underlying rocks of little value. But in the valley of Furnace Brook, where it runs north of and parallel with Adams Street, north of North Common and President's Hills, the Cambrian outcrops, as described on pages 426—428 are matched by equally satisfactory exposures of the conglomerate and newer slate series. The post-Cambrian or Carboniferous outcrops OCCAS. PAPEllS B. S. N. H. IV. 31. 482 are confined to the immediate valley of the brook, north of Adams Street, and between the point where Adams Street crosses the brook and Hancock Street, the extreme length of this group of ledges being about three fourths of a mile. The outcrops west of the main line of the Old Colony Rail- road have been briefly described, as noted above ; and a little repetition is essential to the completeness of this section. Going north from Adams Street between Common and Whitwell Sti'eets, we have the fine granite, forming the northern margin of the granite and Cambrian complex, outcropping obscurely for three hundred to four hundred feet, followed by two hun- dred to over four hundred feet of conglomerate, widenino; rapidly eastward. The conglomerate is chiefly composed of felsite and of fine granite like that of the adjacent ledges, and the granite pebbles are in part distinctly angular. The essen- tially local and arkose character of the conglomerate is thus plainly indicated. The true arkose of Milton is wanting in this section ; but in the absence of any positive evidence that the arkose changes to conglomerate eastward, it appeal's best not to correlate these two rocks, but rather to consider that the throw of the boundary fault increases eastward until it becomes sufficient to conceal both the arkose series and the band of purple and green slate separating it from the con- glomerate, and thus to bring the conglomerate itself down against the border of the complex. The conglomerate is here unusually hard and flinty, the paste appearing to have been completely baked, or, perhaps, silicified, as if by prolonged hydrothermal action. Mr. White ' has specially noted the metamorphic aspect of the conglomerate, stating that while the prevailing rounded forms of the pebbles prove that it is a true sedimentary rock, the groundmass or paste cementing the pebbles appears in thin sections exceedingly like that of a volcanic tufl^, consisting largely of rather large lath-shaped plagioclases and short stubby hornblende crystals witli scat- 1 I'roc. Boston soc. iiat. hist., vol. '2S, p. Vli. 483 tered grains of magnetite and quartz. As previously stated, it is practically impossible to locate satisffictorily the contact between this altered conglomerate and the granite, partly because of insufficient outcrops, and partly because of the arkose and crystalline character of the conglomerate ; but mainly because the granite itself has, near the contact, a more or less marked fragmental structure, having been locally reduced to a crush breccia. The conglomerate is bordered on the north by dark gray slate striking N. 61°— 70° E., with a, vertical dip, and an extreme exposed thickness of 125 feet. These facts: the cutting out of the arkose series, the inter- vening slate and a part of the conglomerate, the resulting oblique strike of the slate north of the conglomerate, the crushing of the granite and metamorphism of the conglomerate, as well as the lifting above the present plane of erosion of the main part of the Cambrian strata north of North Common Hill, not only demand a fault along the border of the granite- Cambrian complex, but a fault of greatly increased displace- ment in the vicinity of North Common Hill, accompanied by crushing of the granite, and probably also of the conglomerate, as well as by pronounced hydrothermal activit}^ Further east the conglomerate becomes of more normal character, and attains a breadth of 450 feet. Northwest of President's Hill it is crossed obliquely by the brook, and is closely bordered on both sides, as previously described, by nearly vertical slates. The structure was formerly ^ inter- preted as anticlinal, the conglomerate being regarded as a denuded axis, and the slate on the north as a repetition of that on the south. The southern slate is in part at least, older than the granite; and since no appreciable difference nor line of demarcation can be discovered in it, the only alternative appears to be to refer it all to the older or Cambrian series, especially as its lithological characters nve entirely in harmony with this disposition. Although the contact of this slate and 1 Occas. papers, Boston soc. nat. hist., vol. 3, p. 209. 484 the conglomerate occupies a drift hollow and is nowhere exposed, it is, apparently, quite abrupt, and may therefore be regarded as marking the position of the boundary fault. The conglomerate is still a firm, hai'd rock, as indicated by its bold exposures, and grades northward through fifty feet or so of alter- nating fine conglomerate, sandstone, and slate into the main body of dark gray slate, which is very distinctly and finely banded, and, on the whole, strongly contrasted with the Cam- brian slate south of the conglomerate. This slate, which is exposed for a breadth of about three hundred feet, and dips N. 85°— 90°, unquestionably belongs to the same series as the conglomerate. About one fourth of a mile east of the section just described, the intervening ground being a complete blank, is the well- known section on Newjiort Avenue, the Old Colony Railroad and Hancock Street. The Cambrian slate, south of the con- glomerate, is wholly wanting here, as well as the southern part of the conglomerate itself. The exposure of the con- glomerate beo-ins about 150 feet north of the brook, and extends north for 160 feet. It is of irregular composition, mostly rather fine, and largely of a slaty character, a super- abundant arenaceous and slaty paste inclosing small and often scattering pebbles of granite (fine and normal), felsite, mela- phyre and, probably, Cambrian slate ; and it finally passes through some 40 feet of mixed conglomerate, sandstone and slate into the true slate, identical with that west of the rail- road — dark gray and finely laminated or banded. It is exposed for a breadth of 150 feet, — distinctly for 75 feet and then very obscui'ely, on the railroad, for 75 feet more. These beds are in line with those of the preceding section, but all dip S. 80°- 90°, averaging about 87°, the series being thus, apparently, slightly overturned. In part the conglomerate resembles that of the North Common Hill section, but it lacks the flinty hard- ness — the silicified and metamorphic character — of that rock. As before, the conglomerate passes gradually northward into 485 the sandstone; and the sandstone, in turn, seems to blend perfectly with the slate, so that it is impossible to draw a line of demarcation between them. The sandstone is thus a mere bed of passage between the conglomerate and slate ; and in the fact that the slate north of the conglomerate is thus unquestionably a part and almost certainly a newer part of the same series with the conglomerate, we have its strongest and most vital or significant contrast with the Cambrian slate south of the conglomerate : for the conglomerate is clearly newer than the granite, and the granite, in turn, undoubtedly belongs to a much later period than the Cambrian slates which it so freely intersects. That we have in these Furnace Brook sections the beginning only of a broad area of the newer slates, which has been previously assumed to underlie the drift plain of North Quincy, is rendered much more certain by the facts that the deep sewers recently constructed in this part of Quincy encountered slate of the same character at a depth of 25 feet under Hancock Street, about eight hundred feet north of the surface exposure on this street, and a thousand feet north of the brook, and again nearly a thousand feet northeast of this point, in the northern part of Merrymount Park, at a depth of about ten feet below high tide. On the railroad the conglomerate is cut near its northern edge by a trap dike two to three feet wide, and what appears to be the same dike cuts the slate on Hancock Street near the conglomerate with a breadth of at least four feet. The Hough's JSFeck Area. — Eastward from Hancock Street in the latitude of Furnace Brook, the mantle of drift is unbroken for about one and a half miles. At this distance, on Sea Street, the road leading to Germantown and Houghs Neck, at the point just beyond Mt. Wallaston where the sand plain is interrupted by a narrow strait of salt marsh, is a low and solitary outcrop of slate, the only rock exposure of any kind, in an area of more than four square miles. It is a thin, gray slate, with ' perfect lamination cleavage, and occasional 486 coarse gritty bands strongly suggestive of the arkose of Milton. The strikers N. 80° W., and the dip N. 80°- 85°. The north-south breadth of the outcrop is about 275 feet, mostly on the south side of the road. On the southern edge of the outcrop a small, irregular dike cuts the slate approximately parallel with the strike ; and there are indications of a similar dike on the northern edge. Nearly a mile farther east, on the south side of Hough's Neck, is the rocky ridge, half a mile long, encircled by the salt marsh, and known as Rock Island. Directly east of and in line with Rock Island, from which it is separated only by a narrow strip of the salt marsh and a small creek, is a second and slightly shorter and narrower ridge, also isolated by the marsh, which terminates eastward in Rock Island Head. And about eight Imndred feet northeast of this promontory is Raccoon Island, a rocky islet, which is joined to Rock Island Head at extreme low tide. These rocky reliefs are clearly, only the higher and unsubmerged portions of a continuous rocky ridge separating two buried east-west valleys ; and Rock Island and Rock Island Head, at least, evidently owe their prominence to the exceptionally hard and resistant char- acter of the rocks. Rock Island and Raccoon Island afford very satisfactory and virtually continuous exposures of the rocks, and the same is true of the narrow western end of the Rock Island Head ridge for about one third of its length, the eastern two thirds being a smoothly rounded drift hill of drumloid form. As the outcrop map (PI. 27) clearly shows, the dominant and determining oeological feature of Rock Island is the heavv bed of melaphyre, which forms the axis and main mass of the ridge ; and melaphyre is the only rock exposed in the Rock Island Head ridge. The melaphyre forms in either ridge a continuous east- west belt, which varies in exposed breadth from two hundred to four hundred feet in the western ridge, and from two hundred to two hundred and fiftv feet in the CO O Q z < 22 O O O Ou 0- O u H D O •487 eastern ridge. The inelaphyres of the two ridges are practi- cally indistinguishable, by either macroscopic or microscopic characters ; and whether referred in origin to one continuous mass or to separate masses, they must be regarded as litho- logically identical. This rock was first described lithologically by Ur. J. E. Wolff ;^ and his description is substantially confirmed by Mr. White. Dr. Wolff says : — " In the hand specimens the groundmass varies in color from green through gray to red, — the last color characteristic of the rock that is least decomposed. It sometimes encloses large green or white feldspar crystals, often indented by the groundmass, or the feldspar crystals may be comparatively minute ; grains and crystals of epidote are occasionally seen. The rock generally contains greenish spots of epidote and of chloritic material, in part true amygdules, and spots of reddish decomposition. There are also amygdules of calcite and quartz. "As seen under the microscope, the rock is composed of large and small feldspar crystals, magnetite, epidote, calcite, and a mass of chlorite, viridite, and opacite. The large por- phyritic crystals are twinned plagioclase, and occasionally Carlsbad twins of sandin. The minute feldspars of the ground- mass flow around them, encroach upon them, and sometimes break them The degree of decomposition that the feldspars have unders^one varies in the different sections : in the freshest rock they contain immense quantities of minute fluid inclusions, characterized by moving bubbles, and occasional larger ones, rounded or irregular in shape, together with inclusions of the base. The latter are cylindrical, or irregularly reticulated in form, often zonally arranged in the interior or exterior parts of the crystal ; they are absolutely identical in shape, and in their relations to the enclosing crystal, with the inclusions of glass or base of the fresh basalts ; they are now altered to magnetite, 1 Bull. mus. comp. zool. vol. 7, pp. 231-242. 488 viridite, and other products The only other original mineral, unless it be part of the magnetite and apatite, is olivine. This was found in well-marked, large, and unde- composed crystals only near the contact of the amygdaloid with the conglomerate Their relations to the groundmass prove an anterior origin ; some of the magnetite and opacite in the sections have probably been derived from the alteration of small fragments of olivine. Between the large and small feldspar crystals lies a mass of greenish alteration products, — chlorite (often dichroic), viridite, magnetite, opacite, consider- able epidote, quartz, and calcite. When some of the large feldspar crystals diverge, the triangular space between them is filled with very small feldspar crystals lying in this greenish mass, showing, as has been often remarked, that it is merely an original, glassy base, much altered, for we find this same relation in the unaltered basalts. Calcite, quartz, epidote, hornblende, biotite, apatite, etc., in the decomposed base, seem to belong to the more advanced state of decomposition. Magnetite is always present. A large part of the magnetite arises from the decomposition of the base, and it is generally difficult to say what part of it is original. "While in some sections true amygdules are wanting, yet they generally occur, characterized by their sharp boundary, and the arrangement of the feldspars of the groundmass parallel to their outline. They are filled by epidote, chlorite, viridite, calcite, or quartz ; the epidote generally on the outside when other minerals occur with it. Besides these true amyg- dules, areas of decomposition occur in the groundmass, con- sisting either of opaque ferritic material, constituting the macroscopical red spots, or of epidote, chlorite, viridite, etc., enclosing the small feldspars." Dr. Wolff summarizes this description as follows: "It is found by study to be a rock which, in the original state, was composed of the feldspars, olivine, magnetite, a base (glassy, microlithic, etc.) and probal)ly some augite (though this can- 489 not be identified now), all in varying proportions, and that these original constituents have been largely replaced by sec- ondary products. It is therefore an altered basalt," — in other words, a nielaphyre. The mekphyre of Rock Island (PI. 27) and the conglomer- ate bordering it on the south rise abruptly from the marsh on tlie west. The conglomerate dips S. 70°, and varies in the breadth of exposure from a hundred and twenty-five . to two hundred feet ; while the exposed breadth of the melaphyre, which slopes down beneath the drift on the north, varies 'from two hundred and fifty to nearly three hundred feet. These conditions continue for about six hundred feet, and then the contact of the melaphyre and conglomerate is abruptly shifted to the south about a hundred feet, apparently by a transverse fault ; and the exposed breadth of the melaphyre is somewhat increased. The road, where it turns to the southeastward, appears to mark the position of another obliquely transverse fault, which breaks the melaphyre-conglomerate contact again, shifting it to the southward about forty feet. This fault quite certainly crosses the melaphyre, for the northern border shows a similar displacement in line with this. From this line east- ward the melaphyre is closely bordered by conglomerate on the north as well as the south, and its full breadth is seen to be about three hundred and fifty feet. The northern conglomer- ate dips N. 85°— 90°, and is exposed for a maximum breadth of a hundred and twenty-five feet, and an extreme length of about five hundred feet, the contact being exposed in part as an abrupt escarpment ten to twenty feet high. Beyond'this, or about midway of its length, the Rock Island ridge subsides to an undulating and ledgy field, in which the melaphyre outcrops frequently and apparently with diminished breadth eastward. About twenty-two hundred feet from its western end, and five hundred feet from its extreme eastern end, the island is divided somewhat obliquely (see map) by a very narrow strait of the marsh, which almost certainly con- 490 ceals a fault ; for beyond this the melaphyre is wholly wanting, the conglomerate, in almost continuous outcrops, extending directly across its path, the total exposed breadth of the con- glomerate in this part of the field being about eight hundred feet. An alternative view is that the melaphyre gradually narrows eastward and comes to an end in its last outcrop on the edge of this strip of marsh. The distribution of the out- crops is quite consistent with either view. It is improbable, however, that so thick a body of melaphyre (350 feet), whether effusive or intrusive, would die out in so short a distance, and the first explanation appears, therefore, the more acceptable, especially in view of the very clear evidence of transverse faulting in the western half of the rido-e. It is interesting in this connection to note that this hypothetical eastern fault is parallel in trend with the others, but apparently reversed in throw, the melaphyre probably passing to the south, but pos- sibly to the north, of the island of conglomerate. Crossing the marsh and creek to the Rock Island Head ridge, it is seen that another transverse fault, partially compensating with reference to the first, would satisfactorily explain the relations of this sec- ond band of melaphyre to the first band. In the Rock Island Head ridge the melaphyre is exposed for a length of seven hun- dred and fifty feet and a breadth of two hundred to two hundred and fifty feet, with no indications of the conglomerate on either side. South of the melaphyre in Rock Island, and including the elongated, isolated ledge in the marsh, the total exposed breadth of the conglomei^ate is about five hundred and fifty feet. The frequent arenaceous and slaty layers show the strike to be throughout about N. 80° W., and the dip S. 70°. The conglomerates bordering the melaphyre on either side are of very normal character — typical pudding-stone. The northern conglomerate is rather fine, few of tlie pebbles exceed- ing an inch in diameter, and very uniform, tlie only noticeable band of finer sediment being a layer of slate three to five feet thick, with streaks of sandstone, which marks the contact vith 491 the melaphyre for the eastern half of the outcrop. The south- ern conglomerate is iii part distinctly coarser, containing many pebbles from two to "three inches in diameter; and within two hundred feet or so of the melaphyre it embraces much arenace- ous material, — perhaps fifty feet in all of more or less distinct and thin-bedded sandstone. We come now to the interesting question as to the relations of the melaphyre to the conglomerates. In 1880 ^ I regarded the melaphyre (amgydaloid) as an igneous rock belonging to an older series (Shawmut group) than the conglomerate, and as marking an anticlinal axis in the latter. In 1882, Dr. Wolff, in the paper cited on page 487, held, in accordance with the title of that paper (The Grreat Dike at Hough's Neck, Quincy, Mass.), that the melaphyre or amygdaloid forms a great dike, or possibly two dikes, cutting the conglomerate parallel with the strike, and necessarily newer than the con- glomerate on either the north or the south. I long ago became satisfied that the melaphyre is not older than all the conglomerate: and since no one doubts its igneous origin, the real question now is as to whether the melaphyre is older than a part of the conglomerate or younger than all of it ; or, in other words, — Is the melaphyre intrusive, as held by Wolff, or effusive (contemporaneous) — a dike, or a surface flow of lava ? Dr. Wolff, although apparently not regarding the evi- dence as entirely satisfactory, but as demanding a more detailed study, summarizes as follows the facts supporting his pro- visional conclusion that the melaphyre is a dike : ' ' At the western end, on the southern side, the junction Avith the con- glomerate and red sandstone is very irregular, — large and small tongues of the dike penetrate into the conglomerate, this rock having a strike N. 60°- 80° W. and a dip 70° south. The junction between the two rocks is sharp and well-marked ; the dike seems often amygdaloidal near the junction. Sections of the contact of the two rocks show that the dike is composed 1 Occas. papers Boston soc. nat. hist., vol. 3, pp. 176, 209. 492 of a mass of very small feldspars, having a beautiful fluidal arrangement, while they are often bent when in contact with the line of the conglomerate. On the northern side a fine vertical exposure of the junction is obtained, which is seen to stand almost vertical ; the dike cuttino^ the slate and cono-lomer- ate a little irregularly, but standing nearly parallel to the stratification." My later observations indicate, on the contrary, that the melaphyre is unquestionably effusive. The facts which appear to compel this conclusion are, briefly, as follows : — First, the structure of the melaphyre itself, and especially the contrast in structure presented by a north-south section of the mass, suggests a lava flow rather than a dike. Along the north side, or in the northern half, it is a homogeneous, dense, and in general non-amygdaloidal, compact to finely crystalline and porphyrytic rock of gray to greenish gray and purplish gray color. Mr. White, who fully accepts the effusive nature of the melaphyre, notes that, "At the base (north part) of the flow the rock almost wholly lacks amygdules, and seems to be a true basalt," the only traces of amygaloidal structure being occasional round amygdules up to 1 cm. in diameter, of a brownish black, uncertain decomposition product. South- ward and especially along the southern margin, as noted also by Mr. White, the peroxidation of the iron in the lava is more general, and dull reddish to purplish tints prevail ; and the rock becomes upward, and especially near the southern or upper margin of the flow, gradually amygdaloidal and scoria- ceous, the superficial portion of the flow being in part pro- fusely amygdaloidal. Mr. White points out that the mela- phyre at the top of the flow, along the irregular contact with the conglomerate, is of a slaty character and exhibits a micro- fluidal structure parallel with the contact, as described by Dr. Wolff. In short, we have here a perfectly normal and almost ideal section of an ancient lava flow — a single, simple flow, with its base against the northern conglomerate and its sur- 493 face covered by the southern conglomerate. The gradation in texture from dense and crystalline to amygdaloidal and scoria- ceous, is, if possible, more perfect, or more satisfactorily exhib- ited, in the eastern than in the western ridge, although the entire thickness of the eastern melaphyre is quite certainly not exposed. Second, the formal relations of the melaphyre to the con- glomerates, although perhaps not wholly unambiguous, are certainly entirely consistent with the view that it is effusive. The irregularities of the northern contact rioted by Wolff are in part, probably, due to faulting ; and this explanation is especially demanded where, on the west, the melaphyre, for more than a hundred feet, cuts obliquely and by a straight line across the conglomerate, as shown on the map. Where not so explainable, the irregularities are no more than we should naturally expect where a stream of lava over three hun- dred feet thick flowed over the unconsolidated gravelly sedi- ments of the Carboniferous sea. At a point about a hundred feet west of the eastern end of the large outcrop of the northern conglomerate, a vertical dike about a foot wide cuts the con- glomerate at right angles to the contact but fails to penetrate the melaphyre ; its apparent lithologic identity with the mela- phyre suggests, however, that it is probably a tongue or apophysis penetrating the conglomerate from the base of the flow. The phenomena of the upper or southern contact are naturally of greater significance and interest ; and it may be noted that the sharp line of demarcation between the melaphyi'e and conglomerate, and the conformity of the fluidal structure of the melaphyre with the contact, as described by Dr. Wolff, are demanded by the effusive no less than by the intrusive theory. As regards the irregularities of the 'contact, the sup- posed tongues of the melaphyre penetrating the conglomerate and inclusions of conglomerate in the melaphyre, I have been able to find absolutely nothing that is not readily and satisfac- torily explained on the supposition that the conglomerate was 494 deposited over the naturally uneven and scoriaceous surface of a lava flow, and especially of a submarine flow, such as this most probably is. AVhen the slaggy surface of a lava stream is filled and covered by sand and fine gravel, we must expect that the chance and imperfect sections which erosion affords Avill often be puzzling and inconsistent. Third, the composition of the conglomerate tells strongly against the intrusive origin of the melaphyre, and in favor of its eff'usive origin. The southern or upper conglomerate, especially very near the melaphyre, contains numerous large and small pebbles of precisely the same kind of melaphyre. In fact, the layer of conglomerate immediately covering the melaphyre is at some points largely or chiefly composed of this melaphyre debris. On the other hand, the melaphyre pebbles appear to be wholly wanting in the lower or northern con- glomerate, and in the southern conglomerate, as a rule, beyond a few feet from the contact. To summarize, the chief facts proving that the melaphyre is effusive and not intrusive, — a record of volcanic activity in this region contemporaneous with the formation of the con- glomerate, and not of plutonic activity at some subsequent period, — are : the amygdaloidal and scoriaceous character of the upper part of the flow ; the highly unsymmetrical section of the melaphyre, so nor.mal for a flow and so abnormal for a dike ; the highly but minutely irregular form of the upper con- tact, and the compact and fluidal texture of the melaphyre near this contact ; and the abundant occurrence of debris of this melaphyre in the base of the overlying conglomerate and its absence in the underlying conglomerate. That the eruption was submarine is probable from the sim- ilarity of the underlying and overlying sediments, the complete- ness of the section, and the absence of evidence of extensive erosion of the melaphyre before it was covered. That the eastern and western ridges are dislocated portions of one and the same flow is indicated by their lithological similarity, and 495 by their great thickness, which would be inconsistent with basic flows of limited area. That the structure, as now revealed, is not anticlinal, as I formerly supposed, is proved by the un- symmetrical section of the melaphyre or the evidence that it is efl^usive, and by the contrast between the two conglomerates, the lower conglomerate containing no melaphyre debris, and but few arenaceous or slaty layers, while both are conspicuous features of the upper conglomerate. • And, finally, that the strata are not inverted is clearly indicated by the fact that this great bed of lava is still right side up. The outcrops are con- sistent with the view that the melaphyre narrows gradually both east and west from the middle of Rock Island, where it attains its maximum breadth (thickness) of three hundred and fifty feet ; and the diverging dips of the upper and lower con- glomerates suggest also that it may increase in thickness down- wards. At all events, it is improbable that this sheet or flow of melaphyre was ever indefinitely or very greatly extended in the direction of the strike beyond its . present exposures ; and hence there is no occasion for surprise at its non-appearance in the sections on the Old Colony Railroad, Furnace Brook, Ran- dolph Avenue, etc. The interesting complex of diabase dikes traversing the melaphyre and to some extent the conglomerate of Rock Island will be described in a later section. Passino- now to Raccoon Island, which is about a thousand feet in extr,eme length by three hundred feet in breadth, we find that it consists wholly of slate of a very uniform character. The slate is dark gray, distinctly and finely banded, but rather massive, the lamination cleavage not being very distinct ; and in the main it is of rather coarse texture, some layers being almost arenaceous. Some layers, also, are minutely contorted, although the adjoining layers and the slate as a whole are remarkably straight and regular. The strike is E.-W., or N. 85^ W. ; and the dip N. 85°- 90'' ; corresponding closely with the conglomerate north of the melaphyre on Rock Island. The slate outcrops practically all over the island, the exposed 550 feet 350 i i 125 i i 496 breadth (equivalent to the thickness) being nearly three hun- dred feet. The south side of Raccoon Island is about in line with the northern edge of Rock Island, as is also the outcrop of slate already described, on the road a mile west of Rock Island, although the finer and softer texture of the last-mentioned slate would indicate that stratigraphically it belongs further from (above) the conglomerate than the slate of Raccoon Island. In the Rock Island monocline, which is, of course, simply the south side of a great anticline, we have, from the south, approximately, Upper conglomerate and sandstone Melaphyre Lower conglomerate 1025 " The melaphyre seems to agree well in character, thickness, and stratigraphic position (low in the conglomerate series) with the great bed of melaphyre in Hingham, and a provisional correlation is clearly justified. Below the melaphyre of Hing- ham the conglomerate is of uncertain but inconsiderable thick- ness, while above it the conglomerate series (conglomerate, sandstone, and slate) measures in several good sections from eight hundred to twelve hundred feet before we come to the great slate series, the lower part of which, as well exposed on Huit's Cove, is related in its coarse and gritty texture to the slate of Raccoon Ishxnd. These considerations indicate that the anticlinal axis of Hough's Neck is not far to the noi'th of the melaphyre ; but the repetition on the north side of this axis of the melaphyre and the great thickness of the upper conglomerate would throw the slate so far to the north as to necessitate two ti'ansverse faults (one at either end of the melaphyre) similar to those dividing the melaphyre, but of far greater throw, involving a horizontal dis[)lacement of certainly 497 one thousand and possibly fifteen hundred feet. This consider- ation, and the entire absence nortli of the Rock Ishind ridge of any indication, topographic or otherwise, of either the mehi- phyre or conglomerate, suggests an alternative solution of the problem in the form of a longitudinal fault along tiie north side of Rock Island, with a down-throw to the north sufficient to conceal both the conglomerate and melaphyre. It might even be suggested that in the irregular vertical contact of the melaphyre and the lower conglomerate we have some indica- tion of such a fault. Between these two views it is difficult to^ choose ; but since the longitudinal fault appears, on the whole,, to agree best with the topography, it has been preferred in the construction of the map. In looking eastward for a possible extension of this broken anticline, it may be suggested that we -have some indication of it in the coarse gritty slate on the extreme north corner of Grape Island (see Part 2, p. 257),. and the strike, N. 70° E., is distinctly favorable to this view. If, as we may reasonably suppose, the anticline pitches east- ward, the absence of the conglomerate and melaphyre on Grape Island is readily accounted for. Still farther to the northeast, as previously suggested (Part 1, page 104), we find the prob- able continuation of this axis in Bumkin Island, Strawberry Hill, Strawberry Ledge, and Harding's Ledge. The slate on the north side of Grape Island dips S. 70°, evidently passing beneath the fine, soft, dark gray and nearly vertical slate on the south side of Grape Island and on Slate Island and the north shore of Hingham. Although this area south of the Rock Island anticline is undoubtedly characterized by great structural complexity, it must be in a general way synclinal, and hence occupied chiefly, as we see in Slate Island, by the soft, black slates. To the south side of this trough may be referred the coarse, gray, massive slates outcropping on the beach at the end of Sea Street, northwest of Great Hill in Weymouth, with E.-W. strike and nearly vertical southerly dip. In the field immediately south of the beach, the slate, OCCA.S. PAPERS B. S. N. H. IV. 32. 498 now black and fissile, outcrops again with the same strike and dip and an eixposed breadth of about twenty feet. These out- crops have more the character of the newer or Carboniferous slates than of the known Cambrian slates : and since there is a possibility, as described on page 422, that the arkose sand- stone is in situ on Neck Street in Weymouth, it appears best to carry the boundary fault of the Blue Hills Complex as far south as Washington Street in Quincy and Old Spain in Wey- mouth, and to regard the absolutely blank area between this line and the Rock Island anticline as the westward continua- tion of the Carboniferous slate basin. Takino; a o-eneral view of the Carboniferous border of the Blue Hills Complex, it is evident that although the Rock Island anticline is, apparently, continuous for its entire length, west of the Old Colony Railroad, the southern limb of the anticline has been more or less completely cut away by the boundary fault ; while east of the railroad the fault, as sketched in the preceding paragraph, bears to the southward and the synclinal basin of the newer slate is gradually developed between it and the anticline. West of Hancock Street, Quincy, the outcrops are all referable to the northern limb of this rather hypothetical anticline, the beds of arkose, sandstone, conglomerate and slate lying in monoclinal fashion against the northern edge of the Complex ; but on Hough's Neck and Raccoon Island the conglomerate- and melaphyre must be referred wholly to the southern limb and the slate wholly to the northern limb ; and nowhere do we obtain a clear or com- plete section of the anticline, which is evidently much broken bv transverse as well as longitudinal faults. That the struc- ture along this line is essentially anticlinal is a reasonably safe inference from the general fact that throughout the Boston Basin the conglomerate, so far as known, underlies the newer slate, and hence marks as a rule in its outcrops, the positions of anti- clinal axes. Probably the best general summary of the stratig- raphy of this border is that the actual but much broken 499 anticline east of the Old Colony Railroad becomes, west- ward, a northward-dipping monocline, due, perhaps, at least in part, to the drag of the boundary fault. COMPARISON OF THE CARBONIFEROUS STRATA OF THE CON- TIGUOUS PORTIONS OF THE BOSTON AND NORFOLK BASINS. Incomplete and scattering as are the outcrops of post-Cam- brian sediments along the southern border of the Blue Hills, they are still more meager and fragmentary along the northern border ; and in the absence of approximately continuous sections a close comparison, not to mention a close correlation is out of the question, unless it be on the basis of a broader survey of both basins than we are prepared for now. The salient characters of the sediments in the eastern part of the Norfolk Basin are: (1) the prevailing coarseness, as expressed in the predominance of conglomerates, the great development of arenaceous beds, and the comparative absence of slates and shales: (2) the apparent great thickness of the section, which may, however, be due in part to repetition by folding and faulting ; (3) the dominant red color of the lower half of the series ; and (4) the absence of contemporaneous lavas and, in general, of dikes. East of the Blue Hills Com- plex, in Hingham and Nantasket, the conglomerate series, to which alone the Norfolk Basin sediments can be regarded as in any sense equivalent, embraces, besides the interbedded volcanics, many alternations of conglomerate, sandstone and slate, the finer sediments, especially, being chiefly of a reddish color, although the color is in general darker and less pro- nounced than in the Norfolk Basin. In the Nantasket area, in Hingham, and in the Norfolk Basin, the base of the conglomer- ate series and the granite floor on which it was deposited are clearly exposed ; and omitting the volcanic sheets, of which there are many in Nantasket, one in Hingham, and none in the 500 Norfolk Basin, a certain general correspondence can be traced between these three areas ; but the iSTorfolk Basin sediments are unquestionably much thicker, average much coarser, are less perfectly assorted, and more intensely colored than those of the more eastern areas. Passing now to the north side of the Blue Hills Complex, the contrast with both the eastern and southern areas is much more marked ; although the melaphyre of Hough's Neck is an important connecting link between this area and Hing- hara. The base of the conglomerate series is, apparently, not exposed ; red sediments are much less in evidence ; the distinct alternations of conglomerates with sandstones and slates are, apparently, less frequent ; and especially the arkose sandstone series, which is the most distinctive feature of this area, cannot be recognized in Hingham and Nantasket, nor as a sharply defined series in the Norfolk Basin, although the coarser sedi- ments of this basin are throughout more or less of arkose character. That the sediments of these different areas are essentially contemporaneous, can hardly be questioned by any one con- versant with all the facts bearing upon this problem ; and we must conclude, in view of the striking lithologic and strati- graphic contrasts which they present, that in Carboniferous time the conditions of deposition in these closely contiguous areas were strongly differentiated. But this sharp differentiation is inconceivable if we must consider that the Blue Hills were then practically non-existent and the sea spread freely over the entire area ; furthermore, the formation of the giant conglomerate, hundreds of feet thick, along the southern base of the Blue Hills demands a bold shore freely exposed to a vigorous surf; and finally, the great thicknesses of these coarse, shallow water sedi- ments, especially in the Norfolk Basin, compels us to postulate a contemporaneous, progressive and localized subsidence ; that is, a graben. In other words, the demarcation of the orographic block known as the Blue Hills Complex, occurred early in the 501 Carboniferous age ; and the development of the boundary faults went on during the deposition of the conglomerate series. The Complex is divided by the great transverse fault already postu- lated in the vicinity of the Old Colony Railroad into two principal blocks ; and while, as will be more fully shown in the next section, the eastern, block probably experienced a massive and somewhat equable elevation, the western block, embracing besides the Blue Hills proper the area now included in the Norfolk Basin, suffered a north-south tilting, the north- ern edge rising and the southern edge sinking. The depression of the southern margin of this block gave rise to a deep trough into which were rapidly washed : first, the coarse detritus worn by the waves from the quartz porphyry and Cambrian cover of the Blue Hills, to make the basal conglomerate ; second, the highly ferruginous residuary chemical detritus of a wide area to form the complex of red beds ; and, third, after the practical exhaustion of the red soil through the rapid erosion incident to the strong elevation of the land contiguous to the graben, the normal gray detritus formed at a time when the denuda- tion of the land went on too rapidly to permit the formation of the red residuary earths. We are thus able to find in this progressively deepening trough, which was filled as fast as it deepened, a natural and satisfactory explanation of the great thickness, the tripartite structure, and all the other original characters of the Norfolk Basin sediments. During the development and filling of the Norfolk trough, the northern edge of the Blue Hills Complex was rising cliff-like above the broader and, apparently, shallower Boston Basin, and the erosion and especially the granular disintegration of this granite escarpment afforded in large part the material of the bordering Carboniferous strata, including the heavy bed of arkose sandstone. Being on the down-throw side of the great displacement, the base of the sedimentary series, as previously noted, is not exposed ; but the available facts suggest a thick- ness materially less than that of the Norfolk sediments. East- 502 ward, the great longitudinal faults appear to terminate in trans- verse displacements, the southern boundary fault being traceable as far as Hingham Harbor, while the northern boundary fault quite certainly does not cross Hingham and may be supposed to terminate with the Blue Hills Complex in the valley of Wey- mouth Back River. Beyond Hingham, in' the Nantasket area, the exposures do not afford a complete section of the conglom- erate series from the granite up to the overlying slates ; but the evidence favors a moderate thickness, as in Hingham, and a corresponding subsidence for both areas, with no indications of a deep local trough or graben. But we do find in these eastern fields the locus of volcanic activity which was clearly contem- poraneous with, and in some way, no doubt, causally connected with, the profound displacements to the westward. To find a parallel to the conditions of deposition proposed here we need go no further than the Connecticut Valley, where, according to Davis, ^ the Triassic monocline is cut off on the east by a fault with a down-throw to the west of, probably, many thousands of feet : and the great thickness of the Trias- sic strata (roughly estimated at two miles), all bearing evi- dence of shallow water deposition, may be accounted for by a progressive development of this displacement during the ac- cumulation of the sediments, the wash of red residuary earths from adjoining land areas, into this trough keeping pace, approximately, with the localized subsidence. According to the as yet unpublished observations of the writer, similar conditions obtained in the Triassic area of the Minas Basin in Nova Scotia. Along the northern or PaiTsboro shore of this basin the gently dipping Triassic beds meet the strongly folded and usually vertical Carboniferous and Devonian strata in an extremely obvious fault having a throw approximating the entire thickness of the Triassic series. That this fault was in process of development during the deposition of the Triassic sediments is practically proved by the fact that it is now ' W. M. Davis, Eighteenth aim. rept. U. S. geol. surv. 503 marked by a chain of elongated or lenticular volcanic necks which supplded the thick beds of contemporaneous lava or trap constituting such an important and conspicuous feature of the Acadian Triassic. Two of these necks, clearly lying in linear fashion in the main fault fissure, are superbly exposed in Was- son's Bluff, five miles east of Parrsboro ; and the volcanic aofo-lomerate fillino- these ancient vents is the great storehouse of the zeolites and other minerals for which this shore is so famous. If these necks and the connecting interbedded sheets of columnar, amygdaloidal, or agglomeratic lava are contem- poraneous, as probably few geologists familiar with the facts would deny, then the fault must also be, in part, contempora- neous ; and we need not suppose that the Triassic series ever extended with its full thickness, if at all, over the older formations on the up-throw side of the fault. This view, it will be observed, minimizes post-Triassic erosion, as well as greatly diminishing the difficulty of finding an adequate supply of red, residuary detritus for the building of the Triassic strata. For still clearer examples, not belonging wholly to a remote past, but in part still in process of development, of deposition strongly localized and accelerated by faulting, we may refer to the Great Basin, between the Wasatch Mountains on the east and the Sierras on the west. The numerous north-south ranges of mountains by which the Great Basin is divided are, almost without exception, monoclinal in structure. In other words, each range is a tilted orographic block, presenting a long gentle bedding slope on one side and a short, abrupt fault scarp on the other ; and the intervening valleys necessarily share the struc- ture of the mountains. The level desert and play a floors of the valleys consist of silt and coarse forms of detritus washed from the bordering mountains, and indicate, in connection with the inclinations of the bordering slopes, that the valleys have been filled to depths, in some cases, of several thousand feet. Ac- cording to Russell ^ and other competent authorities, the throw 1 1. C. Russell, Monogr. 11, U. S. geol. surv. 504 of the more important faults of the Great Basin ranges from 5,000 to 10,000 feet or more, and nearly every fault scarp af- fords evidence of recent slipping : showing that the faults are still growing and probably have been growing during the entire period represented by the silting up of the valleys, which have never had the topographic depth indicated by the converging mountain slopes. If, instead of being an excessively arid re- gion, where erosion is almost at a standstill, the Great Basin enjoyed the hot and humid climate of the southern Atlantic States and the tropics, with a comparatively rapid development of red residuary soils, deposition would have kept pace with the growth of these contemporaneous fault valleys, and they would be now more completely filled Avith a complex of red sediments analosrous to those of the Triassic trouo-hs of the Atlantic sea- board and of the Norfolk Basin. Accepting then, as a working hypothesis, the view that dur- ing the formation of the lower part of the conglomerate series of the Boston and Norfolk Basins, deposition was more or less sharply localized in strongly contrasting areas ; it is, neverthe- less, probable that the deposition of the upper part of the con- glomerate series was continuous, under somewhat uniform con- ditions, across the Complex. But, however this may be, it can hardly be doubted that the overlying slates Avere spread over the entire area north if not south of the southern boundary fault. It will be noticed that both of the boundary faults down-throw to the north, giving a cumulative displacement and at the last differentiating the Boston Basin, by its exceptional depth, from the area immediately to the southward. One important contrast of tlie Boston and Norfolk Basins remains to be stated : The post-Cambrian or Carboniferous sediments of the Boston Basin, notwithstanding the great ex- tent to which they have been folded, faulted, sheared, cleaved and invaded by igneous masses, are rather seldom, in the full and normal sense of the term, appreciably metamorphic. Along the borders of the igneous masses, more or less pronounced local 505 or contact luetamorpliism may often be observed, taking the form especially of baking and induration ; but true regional or normal metamorphism with the development of an indigenous micaceous element, as in the Cambrian slates near the granite, is practically wanting. In the Norfolk Basin, on the contrary, lustrous surfaces and a semi- or incipiently micaceous facies, suggestive of a transition to inica schist, are prevalent charac- ters of the sandstones and to some extent of the shales and the finer part or paste of the conglomerate. This contrast may, perhaps, be attributed in part to the greater thickness and con- sequent more severe compression .and shearing of the sediments in the Norfolk Basin. But the writer is disposed to find the explanation chiefly in the more generally arkose character of the Norfolk Basin sediments, which, in turn, is a necessary correlative, of their more rapid deposition. The essence of the explanation is in the relatively alkaline composition of arkose, the feldspathic detritus insuring the presence in the sediments of the alkalies — potash and soda — so essential to the develop- rnent of mica and normal metamorphism : and we are, per- haps, justified m correlating metamorphism somewhat generally with arkose composition and consequently, as a rule, with rapid and thick deposition. This would explain the absence of met- amorphism (or indigenous mica) in non-arkose sediments, like the slates and sandstones of the Boston Basin, even when they have been buried to a great depth, as in the case of the Pots- dam sandstone of the Appalachian region. In confirmation of, and as a necessary corollary of, this view we have the well- established fact that volcanic tufi^s, which are pre-eminently of arkose character, are so peculiarly susceptible of metamorphism that in the older formations, as a rule, their original structural features have been completely effaced and their histories are difficult to trace. 506 THE BLUE HILLS COMPLEX DURING THE APPALACHIAN EEVOLUTION. Although the Subcarboniferoiis period left in this region no identifiable records either in erosion or deposition, the closing episode of the Carbonifei'ous age and of the Paleozoic era — the Appalachian revolution — is indelibly recorded in strong and salient structure lines : and it must ever stand as the chief event in the geological history of this part of the continent. As else- where, the strata yielded chiefly by plication ; but the displace- ments are also numerous and extensive. The faulting is prob- ably due to the fact that the thickness of imperfectly lithified sediments (2000 to 3000 feet of Carboniferous beds only in the Boston Basin) was insufficient to induce softening of the crys- talline floor by rise of the isogeotherms. The fissures were sufficiently profound to tap reservoirs of basic magma, and the numerous dikes and intrusive sheets of diabase which intersect the rocks of the Boston Basin, and to a very limited extent those of the Norfolk and Narragansett Basins,^ were formed. It is impossible to present here in full even the principal re- sults of this grand crustal adjustment ; but we must confine our attention to the Boston and Norfolk Basins and especially to their dividing wall, the Blue Hills Complex. The Boston Basin has been described as a composite trough. A north-south sec- tion shows belts of conglomerate and melaphyre alternating with belts of the newer slate. The conglomerate marks anti- clinal and the slate synclinal axes ; and in the most general view we recognize one dominant, central, anticlinal arch three miles broad with several subordinate folds on either side of it before we reach the sharply defined northern and southern boundaries, which are undoubtedly displacements of consider- 1 Amer. journ. sci., vol. 20, 1880, p. 417; and Proc. Boston soc. nat. hist., vol. 23, pp. 32,5-3.30. 507 able magnitude, the bordering crystalline rocks being in each case upon the up-throw side. These northern and southern boundary faults of the Boston Basin must be regarded as of primary importance in the geological structure of the region ; for, virtually, they present on the up-throw side two solid walls of crystalline rock between which the great central area of sedi- mentary and volcanic rocks has first settled down and then suf- fered compression as in a vise, by the approach of those walls, producing the folds already noticed and most of the minor faults of the region. The Norfolk Basin, at least in its northeastern extremity, with which alone we have to do, is a simple trough of monoclinal structure. All along the south side of the Blue Hills the coarse basal conglomerate of the Carboniferous series lies, as already noted, directly upon, and dips steeply away from, the quartz porphyry covering of the Blue Hills Complex, the con- glomerate being chiefly composed of the debris of the quartz porphyry. Going directly across the trough, we find that the prevailing dip is southerly ; and, although the southern border is not clearly exposed, it is probable that we pass abruptly across another great displacement to the normal granite of the region. The up-throw is again on the south and has been sufficient to carry the entire Carboniferous series above the present plane of erosion. In the preceding pages these principal displacements have been referred largely but not necessarily wholly to the time when the conglomerate series was being deposited. The later Carboniferous sediments were probably deposited during a rela- tively quiescent period when this region was experiencing a massive rather than a differential subsidence. This period was closed by the Appalachian revolution, during which the growth of the primary faults was renewed, and, in obedience to a power- ful compressive force acting in a north-south direction, the net- work of minor faults and the complex folds of the strata were developed. With the primary faulting may be associated or correlated the effusive volcanic activity which resulted in the 508 formation of the contemporaneous sheets of basic lava (mela- phyre and porphyrite) in the lower part of the conglomerate series ; and with the subsequent period of faulting and folding known as the Appalachian revolution, may be correlated the formation of the numerous dikes intersecting alike the earlier and the later Carboniferous strata. The effusive portions of the later or dike eruptions, if there were any, were not covered and preserved by sedimentary deposits, and hence were long since removed by erosion. Its northern and southern boundary faults, begun in early Carboniferous time and completed during the Appalachian revo- lution, and both throwing in the same direction (northward), make of the Blue Hills Complex an orographic block about ten miles lono^ and increasing in breadth westward from a little more than one mile in Weymouth to more than four miles in the Blue Hills region, embracing in its western half not only the Blue Hills, but also the northeastern extremity of the Norfolk Basin. This orographic block must be conceived as terminated abruptly east and west by transverse faults in the valleys of Weymouth Back River and the Neponset River ; and it is quite certainly divided in the longitude of the Old Colony Railroad by another transverse fault, with the up-throw to the east (p. 411), into two A^ery unequal blocks which appear to have moved some- what independently. That the smaller eastern block has moved en masse is indicated especially b}^ the fact that the fine granite cover of the Complex is preserved over nearly all parts of this block, alike on the north and the south. The large Avest- ern block, on the other hand, has evidently been tilted, the northern edge elevated and the southern edge depressed, and in this depression lies the narrow eastern extremity of the Norfolk Basin, Avhile the northern scarp overlooks the Boston Basin. Concerning the magnitude of the displacements bounding the principal orographic block of the Blue Hills, outlined above, it is impossible, at present, to make definite statements. Al- though this block is, in itself, monoclinal, it is seen in a broader 509 view to be essentially anticlinal in its relations to the bordering Carboniferous basins ; for the strata dip away from it on both sides at high angles. The gigantic and unsymmetrical arch of the later Carboniferous strata finally broke along its northern crest, when displacement by flexure and stretcliing through the renewed growth of the ancient boundary fault had reached its utmost limit ; and the granitic axis, bearing its cover of quartz porphyry and a considerable thickness of the newer Carbonifer- ous strata, rose high above the crumpled beds forming the floor of the Boston Basin. From this northern fault-scarp the con- tact zone of quartz porphyry and overlying strata probably sloped southward very gently at first, and then more steeply as they bent down to form the trough of the Norfolk Basin, finally abutting against the fault-scarp forming its southern wall. The northern slip, forming the southern wall of the Boston Basin, conceals almost the entire thickness of the conglomerate series ; and the base of the conglomerate, considering the high inclination of the beds, is probably not less than two thousand feet below the present surface. To this must be added the thickness of the quartz porphyry and an equal or greater thick- ness of the underlying granite, giving twenty-five hundred to three thousand feet as a minimum or reasonably safe estimate of the total throw. The throw of the southern fault, forming the southern wall of the Norfolk Basin, must be, in view of the great thickness of the Carboniferous beds in this trough, at least double this amount, or say five thousand feet as a mini- mum. Eastward from the fault on the line of the Old Colony Rail- road in the Quincy Adams valley, terminating the quartz por- phyiy and felsite area, and dividing the Complex, we do not find another transverse slip of similar magnitude until we come to Weymouth Back River. East of this north-south valley, in Hingham, we have, besides the granitic series, only an exceptionally complete development of the conglomerate series and the newer slates, or the strata now believed to be of 510 Carboniferous age. Evidence that either the Cambrian or Carboniferous beds cross this line, south of the northern bound- ary fault of the Complex, is entirely wanting ; and since the Carboniferous series on the Hingham side has a measured thickness of about two thousand feet, a slip of at least that magnitude is a necessity here, the down-throw being, in this case, to the east. The slates outcropping near Great Hill, on the north shore of Weymouth, have been referred, with the slate of Raccoon Island and the conglomerate and mela- phyre of Hough's Neck, to the Carboniferous series ; and somewhere beneath the broad expanse of drift and water be- tween these outcrops and the Lower Cambrian beds of Mill Cove and eastern Quincy must extend a continuation of the o-reat displacement forming the southern wall of the Boston Basin ; but east of the Old Colony Railroad the exact course of the fault must remain a matter of conjecture. The fault mark- ing the southern boundary of the Complex is, as previously described, very clearly and instructively exposed in the railroad cuttings in Weymouth, the fissile and lustrous Cambrian slate resting against a nearly vertical, slickensided wall of granite. Among the minor displacements of the Blue Hills Complex described in the preceding pages and probably dating from the Appalachian revolution may be mentioned : first, the longitudinal faults bounding the Hayward Creek and Ruggles Creek belts of Cambrian slate ; and, second, the transverse faults breaking the southern boundary fault in the valleys of Mill Cove, Wey- mouth Fore River and Smelt Brook, and the more hypothetical fault in the vicinity of Braintree Great Pond and Purgatory Road, which we suppose to determine the eastern end of the Norfolk Basin and the common boundary of the quartz porphyry and effusive aporhyolite in the southeastern part of the Blue Hills. The fact that the east-west boundary faults of the Blue Hills Complex so closely follow belts of Cambrian slate is prob- ably due to the circumstance that these narrow zones of slate, extending to a great depth in the granitic floor over which the 511 Carboniferous strata were spread, must have formed planes of weakness in the Complex, and hence determined the position and direction of great displacements. The northern belt prob- ably extends out under the Carboniferous beds, and is not so narrow as it appears on the map. DIABASE DIKES NEWER THAN THE GRANITIC SERIES. Undoubtedly, the closing event in the development of the lithified formations or hard rocks of the Blue Hills Complex was the making of the extensive series of the post-granitic basic or diabase dikes. These dikes are not only contrasted chrono- logically with the earlier or pre-granitic dikes, which they other- wise closely resemble, by being post-granitic ; but they are also essentially post-Carboniferous, clearly dating, in part at least, from the close of the Appalachian revolution, when the Carbon- iferous strata, having become, through severe plication, close- folded and rigid, suffered farther deformation by faulting. That the distinctly basic or diabase dikes of this area belong, chrono- logically, to two series is indicated by their relations to the granitic rocks. The dikes of the fii^st or older series, which have been cleai'ly recognized only in the valley of Pine Tree Brook, as already described, cut the Cambrian sti-ata and are cut and enclosed by the granitic rocks ; while the dikes of the second or later series cut alike the Cambrian beds, the entire granitic series and the bordering Carboniferous strata. It is only where cutting the Cambrian sediments alone that the dikes ai-e ambiguous; and of such dikes there are, fortunately, but few. One of the most striking features of the pre-granitic dikes, as developed in the Pine Tree Brook area, is their close parallelism with the bedding of the Cambrian slate, a parallelism which is per- sistently maintained through an important change of strike (PI. 26) . Although these slates probably once possessed a good lami- 512 nation cleavage, they have become, through the metamorphic influence of the enclosing granite and porphyry, decidedly mass- ive and non-fissile in structure. This is another clear indica- tion that these dikes were formed before the invasion of the slates by the plutonic rocks, although they are, quite certainly, subsequent to the folding of the slate. The post-granitic dikes, on the other hand, show no special disposition to conform witli the bedding planes of either the Cambrian or Carboniferous strata ; but they tend rather to agree in direction, and to some extent in position or distribution, with the fault lines of the dis- trict, which are also the directions of the principal joint systems. In the further consideration of the distribution of the post- granitic dikes, it is interesting to note that they are chiefly con- fined to the Boston Basin side of the Complex, or, in other words, to the region north and east of the main range of the Blue Hills. Over the main or quartz porphyry range of the Blue Hills and the Norfolk Basin dikes are almost entirely want- ing, and but very few" have been noted in the granitic area south of the Norfolk Basin. It has been suggested that this disparity, which was pointed out by Professor Barton and the writer nearly twenty years ago,^ may be attributed to the infrequency of the outcrops. But this explanation is certainly not applicable to the Blue Hills, where the outcrops are broad and continuous, covering, on some of the hills, a large part of the surface ; nor do the Carboniferous beds outcrop any less freely along the southern than the northern side of the Complex, although the dikes are a conspicuous feature of the latter area and entirely unknown in the former. In fact, the only dike which Professor Barton and I discovered in a rather thorough examination of the outcrops of the Norfolk Basin for its entire length of over thirty miles, is the small example previously cited in the long- cutting ill the red beds on the Boston and Providence Railroad, north of Canton Junction. The generalization appears, there- 1 Amcr. joivni. sci., vol. 120(1880), p. 417; see also I'roc. Boston soc. nat. liist,, vol. 23, p. 330. 513 fore, to be well sustained, that in later as in earlier Carbonifer- ous times the Boston and Norfolk Basins were contrasted by the fact that igneous activity was practically confined to the former. The diabase dikes of the Nantasket and Cohasset area are readily referred by their directions to three distinct systems — two approximately east-west (N. 75° E. and S. 75° E.) and the third north-south. These systems are proved by intersec- tions to be also chronologically distinct, the north-of-east system being the oldest and the north-south system the newest. The dikes of the east-west systems have suffered the deep-seated alteration known as chloritization and are composed of typical greenstone ; columnar jointing is seldom well developed in them ; and they are frequently porphyritic. The north-south dikes, on the other hand, have not as a rule experienced the green- stone alteration or chloritization, are usually distinctly cross- jointed, readily decompose to a brown earth, and are never distinctly or coarsely porphyritic. The east-west systems are recognized in Hingham, although exposed intersections are want- ing to establish their relative ages ; and dikes clearly referable to the north-south system are wanting. West of Weymouth Back Kiver, in the area of the Blue Hills Complex, both east- west and north-south dikes are well represented : but, as in Hingham, intersections are nowhere clearly exposed. Both series of east-west dikes are represented, although the younger or south -of-east system appears to prevail ; all of these and some of the north-south dikes are greenstone ; and there is a consid- erable proportion of abnormal trends, especially in the northwest- southeast quarters. The Nantasket classification is clearly applicable to the majority if not to nearly all of the dikes of this more western area ; and it is especially interesting to note that, with practically no exceptions, the cross-jointed, black, rusty-weathering dikes, decomposing readily to a brown earth, have north-south trends. In his microscopic examination of the dikes of the Blue Hills Complex Mr. White has failed, as did Professor Merrill for the OCCAS. PAPERS B. S. N. H. IV. 33. 5M Nantasket area, to discover any distinctive characters for the different systems. This is the more surprising since, macro- scopically, the east-west systems, which are clearly seen at Nan- tasket to be of approximately the same age and possibly chrono- logically distinct only in individual instances, are certainly strongly contrasted with the decidedly younger north-south dikes. In view of all the facts, it appears to be a reasonably safe generalization that the east-west dikes, which often show, besides the general chloritization, unmistakable shearing and crushing to which the absence of the columnar structure or cross-jointing may in part be due, were formed in the main during the general folding or north-south compression of the Carboniferous strata, in the temporary east- west or longitudinal fissures naturally incident to the. plication or bending of the for- mations, with or without faulting ; while the north-south dikes, which are practically free from dynamic metamorphism, may be referred to a somewhat later stage of the Appalachian revolu- tion, when, the formations having become well stiffened, north- south or transverse fissures naturally resulted from the contin- ued north-south compression and consequent east-west tension or stretching. The north-south fissuring, in harmony with this explanation of its origin, was in general unaccompanied by either contemporaneous or subsequent thrust faulting or shear- ing. A few of the east-west dikes hade to the south, and others are vertical, but the great majority have a decided hade to the north. This generalization holds not only for the Blue Hills Complex, but for the entire southern margin of the Boston Basin, as is so clearly shown by the tabulated list of dikes in Pai-t 1. Master joints, shear planes and fault planes of south- erly hade are not at all uncommon ; but under the stresses of the dike-forming period they have seldom opened to afford a pathway for the basic magma. The dominant northerly hade of the dikes appears the more surprising since that is also the attitude of the slatv cleava2:e throus-hout both the Boston and 515 Norfolk Basins, indicating compression normal to the dikes. We must conclude, thei-efore, that the formation of the east- west dikes, as well as of the north-south dikes, post-dates the main period of plication and cleavage development in this region, accompanying the subsequent more massive deformation or period of faulting, when, apparently, the fissures of northerly hade widened downward, and the fissures of southerly hade, as a rule, remained closed or widened upward. It may also be noted in this connection that the north-south dikes throughout the entire region, with few exceptions, either are vertical or have a slight hade to the west. That the hade of either system of dikes has been sensibly influenced or determined by massive movements or tilting of the crust subsequent to their formation is perhaps improbable ; although it is certain that the east-west dikes at least, as previously noted, and to a greater degree than the north-south dikes, suffered dislocation and shearing before the final close of the Appalachian revolution. Among the post-granitic dikes of the Blue Hills Complex the two which accompany, the one the northern and the other the southern, boundary faults, stand out with special prominence by virtue of both size and structural importance. During the prosecution of this work no systematic or thorough search for dikes has been made ; and it is certain that the following de- scriptions, copied from ray note-books, are not exhaustive, although it is probable that every important or distinct type of dike occurring in this area, is represented. On account of their readier decomposition, the nortli-south dikes usually outcrop less prominently than the east-west dikes, and are thus more likely to escape observation. The dikes may be most conven- iently described in topographic order. Dikes in the Region South of the Blue Hills Complex. — In the railroad cutting northeast of King Oak Hill we have ex- posed, south of the lower Cambrian slate and the boundary fault, several hundred feet of coarse reddish granite. It is much crushed and slickensided and profoundly altered ; and it is trav- 516 ersed by several east-v/est dikes of greenstone diabase one to several feet in width, all as badly smashed up and altered as the granite. In the cutting at East Weymouth Station six greenstone dikes are clearly exposed in the granite. They are approxi- mately east-west in trend, vertical, or with a slight hade to the south, and vary in width from eighteen inches to five feet, ex- cept the most northerly dike, which is only four inches. They seem to agree closely in trend ; and parallel with them are two strongly marked shear-planes or faults. On the southwest corner of Middle and Broad Streets, in Weymouth, an ideally regular dike is very handsomely exposed. It is greenish black, cross-jointed, three feet wide, trends N. 75° E. and hades S. 10° approximately. Southeast of this locality, on the east side of Whitman Pond, north of Washing- ton Street, is a fine series of dikes. Only one has been observed in the granite south of the triangular ai^ea of diorite ; one in the diorite ; and half a dozen or more in the granite north of the diorite ; occurring, evidently, at more frequent intervals north- ward. They are composed of fine-grained greenstone diabase, vary in width from six inches to fifteen feet, and trend approxi- mately east-west, belonging mainly at least to the north-of-east system. Mr. White describes a I'epresentative dike of this series as rich in pyrite and showing considerable effects of ci'ushing. East of this group of dikes, on the west side of Pleasant Street, between East Weymouth and Lovell Corners, the normal bio- tite granite has been blasted, and several dikes, two inches to two feet wide, are exposed, trending N. 80° W. Farther south there are, evidently, but few dikes ; and in the seam-fiice o-ranite district of Weymouth and Hingham I have observed but one. This is in Gilbreth's quarry, six hundred feet east of Pleasant Street, in Weymouth. It is only from six to twelve inches wide and trends N. 70° W. West of Whitman Pond dikes in the granitic rocks are cer- tainly relatively fewer. In fact, east of the Old Colony Rail- 517 road only two have been observed, besides those accompanying the boundary fault. One of these is northwest of the pond, midway between Washington and Essex Streets. It is four feet wide and trends N. 70° E. The other outcrops on the summit of Liberty Hill in Braintree, with a breadth of nine feet, and also trending N. 70° E. Both of these dikes are ver- tical, and consist of the normal greenstone diabase ; and the latter is profusely porphyritic. Between the Old Colony Railroad and the Boston and Provi- dence Railroad, a distance of eight miles, the only dikes ob- served are on the highway between South Braintree and Ran- dolph. On the direct road to Randolph, about half a mile south of the road leading west to Great Pond, is an east-west greenstone dike three feet wide ; and nearly half a mile south of this the granite is cut by three vertical north-south dikes from two to six inches wide. In the granite cutting on the Boston and Providence Railroad, immediately north of Canton Junction, several rather irregular dikes of moderate width are exposed. To this area belongs also the great southern boundary dike, which has been described with sufficient fullness on page 405. This is probably the largest dike in the entire area of the Blue Hills Complex. It has been traced, with a maximum breadth of three hundred feet, for about one and a half miles in a gen- eral east-northeast direction from East Braintree nearly to King Oak Hill in Weymouth. It is a true boundary dike for not more than half this distance. Where the boundary fault is thrown to the northward by the Fore River fault, the dike enters the granite. It is clearly later than the main part at least of the displacements, and yet evidently genetically related to them. Mr. White describes the groundmass of this dike as composed of dirty brown alteration products in which are imbedded lath- shaped feldspars. Bending of the crystals and flow-structure are observable. Triangular magnetite crystals are present, and 518 a little ilmenite somewhat altered to leucoxene. In what is virtually a part of this dike, on Mt. Pleasant, Mr, White found a handsome rose-colored augite that appears like garnet except for its low refraction. He also says the rock is much chloritized, the chlorite extending all through the feldspar, while many minute veins of quartz ramify through the rock. Dikes in Worth Weymoiith. — On account of the nearly continuous mantle of modified drift, the dike phenomena of this area are exceedingly scanty. Near the head of Mill Cove, and four or five hundred feet north of the railroad, is a prominent outcrop of trap bounded on all sides by salt water and marsh, and connected with White's Neck by a tide-w^ater dam. The rock is a finely and somewhat obscurely crystalline greenstone in which, according to Mr. White, the original microscopic characters have been almost entirely eifaced. The north-south breadth of the outcrop is nearly two hundred feet. A little more than half a mile to the eastward, in the laro-e s^ravel exca- vation between East and Gi'een Streets, is another large and ill-defined exposure of trap. To refer these two outcrops to one great east- west dike parallel with, or possibly a dislocated portion of, the boundary dike, is a tempting generalization. The only difficulty is that they are not lithologically very simi- lar, the more eastern outcrop being an entirely normal, dis- tinctly crystalline greenstone diabase. The only other dike exposed in this area is that cutting the red Cambrian slate and limestone on the north side of Mill Cove (see PI. 20). It is a well-defined, regular dike of greenstone thirty inches wide, of rather abnormal attitude, trending N. 40° E. and hading N. W. 40°. Mr. White describes it as highly altered. Dikes in the Area between JVeymoiUh Fore River and the Old Colony Railroad. — In the Monatiquot Valley belt of Cambrian strata a few small longitudinal or east-west dikes have been observed, some of which have been noted incidentally in the descriptions of the Cambrian outcrops. In the railroad cut east of Weymouth Station, about six feet from the granite, 519 is a gray, compact dike twelve to eighteen inches wide and parallel with the bedding of the slate. In the cutting on Quincy Avenue south of the railroad, two vertical east-west dikes of similar gray comj^act trap traverse the slate about fifty feet north of the great boundary fault. They are only about one foot apart, and the north dike averages one foot wide and the south dike half a foot. At the Jenkins Rubber Mills, in the green slates along the middle of the valley, is a nearly ver- tical but conformable dike about one foot wide trending N. 82° E. And in the same slates, on the north side of the river, east of Commercial Street, we have exposed at intervals for about six hundred feet what appears to be a continuation of this dike with a width of one foot or less and perfectly conform- able with the slates : but it is here highly chloritized and largely replaced by vein quartz. North of this, in the red and green slate about the East Braintree Station, two east-west dikes outcrop obscurely, trending with the slate N. 77° E. The southern dike is of normal character and three feet wide, but the northern dike, under the railroad, shows much alteration and is distinctly amygdaloidal. The possibility must be recog- nized that alfof these narrow, highly altered, conformable dikes in the slates may be pre-granitic. In the Wyman Hill area of fine granite separating the Cambrian strata of the Monatiquot Valley and Hayward Creek, no dikes have been observed. On the north side of the Para- doxides Quarry, at the mouth of Hayward Creek, a vertical north-south black dike about six feet wide cuts the massive flinty slate ; and in the next quarry immediately to the west of this is an irregular dike ten to twenty feet wide trending about southeast. Mr. White calls this a decomposed diabase in which the augite is mostly changed to chlorite, giving the rock a green color. Pyrite is present in microscopical crystals, with much ilmenite slightly altered to leucoxene. Following up the valley of Hayward Creek to Echo Quarry at the base of Payne's Hill, we find at the east end of the quarry a small N. Fig. 46. — A north-south dike of diabase in the nor- mal granite northeast of Payne's Hill, Quincy. Scale ; 1 inch = 16 feet. 520 but veiy typical example of the north-south dikes. It is vex'tical, cross-jointed, brown-weathering, from three to twelve inches wide and rather crooked. It is faulted about ten feet by the very perfect slickensided joint plane described on page 340. In the principal quarry next northeast of Echo Quarry, in the direction of Quin- cy Avenue, are two larger but equally typical dikes of the black, brown-weathering dia- base. They have the normal north-south trend, are only twenty to thirt}^ feet apart, and are beautifully cross-jointed. The east dike is best exposed, is two feet in normal width, vertical and curiously irregular as shown in the accompanying sketch (Fig. 46), which shows it to be virtually two dikes. The second or west dike hades east some thirty degrees and seems to be narrower than the other. They are probably united below ; and the west dike may be simply the oiFset southern extension of the western mem- ber of the east dike. The large dike of finely crys- talline greenstone diabase sepa- rating; the slate and fine gran- ite alona' the straio-ht northern 521 fault margin of the Eldridge Hill Cambrian area has been described on page 443. Mr. White describes it as very highly decomposed and chloritized. Mr. Watson has traced this dike westward through the normal gi-anite nearly to Quincy Avenue. It appears to be shifted to the northward somewhat, as if by an oblique foult, on the line of Quincy Avenue. Beyond the avenue, Mr. Watson has followed it in frequent outcrops, showing a breadth of thirty feet, to Phipps Street, where it seems to end at a point just south of the double angle in the street. But about six hundred feet farther up the hill, on Phipps Street, Mr. Watson has found what appears to be a displaced continuation of the same dike, agreeing closely in width, trend (E.-W.) and lithological character. This dis- placed member of the dike, if such it be, is exposed on the west side of the street for about a hundred feet, and for twenty or thirty feet on the east side; but beyond that, eastward, on the same direct line are granite ledges, with no trace of the dike in either outcrops or bowlders. It is probable that we have here an original jog or offset in the dike fissure, rather than a true displacement or fault subsequent to the formation of the dike. Where the fine granite on the south side of the Puo-o-les Creek valley has been quarried on the line of South Street, it is traversed by another large east- west dike, very similar in breadth and character to the Eldridge Hill dike. It is exposed for several hundred feet ; but Mr. Watson and I have been unable to find another indubitable outcrop on this line, although three or four hundred feet farther west there is a small expo- sure of trap which is probably a bowlder but possibly a ledge. The only other dike observed in the Ruggles Creek Valley is at the extreme west end of the slate belt. It belongs to the north-south series, is about eight feet wide, and cuts directly across the slate a hundred to a hundred and fifty feet east of New Road ; but it cannot be traced either way into the granite, seeming^ to end as^ainst the faults boundinsf the slate. 522 No dikes have been noted in the granite north of Ruggles Creek : but north of the granite, in the Union Street Cambrian outcrops is another important east-west dike. It is about ten feet wide, and, crossing Union Street at the corner of Edwards Street, can be traced eastward, parallel with the strike of the slate, for nearly four hundred feet. This dike and the greenstone dike near the Paradoxides Quarry might be regarded as pre-granitic, but it appears best to class them with the very similar Eldridge Hill and Ruggles Creek dikes, which are very clearly post-granitic. Dikes in the Pine Hill and JSTorth Common Hill A.rea. — Three dikes only have been observed in the Pine Hill area, or in the angle between the railroad on the northeast and Willard and West Streets on the northwest. Advancing northward, the first dike is a very important member of the east- west series, cutting the normal gray granite along the line midway between the quarries of I'ed granite on the southern margin of this tract and the quarry of gray granite on the northeastern margin, opposite the end of Liberty Street. Where most clearly exposed, the dike, which is about thirty-five feet wide and hades N. 30° to 40°, forms the bottom of a well-defined chasm, fifteen to twenty feet deep and trending N. 75°- 80° W. It is a dark green, crystalline diabase, showing a very normal grada- tion in texture from the middle toward the walls, and quite distinctly cross-jointed. According to Mr. White the ophitic texture is strikingly developed ; the plagioclases are in lath- shaped forms, sometimes fibrous, and of a greenish color; the interspaces are filled by masses of a handsome pink augite : and iron ore is abundant, chiefly in grains of irregular outline, hence presumably ilmenite. The chasm is a distinct topographic feature for some three hundred feet ; and beyond this, east- ward, the dike can be traced by the drift for about a quarter of a mile, or nearly to the margin of the swamp which here bor- ders the Pine Hill tract on the east. At the west end of the chasm the trend of the dike changes abruptly to N. 25° W. 523 and continues with the same breadth and hade as before for fully three hundred feet farther, where it is lost beneath a well- defined and swampy transverse depression, which probably marks a fault, for no trace of the dike could be found in the rocky northwestern wall of the depression. Going northerly from this point we come, just before reaching the quarry oppo- site the end of Liberty Street, to the second dike, which, having about the same trend (N. 25° W.) as the first dike where it disappears beneath the swamp, may be regarded as possibly the displaced continuation of it. It is a true greenstone diabase, very compact on the borders, and sends branches into the bordering granite and porphyry. It is forty to fifty feet wide, and is exposed along the northeastern margin of the Pine Hill tract, parallel with the railroad, for nearly two thousand feet. Farther to the northwest it is clearly cut off by the normal granite, being first shifted to the west several hundred feet. The third dike, having approximately the same aberrant trend, cuts the basic porphyry near the eastern end of the street running into the Pine Hill tract from the junction of Willard and West Streets. It is of the same character as the preceding dikes, but not as wide. Obviously some cause has operated to give the east-west dikes distinctly abnormal trends in the main part of the Pine Hill tract ; and this cause is, no doubt, closely related to or dependent upon the northwesterly trend of the bands of basic porphyry and fine granite which we suppose to indicate the former extension of the Pine Hill slate belt. Northeast of the Pine Hill tract, between the railroad and Center Street, is the hill on which Wilson's Quarry is situated. A north-south dike one foot wide crosses the east wall of the quarry, hading W. 5°. It is black, weathering brown, com- pact in texture and cross-jointed. About ten feet east of this dike is its exact duplicate in every respect, but not so well ex- posed. Crossing the valley northward, we find in the numerous quar- 524 I'ies of North Common Hill several interesting dikes. Com- mencing with the Hardwick Quarry, on the north side of Quarr}' Street, we find that a north-south dike forms its eastern wall and, crossing the ridge, forms also the eastern Avail of the next quarry to the north. It is normally nearly three feet wide, trends almost due north-south, and hades W. about 5°. After the manner of its class, it is distinctly cross-jointed ; but it belongs lithologically with the east- west dikes, its dark green color indicating a good degree of chloritization. The trap is very compact and slaty, and minutely pseudo-amygdaloidal with calcite, which, weathering out, leaves the rock cellular, the cavities being distinctly angular. The calcite is evidently sec- ondary and, probably, a necessary concomitant of the chloriti- zation of the original augite. The dike, although very regular in form, is really a series of beautiful lenses, tapering to a thin edge in both quarries. It is exactly parallel to a system of north-south joints in the granite, and incloses long, thin sheets of granite one fourth of an inch to three inches thick and parallel to the walls of the dike. These inclusions are ex- tremely dike-like, suggesting at first dikelets of granite cutting Cambrian slate ; but a close examination shows that this cannot be the case. Immediately beyond the Hardwick Quarry, on the south side of Quarry Street, is a large quarry at right angles to the street. Across the street end of this quarry is another north-south dike. It is really a double or composite dike, consisting of two dikes, each about a foot wide and separated by about a foot of granite ; and, as before, it hades west about 5°. It is the same compact, slaty, greenish chloritic trap ; and yet both members of the dike are beautifully cross- jointed. On the north wall of this quarry is a dike of slaty, greenish trap, four feet wide, which trends N. 80° W. and hades S. about 10°. It is much broken, but sliows some columnar jointing. In the Lombard Quarry immediately northwest of this, the same dike reappears, forming the south- 525 ern wall. It is admirably exposed, is four feet wide, cross- jointed, trends exactly E.-W. and hades S. 10°. The north wall of this quarry is formed of a dike of the same chai'acter and trend, but hading N. 5°. Passing several quarries which afford superb examples of parallel and master joints, we come to the only other dike observed in the quarries on this hill. It is the small north-south dike crossing the middle of the Craig and Richards Quarry, on the northeast side of Quarry Street. It is six to possibly eighteen inches wide, irregular, and divided on the north wall of the quarry into two overlapping parts. It trends N. 10° W. and hades W. 5°. The east and west ends of the quarry are broad, flat and very rusty joint faces exactly parallel with this dike. This dike is especially remarkable for its deep and complete kaolinization, to which my attention was first called by Professor George H. Barton. In fact, it is, in this respect, rather unique among the dikes of the Boston Basin. On account of its small size it is naturally very fine-grained and compact; and to the greatest depth of the quarry, nearly fifty feet, it is more or less completely decomposed to a greenish and brownish earth. The chemical decay is, how- ever, somewhat localized, being most marked in the central and southern part of the quarry, where a coarsely crystalline vein of quartz occupies the same fissure. The vein, which is not clearly exposed, appears to present a series of pockets, yielding well-formed crystals of quartz from one to several inches in diameter, and affording meteoric waters easy access to relatively deep portions of the dike. Among the residuary products of the decay of the dike, decidedly the most striking and interest- ing is the snow-white and plastic kaolin which appears to belong normally between the quartz vein and the green residuaiy earth or clay representing the main part of the dike. Where most perfectly developed and when freshly dug, the kaolin is not only perfectly white but as soft and plastic as butter; with, appar- ently, a rather sharp line of demai'cation between it and the green clay or earth. Its distribution seems to be very localized 526 or bunchy : and it has a thickness in some cases of several inches. That the kaolin really is a product of the greenish black trap appears certain. The quartz vein shows no sign of a peg- matitic character, that is, of carrying feldspar which might have yielded the kaolin ; and although the enclosing granite is some- what decomposed in places, as would be expected, there is no indication of a passage into the kaolin ; and, furthermore, the kaolin is absolutely free from grains of quartz or any gritty material. The following p.nalyses were made in 1894 by Miss H. L. Gates, then a student in the Massachusetts Institute of Technology. A soft, plastic, freshly dug sample of the white kaolin lost 26.38 per cent, of water by drying in the hot closet for one week, and then yielded : — Silica (SiO.) 45.44 per cent. Aluminum sesquioxide (Al^Og) 37.85 " Water (H,0) ^ 13.93 97.22 A proper distribution of the loss error would make this a very close approximation to the normal composition of kaolin : and it is evident that the process of kaolinization has here accom- plished its perfect work. The kaolin becomes quite hard and firm by ordinary drying ; and such an aii'-dried sample was also analyzed, yielding essentially the same results as the plastic sample. The associated greenish clay, free from grit and cut- ting with a knife like cheese, gave on analysis : — Silica (SiO.2) 51.44 per cent. Aluminum sesquioxide (xVl.Os) 35.21 " Iron oxide (estimated as F.^Og) 1.14 " AVater (HoO) - 6.62 94.41 The silica and water both suggest incomplete kaolinization 527 and the probable presence of residual portions of protoxide bases (alkalies or alkaline earths) which were not determined. The coloring agent is undoubtedly iron in ferrous form. The unusual features of this occurrence of residuary earth are: (1) the extent and completeness of the decomposition; (2) the prevailing green or ferrous oxide color of the residue, instead of the brown tints commonly observed in the decomposed rocks and residuary earths of this region and due to the peroxi- dation of the iron : and (3) the complete bleaching of a part of the residuary kaolin contiguous to the quartz vein. The condi- tion of the outcrop of this dike before the opening of the quarry is not apparent ; but the rather unique features of the decom- posed part of the dike seem to find a ready explanation in the not improbable assumption that the surface was locally somewhat of a swampy character. This condition, in connection with the open vein of quartz, insures (1) an abundance of carbonic acid and the various humus acids for the kaolinization of the trap, (2) the organic matter requisite to prevent the peroxidation of the iron, and thus to permit its solution and removal as a car- bonate or organic salt. Dilces in the West Qaincy Quarry District. — In the numerous quarries and ledges of this area four dikes of the east-west series have been observed, but none that can be referred to the north-south system. The most southerly dike is also the most important in length of outcrop, being traceable for nearly a mile. It first appears along the north side of the branch of the Quarry Railroad which runs oflf to the eastward toward the Bunker Hill Quarry. The outcrop continues for several hundred feet, trending N. 75°-80° W., but does not show the hade or full width of the dike. Proceedino; westward on the line of the dike, we find, north of the. main line of the Quarry Railroad, and some five or six hundred feet beyond the junction, considerable diabase in the drift, of similar char- acter to the dike, and again nearly a quarter of a mile farther west and a few rods northwest of the Jones and 528 Desmond Quarry. The dike next appears beyond Blueberry Swamp, in the northern pai't of the Fitzgerald and Lyons Quarry, with a thickness of eight to ten feet and hading N. 20°. Its next and last appearance is on the south side of the Fuller and Clarke Quarries, with the same hade ; but the full width is not exposed. That all of these exposures are referable to one dike is further indicated by their lithologic uniformity. The rock is, throughout, the normal dark green diabase with very compact or slaty margins. The second dike of this series is the small example forming the southern walls of the Berry Quarries, some eight hundred feet southeast of the Granite Railway Quarry. It is only from six to twelve inches thick, trends N. 85° W., and hades X. 15°. The third dike, which is also the largest and most con- spicuous, is that forming the southern wall of the Granite Railway Quarry. It trends X. 70° W., hades JST. 30° to 35°, and is about fifteen feet thick. Lithologically it is a very normal east-west dike. Converging downwards toward this dike, with a hade S. 25°, is the strongly marked shear and probable fault plane described on page 339. This is simply one striking example among many master joints of southerly hade which serve by contrast to emphasize the prevailing northerly hade of the dikes for the entire southern mai'gin of the Boston^ Basin. The fourth and last dike of this series crosses the extreme north end of the quarry on the west side of Willard Street, north of West Quincy Station. It is about four feet wide, is very regular, trends almost due east-west, and hades N. about 10°. It is nearly black in color, but not deeply weathered. Dikes in the Ma{7i Range of the Slue Hills. — This sec- tion may be very brief; for in the Blue Hills proper, as pre- viously stated, there are practically no dikes. In fact it is probably the most nearly dikeless area in the Boston Basin ; and in the entire area south of the Pine Tree Brook belt of slate and west of Pine Hill and Willard Street I have noted but one example. This is a large and imperfectly exposed body of 529 diabase in the extreme northeast corner of the area, contiguous to the West Quincy quarry district, to which it might properly be referred. The chief outcrop forms the south side and sum- mit of the small hill known as Babel Rock, or bearing a granite bowlder distinguished by that name, south of Old Furnace Brook and west of the junction of Willard and West Streets. The breadth of the outcrop would indicate an east- west dike some two hundred feet or more wide ; and that it really belongs to the east-west series is clear from its lithologi- cal character — fine-grained greenstone — and from other out- crops to the westward. The first one of these is about five hundred feet distant, on the north side of the road leading to the Rattlesnake Hill quarries. A little more than a quarter of a mile beyond this, in a west-northwest direction, on the south border of Furnace Brook swamp and on an island in the swamp crossed by the path to Sawcut Notch, is more trap, so distributed as to bear out the theory of a large east- west dike. Farther west and east the outcrops are all granite, and the dike cannot be traced. Dikes along the Northern Border of the Blue Hills Com- jplex. — The most westerly dike of which we have any knowledge along this line is the great boundary dike of the Randolph Avenue area described on pages 438 and 478. It outcrops on Randolph Avenue wuth a breadth of a hundred and twenty-five feet ; and it can be traced westward a thousand feet to Reedsdale Road, and eastward a little more than two thousand feet to Pleasant Street, where its breadth is reduced to about forty feet. It has a strong northerly hade, and its normal trend is north of east ; but east of Gun Hill Road it apparently shifts to a south-of-east trend, the only alternatives being to refer the Pleasant Street outcrop to a branch or to an independent and narrower dike. The view that both of the east-west systems are represented here by independent dikes is, perhaps, the best, everything considered. It derives strong support from the fact that in the granite south of the junction of Pleasant OCCAS. PAPERS B. S. J^. H. IV. 34. 530 Street and Gun Hill Road is another dike with approximately the same south-of-east trend. It is a typical greenstone, forty- five to fifty feet wide, which is being worked at the pres- ent time for road metal. Its trend intersects Pleasant Street at a point six hundred feet west of Gun Hill Road. Farther east these south-of-east dikes, as well as the north-of-east Ran- dolph Avenue dike, supposing that to be independent, are lost beneath the drift ; but that the former are in some way repre- sented in the south-of-east dikes of the West Quincy quarry district seems most probable. Among the ledges of Cambrian slate and granite in the valley of Furnace Brook south of Edge Hill Road there are indications of a large east- west dike. One of the ledges in the complex of granite and Cambrian slate south of Furnace Brook and west of the Granite Branch Railroad, shows an east-west dike eis^hteen inches wide cutting- the slate. This appears to exhaust the exposed dike phenomena along this border. Dikes in the Rock Island Anticline. — This broken fold may be considered to begin on the west in the outcrops of con- glomerate and bordering slate in the valley of Furnace Brook west and east of the Old Colony Railroad ; and in its eastern extension it embraces the ledges of Houo-h's Neck and Raccoon Island. Between the railroad and Hancock Street an east-west dike, widening eastward from two to four feet, runs through the conglomerate and slate, as previously noted. The ledge of slate nearly three hundred feet wide on the road to Hough's Neck is bordered on the north and south by obscurely exposed east-west dikes, as described on page 486. So far as known, no part of either the Blue Hills Complex or the contiguous Carboniferous areas is so much cut up by dikes or presents such an interesting group of fissures in a lim- ited area as the western half of Rock Island (PI. 27) ; and it is a significant fact that, as the map so plainly shows, the dikes, or at least their outcrops, are almost entirely confined to the great bed of melaphyre. Still more curious are the facts that 531 in their relations to the melaphyre and the geological structure generally they are neither distinctly longitudinal nor transverse ; and that, while showing little close agreement in trend, they are all confined to the northwest and southeast quarters of the compass. The hade of the dikes is, with one or two vertical exceptions, distinctly to the north and east. Lithologically they ai-e nearly all normal greenstone ; the one important exception being the great dike (No. 12). This weathers brown and has the general facies of the north-south dikes. It is sixty and possibly seventy feet wide, of uncertain hade, and cuts both the melaphyre and conglomerate, as well as several of the greenstone dikes. So far as known, this is the largest dike of ~the north-south system on the south side of the Boston Basin. The small dike (No. 11) agreeing most nearly in trend with the great dike should probably be referred to the same system. All the remaining dikes should, apparently, be referred to the east-west systems, although several of them are quite aberrant in trend. The east-west dikes are more or less interrupted by transverse faults, as the map shows ; but no clear intersection of dikes of this system has been observed. The apparent intersection of dikes 3 and 4 is quite possibly a branching. This explanation is suggested by the facts that both dikes are porphyritic and that they are the only por- phyritic dikes in the entire area. The Rock Island dikes have been studied microscopically by Dr. J. E. Wolff' in connection with the melaphyre. He de- scribes the great north-south dike (12) as a coarse-grained, dark green rock containing crystals of feldspar, pyrite, magne- tite and hornblende, in a dark green groundmass. The section shows rather large feldspar crystals ; fibi-ous, greenish, dichroic hornblende ; crystals of magnetite and pyrite ; decomposed crystals of olivine ; epidote ; and viridite, quartz, apatite, etc. The feldspars are to a great extent kaolinized. The hornblende occurs in irregular masses and contains a great deal of epidote 1 Bull. Mus. comp. zool., vol. 7, pp. 231-242. 532 in rounded grains. Some of the feldspar crystals lie imbedded in the hornblende, or cross it, just as they do in the case of the augite of the less decomposed diabases, so that this and the whole character of the hornblende indicate that it is (in part at least) a product of the decomposition of the original augite. The olivine occurs generally in shattered crystals, with the usual blackened borders. The magnetite is found in extremely irregular forms, while the pyrite grains often contain magnetite, and therefore arise probably from its decomposition. The large, coarsely porphyritic dike (No. 4) is described as consisting of a grayish green groundmass holding crystals of greenish feldspar and grains of pyrite. The section shows the rock to be much altered. The feldspars retain their outlines, but are filled with chloritic material, kaolin, epidote and calcite. Magnetite and menaccanite are very plentiful in crystalline and irregular forms. Pyrite occurs in occasional grains and square crystals, generally close to or mingled with the iron oxides, and is therefore probably an alteration product. The remaining portion of the rock is a confused mixture of chlorite, epidote, quartz, viridite, hornblende, calcite, and colorless needles, in part probably apatite, — all products of alteration. Dr. Wolff adds, " This rock is the most coarsely crystalline and the most decomposed of any examined." A non-porphyritic east-west dike is described as a greenish gray, felty-looking rock, containing minute grains of pyrite and small feldspar crystals, and traversed by veinlets of epidote. The section shows white, opaque, feldspar crystals and masses of opacite, magnetite, and pyrite, in a green, chloritic groundmass. The feldspars have generally the long ledge form of the basaltic triclinic feldspars, but occasionally the form of Carlsbad twins of sanidin. They are entirely altered to a fibrous and scaly aggregate. Colorless needles of apatite occur occasionally in the feldspars, and also aggregates of quarts. Between the feldspars lies a mass of green fibrous products, — chlorite, viri- dite, etc., considerable epidote, magnetite, quartz, etc., rarely hematite and biotite. The feldspars occasionally have a fluidal arrangement. Dikes of Hock Island^ IIougKs Neck. Width No. Trend. Hade. in feet. Remai'lis. 1 N.39" W. Vertical. 0.75 Wholly in conglomerate. 2 N. 65° W. to K 86° W. N. E. 0°-5°. 2-3.5 Eatilted and cut by No. 12. 3 N. 85'- W. N. 20°. 3 Sparingly porphyritic; cut by No. 4 or possibly a branch. 4 N. 56° W. N. E. 15°-20° 15-20 Coarsely porphyritic on south side. 5 N. 62° W. N. E. 5°. 7 Faulted to left, 60 feet. 6 N. 52° W. to N. 61° W. N. E. 22°. 5 Faulted and cut by No. 12. 7 N. 50° W. N. E. 27°. 1-1.3 Only one outcrop. 8 N. 32° W. N. E. 15°-20°. 16 Probably faulted. 9 N. 32° W. N. E. 10°. 3.5 Probably faulted twice. 10 N. 45° W. ? '? Irregular and not well exposed. 11 N. 23° W. N. E. 10°-15°. 2-3 Exposed for 50 feet ; dies out southward. 12 N. 20° AV. ? 60-70 Brown-weathering ; cuts Nos. 2 and 6. > No dikes have been observed in either the melaphyre or con- glomerate in the eastern half of Rock Island, nor in the Rock Island Head ridge of melaphyre. There are, however, two east- west dikes on Raccoon Island. They are exactly parallel with the bedding of the slate (trend IST. 85° W. and hade N". 0°-5°) and run directly through the middle of the island, under the house, from the extreme easterly point to the west shore. They are composed of normal, fine greenish trap ; and are fif- teen feet apart east of the house and about thirty feet west of the house. The south dike averages about one foot wide, and the north dike varies from two feet to nearly five feet, averaging about three feet. 534 POST-CARBONIFEROUS EROSION OF THE BLUE HILLS COMPLEX. " The hills are shadows, and they flow From form to form, and nothing stands ; They melt like mist, the solid lands, Like clouds they shape themselves and go." Structurally, the Blue Hills were now complete in the two principal fault-blocks, extending from Weymouth Back River to the Neponset River, out of which they were to be carved. Each of these blocks embraced in vertical section from the base upward : (1) the batholite of normal granite with its covering of fine granite and quartz porphyry; (2) an unknown and vari- able thickness of the strongly folded Cambrian slates ; (3) over a part of the area at least a remnant of the effusive felsites ; and (4) several thousand feet of Carboniferous strata. The eastern block, having been lifted bodily, without important tilt- ing, presented a plateau character ; but the strongly tilted west- ern block must have formed from the beginning, a sharp- crested monoclinal mountain range. We thus account for the existing topographic contrast of the two blocks, and learn that, prominent as the Blue Hills are in the modern landscape, they possess scarcely a tithe of their original mass and height. So far as we know, this area has been above sea-level con- tinuously since the Appalachian revolution, with the exception of a brief partial submergence during the close of the glacial epoch. There are no indications that the sea invaded eastern Massachusetts in Triassic times, when the Connecticut Val- ley was a Bay of Fundy ; nor, apparently, did the elevation and faulting of the Triassic beds add to the complexity of our local geology. The Appalachian revolution, as previously noted, gave eastern Massachusetts mountain relief, which ex- 535 plains the non-occurrence of later sediments ; and it also devel- oped a high degree of rigidity in this part of the earth's crust, which has, at least in large measure, precluded subsequent flexures and fractures. The grand result accomplished by post- Carboniferous erosion has been the substantial efFacement of the Appalachian reliefs and the sweeping away of the entire Carboniferous series from' many hundreds of square miles of the granitic and Cambrian floor on which it was deposited, these strata remaining now only in the deepest troughs into which they were down-folded and faulted during the Appalach- ian readjustment. That the subjacent floor has itself been deeply planed down is obvious from the fact that even the axes of the folds which must once have existed outside the present basins can, as a rule, no longer be traced. We have one im- portant exception, however, in the Blue Hills Complex, the denuded but still salient axis of the faulted anticline three thou- sand feet or more in height and four to five miles broad which once divided the Boston and Norfolk Basins. The details of the erosion of this axis can be followed most closely and are most intei-esting for the tilted western half. The crest of the range was here the brink of the great northern fault-scarp ; and it must have so continued until erosion had cut down to and through the contact zone of quartz porphyry. In the existing relations of relief to oreoloajical structure in the western fault block two facts are especially prominent. First, the main range, including all the summits from Great Blue Hill to Rattlesnake Hill, is in the southern half of the block. Second, the main range embraces, superficially, but little granite, consisting almost wholly of quartz porphyry, and follows quite closely the northern edge of the porphyry. Since the time when erosion first uncovered the quartz porphyry, its elevated northern edge has, apparently, formed the crest of the Blue Hill Range. In other words, during the gradual wearing away of the quartz porphyry this crest-line has migrated south- ward, and its migration will probably continue until, through 536 the more rapid descent of the sheet where it bends to pass be- neath the Norfolk Basin, it is finally reduced to the base-level of the region. This must mean that the quartz porphyry is a harder and more resistant rock than the granite ; for, other- wise, as soon as the granite was exposed by the wearing away of the northern edo;e of the felsite the rano-e would either cease to present a single well-defined and dominant crest, or the crest would be transferred to the granite. This differential erosion, so clearly indicated by the fact that at all points along the range the quartz porphyry rises above the granite, the difference of level varying from less than one hundred feet at the eastern end of the range to more than four hundred feet at the western end, is confirmed by observations upon the present rates of de- cay and waste of these rocks. Tlie coarse granular texture of the granite is its weak point. Differential expansion and con- traction of its component minerals gradually loosen the grains and permit the ingress of water and other agents of decay ; and both the chemical decomposition and disintegration by frost action become more and more marked with increased absorp- tivity. Even where the surface of the granite is comparatively hard and fresh, the angles of the ledge are usually well rounded off, and in the soil at the base of the ledge or in depressions on its surface the detached grains of quartz and feldspar have ac- cumulated in the form of a coarse sand. This detritus is really much more abundant than it usually appears, since it is easily overlooked after becoming incorporated with the vegetable mold. The quartz porphyry, on the other hand, does not suffer gran- ular disintegration to an appreciable extent ; and its outlines are much more angular. It is a more brittle rock than the granite and therefore more closely jointed ; and the action of frost in incipient joint-cracks tends to reduce the surface of the ledge to a mass of angular fragments. A quite remarkable talus of this kind can be seen on the precipitous western slope of Hancock Hill. It does not appear, however, that these an- gular blocks of quartz porphyry are readily reduced to finer detritus wliich can be transported by rain wash. It [)rcsents, in consequence, broader continuous exposures, and the traces of grlaciation have been less erenerallv effaced than on the ijran- ite. The relations of these rocks to the agents of erosion are thus seen to be in harmony with and to explain their topographic relations. The more rapid erosion of the Pine Tree Brook band of Cambrian slate than of either the quartz porphyry or the granite graduallj^ developed the broad depression or valley separating the elevated quartz porphyry range on the south and the low granite range on the north ; and we have in the limited length of this belt of slate an explanation of the fact that the duality of the Blue Hill Range cannot be traced either way to the end of the Complex, but only so far as the slate belt extends. In its geological relations Great Blue Hill is comparable with Mt. St. Elias, which Russell has shown to be the high corner of a tilted orographic block. EELATIONS OF THE BLUE HILLS COMPLEX TO THE PENE- PLAINS OF EASTEEN MASSACHUSETTS. If the entire region were base-leveled or worn down to an ideal peneplain, as it may have been more than once since the Appalachian revolution, the Blue Hills Complex would still have, by virtue of its boundary faults, great sti-uctural relief. This we have duly considered ; and we come now to regard the Blue Hills as an erosion monument, as measuring in their actual topographic relief the erosion of the country which they domi- nate. It is not proposed to make this phase of the subject a prominent feature of the present study, and a few summary statements must suffice. The most perfect peneplain must slope seaward. The mass- ive elevation which often follows the development of the pene- plain may be and usually is attended by crustal deformation or 538 tilting, thus accentuating the seaward slope ; and where during subsequent base-leveling the inter-stream surfaces are reduced to mere peaks and crests the unequal lowering of these give& rise to still other differences of level beyond what is implied in the name peneplain (approximately a plain). The greatly dissected, but still readily recognized and approxi- mately continuous. Tertiary or coastal peneplain of eastern Massachusetts has a normal height near the coast of from one hundred to two hundred feet. But as we recede from the shore the elevation of this coastal peneplain gradually rises to three hundred feet and more. Rising from the coastal peneplain are widely isolated remnants (monadnocks), including the Blue Hills "^(635 feet), Moose Hill (536 feet), Nobscot Hill (606 feet). Prospect Hill (460 feet), etc., of an older land surface. Farther inland this older and higher Cretaceous or submoun- tainous peneplain is represented by many summits, which are to a large extent connected in more or less continuous ridges and plateaus, and the coastal peneplain is represented only in the lowlands or valleys. Northwestward from Boston Harbor, the first approximately continuous development of the submountainous peneplain i& found in the eastern rim or ridge of the Nashua Valley, with a normal heio;ht of from five hundred to seven hundred feet. East of this rido;e elevations exceedino- five hundred feet are few and widely separated, and continuous areas above the upper limit of the coastal peneplain, although numerous, are individ- ually limited, except on the Marlboro ridge or height of land between the Assabet and Sudbury Rivei's. Westward, along the southern rim of the Nashua Valley, the submountainous peneplain gains gradually in height and con- tinuity as we approach the great water-parting of eastern Massachusetts, the divide between the Atlantic and the Con- necticut, with a normal elevation of a thousand to twelve hun- dred feet. From this height of land rise Watatic (1840 feet), Wachusett (2000 feet) and Little Wachusett (1560 feet), as- 539 the last vestiges, in all this region, of a still oldei' and higher land surface, not necessarily a peneplain, which is represented by the mountain peaks of New England. The Nashua Valley has, then, been carved out of the sub- mountainous peneplain during the development of the coastal peneplain. The coastal peneplain is represented by its floor, the submountainous peneplain by its eastern, southern and west- ern rims, and the still more ancient surface from which the sub- mountainous peneplain was carved, by the sentinel peaks or monadnocks risinsr from its hio;h western rim. The trenching or dissection of the coastal peneplain, many of the valleys near the coast extending far below sea-level, probably dates chiefly from early Pleistocene times, that is, from the period of eleva- tion that finally ushered in the great ice age. This last distinct or strongly accentuated stage in the topographic development of eastern Massachusetts, omitting the effects of the Pleistocene glaciation, is also represented in the Nashua Valley ; for the recent explorations of the Metropolitan Water Board in the vicinity of Clinton have revealed the existence beneath the drift-covered bottom of the valley of a deep and narrow rock a;orge, A peneplain must not only slope seaward ; but after a mass- ive elevation has inaugurated a new topographic cycle the old surface disappears most rapidly in its seaward portion. Hence the submountainous peneplain is no longer traceable east of the Nashua and Bkckstone Valleys in a series of nearly continuous crest-lines, but only in widely isolated summits, which rise abruptly, by crowded contours, from the broadly and gently undulated surface which marks an advanced stage in the devel- opment of the coastal peneplain, to heights ranging from three hundi'ed and fifty to seven hundred feet. The exceptional prominence of the western end of the Blue Hill Range indicates that, probably on account of the intense hardness of the quartz porjjhyry, it was not com[)letely base- leveled on the submountainous peneplain. 540 The second base-level or coastal peneplain which, west of Worcester, we have only in a very immature stage, as narrow valleys, is broadly developed in the coastal area, its normal ele- vations rano'iDor from a hundred feet and less near the shore to five hundred feet and more forty miles inland. The consider- able slope of the first and second ( submountainous and coastal ) peneplains is an indication, as Davis has pointed out, of crustal deformation, the elevation to start a new cycle being differential and not strictly massive. Nowhei^e, perhaps, is the coastal peneplain more perfectly developed than along the southern border of the Boston Basin, from Cohasset to the Neponset Kiver, the main or quartz porphyry range of the Blue Hills Complex alone rising above it. It is not the smoothness of the inter-stream surfaces, for that may be wholly wanting, but the uniformity of the rock elevations that proves the peneplain ; and this we recognize in the level horizon lines all over the coastal area, where the drift hills or drumlins are not too thickly planted. The development of the coastal peneplain was interrupted by the marked elevation of the land which, according to Upham's theory (accepted by the writer), ushered in the glacial epoch. The elevation was sufficient to permit the formation of the deep- est fiords of our coast, including the remarkable submarine channel of the Hudson, nearly three thousand feet deep ; and sufficiently prolonged to enable the streams to cut back their channels into the land and develop the complex and mature drainage systems which characterize our topography today. From the close of the first cycle ( the development of the sub- mountainous peneplain ) dates the crest of the main range of the Blue Hills. From the close of the second cycle (the de- velopment of the coastal peneplain) date the relief and the notches of the main range and the numerous summits of the northern range and of the eastern half of the Complex. From the massive eleva,tion that inaugurated the glacial epoch, date the valleys of Weymouth Back liiver, Weymouth Fore River, 541 Monatiquot River, Pine Tree Brook, etc., and the general sub- division of the peneplain and isolation of the rocky hills so well seen where the accumulation of drift has not been excessive, as in the northern range from West Quincy westward. It can- not be affirmed that the glaciation profoundly modified the rock contours of the Blue Hills Complex ; but the chief topographic function of the ice-sheet in this area consisted in obstructing; the drainage and generally masking the outlines of the rock surface. The post-glacial depression finally left the land, after various oscillations, one hundred to two hundred feet above its level during the development of the second or coastal peneplain. The lower portions of the deepest glacial valleys, such as the two Weymouth Rivers, were submerged and a new base-level established ; and we may consider that a new cycle of topo- graphic development is now well inaugurated, at least in imper- fectly consolidated formations such as the drift hills of Boston Harbor. The bottoms of valleys now near the sea-level con- stitute the be2:innino;.of a fourth base level, holdino- the relation of a terrace to the rock gorges formed during the glacial uplift, which it is not improbable will be matured, if the existing equi- librium proves sufficiently stable, before the monument of the first cycle which we have in the main crest-line of the Blue Hills has been wholly effiiced. This section may, then, be summa- rized as follows : — Southern New England exhibits besides the bottoms of the buried glacial gorges, and the ancient land represented by Mts. Wachusett, Monadnock, etc., three base-levels in different stages of development and effacement, the development of the younger plains and the effacement of the older plains being in general inversely proportional to the distance from the sea. The first base-level or peneplain is still well developed in west- ern and central Massachusetts, but is represented only by iso- lated hills, including the Blue Hills, in the coastal area. The development of the second peneplain, which arrested the devel- opment of the first, was well advanced over eastern and slightly 542 80 over central and western Massachusetts when it was, in its turn, interrupted by the great oscillations of the glacial epoch with the final establishment of the existing base-level and the advent of the third cycle, or the fourth cycle, if we take account of the buried valleys of the Charles, Neponset and other rivers and the floor of Boston Harbor. THE SURFACE GEOLOGY, OR THE BLUE HILLS COMPLEX IN PLEISTOCENE TIMES. Although the Blue Hills Complex presents, superficially, some general contrasts with the areas described in Parts 1 and 2, such as its greater extent, more inland character and more diversified and rugged topography, all of which must tend to influence and differentiate its surface geology, this chapter may still be cast in the same general mold as before, except that it is necessary now to take more particular account than in Part 2 of that most important factor in the later glacial history of this part of the Boston Basin — Lake Bouve. This is the name which, in honor of Mr. T. T. Bouve, former President of the Boston Society of Natural Histoiy, friend of science and humanity, and a life- long, enthusiastic and successful student of the geology of the South Shore, I have had the satisfaction, in association with Mr. A. W. Grabau, of giving to the considerable body of water which was impounded along the margin of the great ice-slieet dur- ing its recession northward from the southern water-parting of the Boston Basin. Its intricate and chano-ino- outline, its liead- lands, islands, successive outlets and levels, its great ice barrier and the icebergs which must have adorned its surfiice, and above all the tributary torrents of glacial water bursting forth from the ice, surcharged with glacial detritus, and the wonderful grace and beauty of the resulting winding eskers and lobate and dimpled delta plains, certainly make the history of Lake Bouve 543 one of the most interesting and romantic chapters in the geology of this region. The features and history of Lake Bouve have been worked out in detail by Mr. Grabau ; and it has afforded me much pleasure to be able to intrust this task to such compe- tent hands. GLACIAL GEOLOGY — OCCUPATION PHENOMENA. It is proposed to describe in this section those phases of the surface geology due to the actual and complete occupation of this area by the great ice-sheet ; or, in other words, those features of the bed-rock surface and of the drift which were developed while the ice still rested heavily upon the land and was capable of abrading and disrupting the hard rocks and transporting the resulting debris. More specifically, these phenomena include : glacial striae; glacial pot-holes; the till or unmodified (un- washed) drift and the special accumulations of the till. known as drumlins ; and erratics or glacial bowlders. Grlacial Striae. — As for the more eastern areas, these rec- ords of the movement of the ice-sheet fall naturally into two groups ; first, those agreeing closely with the normal direction of the ice movement for this region (S. 20°-25° E.) ; and, sec- ond, those deviating from the normal in an easterly direction. The first group is especially characteristic of the more elevated tracts on the rim of or wholly south of the Boston Basin, or more generally still, of the areas, be they high or low, where the movement of the ice was uninfluenced by the easterly trend of the Boston Basin. The second group, on the other hand, is almost wholly confined to the basin, at least in its topographic sense ; and they record the influence of the southern wall of the basin in deflecting the ice currents. In the following table, the normal striae are given first, followed by the deflected striae ; and the latter, it will be observed, occur chiefly on the lowlands, or, more exactly, just where they should, with reference to the . 544 higher land, to accord with the theory of glacial deflection. It is clearly not necessary to suppose that the normal and deflected striae are strictly contemporaneous ; but it is quite as probable that the deflected striae date in part from a time after the nor- mal movement of the ice had ceased : on the same general prin- ciple that the striae on stoss slopes must be, as a rule, later than those on lee slopes, the ice, in its later stages, having refused to follow closely and forcibly the lee contours. Close compari- sons are, of course, difficult and unprofitable on account of the comparatively wide range of directions often to be observed on a single outcrop ; and an attempt has been made to obtain the average trend in each case. Directions of Glacial Striae. North side of Great Blue Hill, on quartz porphyry . . S. 23° E. Summit of Buck Hill, on quartz porphyry . . . S. 25° E. Pine Tree Brook, on Cambrian slate, .... S. 20"- 23° E. Jones and Desmond's granite quarry, West Quincy . S. 27° E. Dean and Horrigan's granite quarry, West Quincy . S. 25° E. Babel Rock, west of Willard St., on trap . . . S. 20° E. Pine Hill, on Cambrian slate S. 17°-23° E. Pine Hill on Cambrian slate, another ledge . . . S. 25°-28° E. Mill Cove, on Cambrian slate S. 23°-28° E. North end of Weymouth Great Pond, on granite . S. 22° E. Main St., south of Washington St., Weymouth, on granite S. 25° E. Corner Winter and Main Sts., Weymouth, on granite . S. 25° E. Washington St., south of Whitman Pond, on granite . S. 23°-25° E. Washington St., west of Whitman Pond, on granite . S. 30° E. Between Commercial and Middle Sts., Weymouth, on granite S. 30° E. Pleasant St., east of Whitman Pond, on granite . . S. 33° E. Northwest slope of Broken Hills, on quartz porphyry . S. 30° E. Northeast base of Wampatuck Hill, on aporhyolite . S. 35° E. North side of North Common Hill, on Cambrian slate . S. 30° E. East side of North Common Hill, on Cambrian slate . S. 38° E. North side of Furnace Brook, Quincy, on slate . . S. 33° E. Hancock St., north of Eurnace Brook, on slate and trap S. 37° E. Washington St., Quincy, on Cambrian slate . . . S. 35° E. East Braintree, on the boundary dike . . . . S. 35° E. Raccoon Island, on slate S. 50° E. Grape Island, on slate S. 60° E. 545 Undoubtedly the most notable exposure of a glaciated surface in this area is the high sloping ledge of granite with inclusions- of basic porphyry on the northeast side of the Pine Hill tract,, on the southwest side of the railroad between West Quincy and Braintree, opposite the end of Liberty Street. It forms a steeply sloping and undulating wall thirty to forty feet high and several hundred feet long, which has been uncovered during the grading of the railroad, and which was so well polished by the movement of the ice that it fairly shines in the sunlight. Glacial Potholes. — -Although situations extremely favor- able to the formation of potholes by subglacial streams, or streams falling through crevasses in the ice-sheet (moulins) to become subglacial, are very numerous, only one good example has been observed. This was brought to my attention by Mr. John J. Loud of Weymouth, while it was temporaril}^ exposed during the excavation for the foundation of the new Congrega- tional Church on Commercial Street near Quincy Avenue, in East Braintree. It was, unfortunately, destroyed by the blast- ing of the ledge, and its site is now covered by the southeast corner of the church. It was formed in the large inclusion of slate in the great boundary dike (p. 405), and close to the con- tact with the trap, which rose eight to ten feet above the pot- hole, on the north, making the situation almost an ideal one. The hole or basin was 3 feet deep, 12 feet long (east-west), and from 4 to Q^ feet wide. It resembled a huge bathtub, ex- cept that it was entirely open at the west end, the smooth, water-worn surface of the pothole rounding out in tliat direction and blending with the general surface of the ledge, which here showed several incipient potholes. It is quite evident that the ice was broken by its passage over the crest of this east-west rock ridge, and that the pothole marks the point where a super- glacial stream fell through the resulting crevasse and, impinging as a glacial moulin upon the sloping surface of the ledge, exca- vated the pothole and flowed away to the west and south in its subglacial course. OCCAS. PAPERS B. S. N. H. IV. 35. 546 There are many points in this region, and these are not wholly confined to low levels or areas of modified drift, where the recently uncovered ledges present in part at least a distinctly water-worn aspect. This phenomenon is so wide-spread that it is needless to discuss its distribution or cite many examples. It first distinctly attracted my attention while I was studying the slate of Raccoon Island ; and the following description is taken directly from my field notes. " Wherever the slate has been exposed in recent times by the washing away of the drift, it presents a smoothly rounded or flowing and unmistakably water- worn surface, with many incipient potholes. My first thought was that it must be due to the action of the sea in recent times ; but I soon found that it extends considerably above the high tide level and probably all over the island. I also observed that these same surfaces are distinctly glaciated. The striae (S. 50° E.) are not very strongly marked, but can be seen on all stoss and the gentler lee slopes. The whole eflfect is clearly due to the combined or simultaneous action of the moving ice and subglacial waters at the close of the ice age, when the rap- idly wasting ice-sheet was moving feebly and with but little abrasive power." It was not so much a cooperation as a con- flict of ice and water, each agent tending to efface the distinctive characters engraved upon the slate by the other, the water, in this instance, being sufficiently in the ascendant, so that lee slopes are nowhere craggy or broken. Till and Drutnlins. — In the more eastern areas — Nantas- ket, Cohasset and Hingham — the till or bowlder clay not cov- ered by modified drift occurs chiefly in the form of drumlins ; and this relation holds also for the lower eastern and northern portions of the area now under consideration, or, more generally, within the range of the lower levels of Lake Bouve. But it is safe to assume that the broad delta plains of Lake Bouve have throughout a more or less continuous foundation of bowlder clay, although few sections are deep enough to expose it ; and the nearly continuous mantle of bowlder clay, existing independ- 547 ently of the clrumlins, over the higher land to the south and west, and notably in the Blue Hills, means simply the absence of the lake deposits. The general map (PL 13) is a graphic presentation of all the essential facts concerning the distribution, forms, sizes, elevations and trends of the drumlins in the area of the Blue Hills Complex. It shows at a glance that in their distribution the drumlins favor the lowlands along the north or Boston Basin side of the Complex, and the east rather than the west, having their largest and most perfect development in the district contiguous to Boston Harbor. These data are I'epeated on the map of Lake Bouve (PI. 25), which also shows the rela- tions of the drumlins to the sand plains or deltas and the succes- sive shore lines of the lake. Closer attention to the facts shows that the drumlin form of the till fades out somewhat gradually as we recede from the great valley of the Boston Basin. On many of the more promi- nent rock hills the stoss slopes, and to some extent the lee slopes, bear distinctly drumloid accumulations of till, suggesting that the normal or fully developed drumlins probably have, as a rule at least, elevated rock foundations or cores, which served as gathering-points for the till. As examples of these embryo drumlins we have the northwest or stoss slopes of Payne's Hill in Quincy, Liberty Hill in East Braintree, Monatiquot Heights in North Braintree, ChickataAvbut Hill, Hawk Hill, Buck Hill, and to a less extent other members of the main range of the Blue Hills, and Ponkapog Hill in Canton. Where the rock contours are not strongly accentuated or dif- ferentiated, the drumlins, especially if closely grouped, become subdued in relief, tending to flatten out into gently undulating plains of till. This is well illustrated by the drumlins of North Braintree, east of the railroad, and especially by the group crossed by Middle Street, east and south of the Monatiquot River. In other cases there is a manifest tendency to form low ridges of till, like long-drawn-out drumlins, trending in the di- I'ection of glacial movement ; and we have an exceptionally good 548 example of this type in the low, but distinct and continuous rido'e extendino- alono- the west side of the railroad, and on which Washington Street runs, for two miles from the bridge over the railroad in Xorth Braintree to South Braintree. The district south of the Blue Hills, between Great Pond and Ponkapog Pond, is a good example of a broadly undulating to flat till sur- face, devoid of drumloid character, and probably due to a moderate thickness of till rather evenly distributed. From the map it can readily be learned that, as usual, the drumlins trend rather more to the easterly, as a rule, than the glacial striae, showing that the formation of the drumlins represents a later period in the history of the ice-sheet, when it was experiencing basal melting and was depositing the ground moraine or till, and when its thinness made it more susceptible to topographic influences. A more particular discussion of the drumlins of this area may be conveniently reserved for Part 4, Avhere we shall have a broad- er and safer basis for generalization. In closing this section it may be noted that, almost without exception, the hills compos- ing the main range of the Blue Hills, and moi-e especially those of the higher western half of the range, are more or less drum- loid in form, or, more exactly, they exhibit typical roche mou- tonee contours. The northwestern or stoss slopes are long and comparatively gentle, this eflect being heightened by an increas- ino- accumulation of till downward : while the southeastern or lee slopes are usually less encumbered by drift and are dis- tinctly more abrupt, or even broken and craggy. These features are beautifully shown on the topographic map of the Blue Hills Reservation (PL 14). The drumlins fairly included within this area afford no natural or artificial sections deep enough to satisfactorily expose the unoxidized or blue till, which in other parts of the Boston Basin has yielded a goodly variety of fossil shells. Such scanty paleontological data as have come to light in the drumlins of this area will therefore be reserved for consideration with tlie more abundant data of Part 4. 549 Bowlders and DiKviqited Ledges. — In a glaciated area so diversified topographically and geologically as this, and embrac- ing several formations, like the conglomerate of the Boston Basin and the normal granites, exceptionally well adapted to the formation of large bowlders, erratics are necessarily a promi- nent feature of the sm^face geology. A few of the bowlders scattered over the Blue Hills Complex are interesting on ac- count of their size, and this is especially true of House Rock, in Weymouth ; others are interesting for the distance or direc- tion of their transportation ; and many others simply for their pi-ofusion on limited areas. As in Hingham and Cohasset, the great majority of the bowlders ai^e of extremely local origin, being within, say half a mile, of the parent ledges, and only a small proportion of the whole can be proved to have travelled as much as five miles. Along the north side of the Blue Hills, in Milton, and espe- cially near Canton Avenue, west of Randolph Avenue, there is a great profusion of conglomerate bowlders, some of which are of considerable size ; but in walking over the main range of the Blue Hills I have been impressed by the infrequency and the small size of the conglomerate erratics, although they are fairly common over the lower northern hills of normal granite. This shows most conclusively that in its later movement, at least, the ice transported but little coarse drift directly across the Blue Hills; although, as we have seen, the glacial scorings on the main range hold rather closely to the normal trend. The ex- ception that proves the rule is a block of coarse conglomerate twenty feet square and twelve feet thick on the northeast side of Chickatawbut Hill. The largest other conglomerate erratic which I have observed along this line lies a little beyond the eastern end of the range, on the northwest side of West Street, about midway between Willard Street and the railroad, and about 110 feet above the sea. It consists of normal pudding- stone alternating with layers of sandstone ; and its horizontal dimensions are approximately fifteen by twenty feet. It really 550 lies in the transverse West Quincy and Quincy Adams valley, and, with many smaller masses of conglomerate in this broad depression, it testifies to the influence of the Blue Hills upon the later glacial movement. Of conglomerate bowlders which appear to have passed through this depression there are some notable examples. One of these, some twenty feet in length, lies at the northeastern base of Monatiquot Heights, and a still larger one rests upon the highest part of the hill, directly north of the stand-pipe ; while another, not so large as these, lies upon the northwest slope of the drumlin south of Payne's Hill and crossed by Elm and Commercial Streets, Perhaps the most striking evidence that the main range of the Blue Hills was an important factor in differentiating the movement of the ice-sheet is afforded by the bowlders of red felsite derived, pre- sumably, from the ledges of this unique and attractive rock in Hyde Park and the Mattapan district of Milton. The most easterly outcrops of the red felsite are in the vicinity of Blue Hill Avenue. It is possible, of course, although highly im- probable, that it also occurs in place beneath the broad drift plain of East Milton and North Quincy ; but in my opinion we are justified, provisionally at least, in referring the erratics of red felsite to the known outcrops of the Neponset Valley. Now this felsite is a hard, massive rock, making, as a rule, rather small bowlders, which may be described as long-lived and capable of withstanding the wear and tear of glacial trans- portation. The noi'mal movement of the ice would have car- ried these bowlders, together with the conglomerate bowlders, directly aci'oss the main range of the Blue Hills. A single example has been observed at the northern base of Great Blue Hill, just above the junction of Canton and Blue Hill Avenues with Brush Hill Road, and several others were noted near the eastern end of the Pine Tree Brook belt of slate. But on the main range of the Blue Hills the red felsite, even more than the conglomerate, erratics are conspicuous by their absence ; while around the eastern end of the hills and throuo-h the West 551 Quincy and Braintree valley tbey occur with a fair degree of frequency. No notes have been made on the smaller examples, but I have been particularly impressed by two unusually large bowlders of the red felsite occurring on the low drumlins of North Braintree, in the area east of Middle Street, north of Union Street and south of Monatiquot River. Lithologically, they are identical with the banded and brecciated red felsite of the ledges in Hyde Park and Milton. One of them is on the northeast slope of the most easterly drumlin of this group, about seven hundred and fifty feet south of the railroad and six hundred feet west of the mill-pond. It measures eight by ten feet horizontally, and is evidently deeply buried. The other one is about half a mile south of this, very near Union Street and nearly a thousand feet east of Middle Street. It measures ten by eighteen feet and is seven feet high. These bowlders are fully six miles in an east-southeast direction (S. 65°-68° E.) from the nearest known outcrop of this rock, on Blue Hill Ave- nue ; and in this course, which crosses Rattlesnake and Pine Hills, no red felsite bowlders have been observed, except in the low land north of the Blue Hills and in the upper valley of Pine Tree Brook. There can be little doubt, therefore, that the course actually travelled by these two bowlders during their glacial transportation was first nearly east for three miles to East Milton and the West Quincy valley, and thence about southeast through this transverse valley three miles or more to the points where they now rest. These facts of glacial trans- \ portation confirm the indications of tlie glacial striae and the drumlins, and make it reasonably certain that the main range of the Blue Hills was a controlling' factor in determinino- the local movement of the ice-sheet, at least in its lower levels. Above the level of the crest of the range (three hundred to six hundred feet) the ice probably held closely to its normal trend, due to the general slope of southeastern New England ; but below the sky-line of the hills, on the north or Boston Basin side, the prevailing movement, at least during the closing stages 552 of glaciation, was easterly to the first important break in the range, where the contiguous portion of the basal ice regained to some extent the normal trend, although the main body swept on to the eastward, impelled by the deepening basin of Boston Harbor. In his study of the distribution of the bowlders of red sand- stone derived from the eastern end of the Norfolk Basin, Mr. M. L. Fuller^ has observed for the eastern or left-hand boundary of the train a decided eastward trend, the easting increasing from about S. 40° E. near the source to nearly south 80° E. at a distance of six miles, thus describing a curve concave to the north, with an average easting of S. 60° E. This agrees very closely with the results of my study of the distribution of the red felsite bowlders in Hingham,^ the average trend of the north- ern border of the fan being N. 67° E. for the first four miles and still more easterly for greater distances. These examples of the distribution and dispersion of erratics evidently demand a very different explanation from that pro- posed above for the red felsite and conglomerate of Milton, and an explanation in which topography is not a prominent factor. It is obvious that the erratics near the parent ledges, say within five miles, must record, in general, either the movement of the basal portion only of the ice-sheet, in case the detritus was sub- glacial or confined to the lower part of the ice, or the later movement chiefly of the upper part of the ice-sheet, in case the detritus had o-ained a hio;her level in the ice and was truly engla- cial. That the motion of the basal and upper portions of a glacier, as of a river, may be differential in both velocity and direction, is very generally conceded ; and this principle is be- lieved to account satisfactorily for the occurrence of the bowlders of red felsite in Braintree, the coarser part at least of the detritus of both the felsite and the conglomerate of the Neponset Valley having generally failed, before reaching the latitude of the Blue I Proc. Boston soc. nat. hist., vol. 28, pp. 251-26-1. ■■iPart 2, p. 272. 553 Hills, to gain a level in tlie ice liigli enough to free it from the localized basal currents and enable it to share the normal on- ward movement of the ice. South and east of the Blue Hills, the reliefs of the low-lying peneplain were manifestly insufficient to exert any important control over the movement of even the basal kiyers of the ice ; and some other cause of the decided eastward movement of the erratics must be invoked. The general lack of agreement be- tween the indications of glacial movement afforded (1) by the glacial striae (S. 20°— 30° E.), and (2) by the drumlins and erratics, is readily explained by the principle that the striae date mainly from the period of maximum glaciation, before the beginning of glacial melting and the deposition of the ground moraine, while the building of the drumlins and the final dis- position of the erratics occurred mainly during the waning and final disappearance of the ice and after it had ceased to effec- tively abrade the bed-rock. Xow it is a reasonable, if not a necessary, assumption that during this waning stage, when the ice-sheet was greatly reduced in thickness and its margin was near instead of remote, and, lubricated by subglacial waters, it slipped over the ground moraine instead of dragging it along, it was somewhat diverted from its former course by the basins of Boston Harbor and Massachusetts Bay. This change in direction, which, whatever its cause, is as certain as any fact in the geology of this region, and cannot be reasonably referred to the details of the topography, goes far also to explain the fanning out or radial dispersion of the erratics. During the period of maximum glaciation the bowlders were dragged by the ice along lines corresponding to the glacial striae, while dur- ing the period of waning and more easterly movement, when striation had practically ceased, they were dropped by basal melting and the ice flowed over them. It appears entirely un- necessary, therefoi^e, in this region at least, to rely for the explanation of radial dispersion upon the principle proposed by Chamberlain that the movement of the ice-sheet tended every- 554 where to be normal to the adjacent portion of its lobate margin. In fact, it is probable that during the waning of the ice-sheet, which is the only period we need take account of, the ice upon this plain country had gradually ceased to move befoi'e ablation had brought its margin near enough to the ledges in question to exert an appreciable influence upon the direction of flow. On the contrary, it is probable that, in accordance with the principle, also proposed by Chamberlain, that the ice tends to become stagnant first upon the uplands, where it has the least thickness and power, the ice diverted toward Boston Harbor by the Blue Hills, continued to flow eastward through that trough after it had become stagnant or had completely disappeared upon the peneplain to the south. The most favorable opportunity for observing far-travelled erratics and their courses in crossing the Boston Basin, is afforded by the marine sections of the drumlins of North Wey- mouth, Hough's Neck, and the neighboring islands. These are on a broad area of slate ; and, in accordance with the prin- ciple that the drift is mainly local, we find that the slate detri- tus very largely predominates. There are, however, many large bowlders of conglomerate from Hough's Neck, Moon Island and Squantum, one of these on the extreme west end of Grape Island measuring twelve by twenty feet, and six to nine feet thick. But of more special interest are the occasional bowl- ders of granite and felsite, which it is certain have crossed the entire breadth of the Boston Basin — twelve to sixteen miles — from the ledges of Medford, Maiden, etc. The most west- erly outcrops of felsite on the north side of the basin are in Medfoi^d ; but the lithologic character of the felsite erratics shows that very few have come from further west than Maiden and Melrose; and it is, therefore, a safe conclusion that these rather far-travelled masses have, during their transit across the basin, held very closely to the normal south- southeast course of the ice-sheet. A large majority, if not all, of tliese masses were certainly englacial during the main part of their transit ; . 555 and referring once more to the differential motion of the upj^er and basal portions of the ice due to the topographic influence, the generalization is suggested that for the drift as a whole, with, of course, many exceptions, the transportation of the local and relatively or actually basal drift was more aberrant in trend than that of the far-travelled englacial drift, so that the latter may sometimes be the more readily and accurately referred to its source. In other words, all drift is at first subglacial or basal, and its movement is controlled to' some extent by the rock contours ; but before it has travelled far it becomes englacial and shares the more rapid and rectilinear normal movement of the ice. I have elsewhere ^ shown that the modified drift is probably derived chiefly from englacial drift during the superficial and basal melting or ablation of the ice ; and hence, in accoi'dance with what precedes, it should be in general of less local char- acter. To test this influence, I examined, with the aid of several students in the Massachusetts Institute of Technology, the composition of the esker forming Hunt's Point on the northwest shore of Weymouth, in the part which has been extensively excavated. North of this point, in the line of gla- cial movement, are three broad belts of rocks : First, slates and conglomerates of the Boston Basin (Carboniferous), about thirteen miles ; second, hornblendic granites, diorite and fel- site, with some Cambrian slate and quartzite, eight to ten miles ; third, mica schists, muscovite granites and gneiss, pegmatite, etc., extending into New Hampshire. We found on lookino; over some tons of material that of all v/hich was coarse enough for easy identification about 50 per cent, is from the first belt, 40 per cent, from the second, and 10 per cent, from the third. Hence it is probably safe to assume that more than half the coarse material (cobbles and bowlders) of the modified drift is five to ten miles from its source, and a good fraction as much as twenty miles. 1 Amer. geol., vol. 17, pp. 203-234; Tech. quart., vol. 9, pp. 116-144. 556 "Of conspicuous bowlder trains or patches which cannot be regarded as mere talus accumulations, the Blue Hills afford a number of good examples, which are quite clearly shown on the topographic map (PI. 14), occurring, as a rule, on the lee or southeast sides of prominent rock elevations. In other cases they are clearly residuary portions of the till w^here it has been exposed to the wash of subglacial or other torrents. A very typical example of this kind occurs on the north side of West Street, in Braintree, southeast of the Quincy Reservoir; and there are many others, as, for instance, between the drum- lins on the north side of Elm Street near Adams Street in Braintree. Of bowlders of local origin arid especially notable only for size or situation, there are many besides those which have been mentioned. Along the south side of the Blue Hills, and par- ticularly of Great Blue, are many large and often quite angular bowlders of quartz porphyry indentical, lithologically, with the neighboring ledges. Some of these are fifteen to twenty feet in diameter. On the southern slope of Rattlesnake Hill is a block of the black quartz porphyry (p. 363), approximating a twenty-foot cube ; but, although detached, it rests upon a ledge of the same character and is virtually in situ. In the little swamp immediately west of Wampatuck Hill is another bowl- der of quartz porphyry, which is nearly a fifteen-foot cube. One very large bowlder of the greenish fluidal felsite was noted on the northeast side of Cedar Swamp, near the angle in the Braintree and Quincy boundary. Of exceptionally large granite bowlders in the Blue Hills several may be mentioned. Two of these occur lying side by side a little south of Pine Tree Brook and nearly half a mile east of Randolph Avenue. A granite bowlder twenty feet long lies on the north side of the carriage road south of Rattlesnake Hill ; and another, fifty-six feet in circumference and eight feet high, rests upon the sum- mit of Babel Rock, northeast of Rattlesnake Hill. South of the Blue Hills, on the coarse, basal conglomerate 557 of the Norfolk Basin, are many large and handsome bowlders, which are to a large extent practically hi situ • — residuary masses and not glacial erratics. They are commonly from fifteen to twenty-five feet or more in diameter ; and in some cases, as in the woods north of Ponkapog Pond, they are so closely grouped as to constitute very typical and picturesque " rock cities." East of the Blue Hills there are but few large bowlders ; and the only one demanding farther mention is that giant amono^ the bowlders of Massachusetts — House Rock. This majestic erratic stands, a conspicuous feature of the landscape, about half a mile northwest of Whitman Pond, in the angle east of Essex Street and south of Broad Street (see the map), at the northwestern base of a low hill of the biotitic normal granite (PL 28, Fig. 1), while near by on the north rises a small but very typical sand plain of the Whitman series (95 feet), (PI. 25). The top of the bowlder overlooks both of these topographic features. It rests upon a northward sloping and distinctly glaciated ledge of granite : and it is, therefore, a true erratic and not in any sense a merely residuary mass. Viewed from the northeast (PI. 28, Fig. 2), the bowlder pre- sents a rather striking profile, first noted by Mr. John J. Loud, to whom I am indebted for this illustration, which has caused it to be known locally as the Weymouth Sphinx. The max- mium horizontal dimensions of the bowlder are : north side, 37 feet ; east side, 25 feet ; south side, 42 feet ; and west side, 42 feet. The circumference, obtained by applying a tape line at a height of six feet, is 126 feet. The height, determined by a hand level, is 37 feet, the top of the bowlder rising 3 feet above the summit of the hill at the base of which it lies. In volume, House Rock must approximate a cube 35 feet on an edge, or over 40,000 cubic feet. At 170 pounds per cubic foot, the weight is about 7,000,000 pounds (3,500 tons). It is thus nearly half as large as the great Madison bowlder of New Hampshire. Lithologically, House Rock is indistinguish- able from the ledofe on which it rests and the orranite which is 558 quarried on the hill above ; and similar granite occurs in ledges to the northward for three quarters of a mile, or nearly to Commercial Street. Hence, although the great bowlder has clearly not been derived from the hill on the southeast, and is an undoubted glacial erratic, the probabilities are very great that it is not far-travelled. In fact, the parent ledge is quite certainly one of the group of ledges on the west side of Essex Street north of the almshouse ; and the Weymouth Sphinx is now, probably, between one half and three fourths of a mile from its birthplace. The power of the ice-sheet to rend the solid ledges and bear away great blocks of stone is well illustrated by many half disrupted ledges on the lee slopes and in the chasms of the Blue Hills ; but the most striking example of all is that to which my attention was first directed by Mr. John J. Loud. It is in the woods on the east side of Whitman Pond, just south of the southern border of the diorite, where the biotite granite is well sprinkled and in part crowded with the segregations and inclusions of diorite. The western face of a hig-h, bold ledo^e of the granite has been sufficiently disrupted by the last efforts of the ice, to form what is virtually a roofless cavern which may be entered through a narrow crevice about twenty-five feet long and closed above. Another example, equally instructive but less impressive, occurs on the south side of Washington Street about one fourth of a mile east of Lovell Corners. In view of these facts it is plain tfiat in the production of its freight of bowlders the ice-sheet must in some instances have quarried away whole hills of solid rock, so that it is not sur- prising that in the case of such a giant as House Rock the parent ledge was destroyed in its making, and nothing remains to mark the site. In closing this section, we may notice an interesting perched bowlder of granite beyond the southern limit of the map. It rests (PI. 29) on the sloping surface of a bold ledge of granite on the south side of Oak Street, near the Hingham boundary. It is 13 feet long, to 8 feet wide, and 7 to 8 feet high. Occas. Papers, Boston Soc.Nat.Hist.,Vol.Iv: Plate ?.B. THE HELIOTYPE PRINTING CO.. BOSTON. HOUSE ROCK, WEYMOUTH. Occas. Papers, Boston Soc. Nat. Hist., Vol. IV. Plate 29. PERCHED BOWLDER ON OAK STREET, WEYMOUTH. 559 GLACIAL GEOLOGY — RETREATAL PHENOMENA. The retreatal phenomena of the ice-sheet are recorded with greater or less clearness in the varied phases of the washed or modified drift, including : first, the eskers and kames, marking the courses of glacial streams ; second, the over-wash plains and marginal plains or terraces ; third, the delta plains, or the de- posits of glacial lakes and the kettles , by which they are diversi- fied ; and fourth, the glacial lakes themselves, with their succes- sive outlets and shores. For this area, however, the other phenomena are, practically, all comprised in a study of the gla- cial lakes, and do not demand separate consideration. Besides Lake Bouve, w^hich occupied the entire sweep of the southern watershed of the Boston Basin for nearly a dozen miles, from Randolph Avenue south of the Blue Hills to the eastern boundary of Hingham, a much smaller lake — a glacial lakelet — nestled in the concave curve of the northern slope of the main range of the Blue Hills during the -time when the mar- gin of the ice-sheet formed the chord of this arc. The Glacial Lake of the Blue Hills. — This little lake, which was probably contemporaneous with and tributary to the earlier stages of Lake Bouve, left its record in the modified drift on the northern slope of the main range, midway of its length, or between Hancock Hill on the west and Chickatawbut on the east. The extreme length of the free ice-barrier was about one and a half miles. But at no time, probably, did the water area much exceed a mile in length by half a mile in breadth, following approximately the contour of 230 feet (see PI. 14) . The highest outlet of the lake is the one which, in the general northeastward recession of the margin of the ice- sheet, was, naturally, uncovered first, — viz., Ponkapog Pass between Hancock Hill (510 feet) and Tucker Hill (449 feet), with an elevation of about 230 feet. The deepest line or axis of the pass lies close along the base of the abrupt southeastern 560 slope of Hancock Hill, some six hundred feet northwest of Hill- side Street. During the existence of this outlet the water was tributary southward, down the valley of Marigold Brook to Hoosicwhisick Pond (165 feet), and thence eastward by the valley of Monatiquot Stream to Lake Bouve. The Hoosic- whisick plain (175 to 180 feet) enclosing this large kettle pond, indents the southern margin of the Blue Hills and borders the valley of Marigold Brook as a well-defined terrace northward for about one third of a ,mile, beyond which the valley rises fifty feet in another third of a mile to the summit of the pass. This outlet, although quite certainly the principal one, was probably short-lived : for, unless the ice margin were stationary for a time, its recession must soon have uncovered Randolph Pass (250 feet), and, a little later, Braintree Pass (220 feet) at the head of Blueberry Swamp and the southern base of Chickatawbut Hill ; and the latter event involved, necessarily, the abandonment of the Ponkapog outlet, the glacial waters becoming tributary to Monatiquot Stream and Lake Bouve, east- ward through Ced'ar Swamp, instead of southward thi'ough Hoosicwhisick Pond. Ponkapog Pass was an ideal outlet of the temporary moun- tain lake : and adjacent to it on the north we find the chief de- velopment of deposits formed in that lake by the glacial streams tributary to it. The small hill between the pass and Hillside Street (271 feet) is covered with till ; but immediately north of this, in the angle between Hancock Hill and Hillside Street, the modified drift is abundant, taking the form, chiefly, of dis- continuous eskers, kame-like knolls and immature plains. Two systems of eskers appear to be recognizable : a low, more or less indistinct and apparently older line trending in a south- southwest direction toward the pass, but not reaching it ; and a strongly marked and typical northwest-southeast series, cross- ing the first and tributary to an irregular and distinctly imma- ture delta plain lying northeast and east of Hillside Pond. This series really embraces two eskers — one on either side of the 5G1 brook flowing from the pond, and converging on tlie nortliwest corner of the phiin. The western esker begins on the northeast slope of Hancock Hill, and, after following this for several hundred feet, turns abruptly to the eastward, ending on the line of the supposed earlier south westward esker; beyond this it continues as a hioh and bold embankment extending; some six hundred feet to Hillside Street. The second esker of this series is closely parallel with this last link of the first one, with only the brook between ; but it cannot be follow^ed beyond or north- west of the brook. The ice margin and delta front of the plain are normally developed : but the surface of the plain is very uneven, with knobs and bowlders of the underlying till project- ing through it ; and bowlders are also a feature of the eastern esker. Both the plain and its tributary eskers are diversified by kettles, and closely correspond in height with the Ponkapog outlet. The less distinct line of esker with the southwestward trend may be supposed to date from the time when the ice al- most completely filled this basin, and the subglacial stream was necessarily diverted, at the last, from its normal southeasterly course toward Ponkapog Pass, which is fully eighty feet lower than the more direct pass between Tucker and Boyce Hills. With the birth of the glacial lake, however, the course of the glacial stream and the location of its mouth became inde- pendent of the outlet, although the latter determined even more rigidly than before the upper limit or maximum elevation of the resulting deposits. East of the eskers, on the north side of Hillside Street, is a second immature plain with a strongly undulating surface and a high ice margin on the edge of the deep valley to the north. Northward down the valley of Hillside Pond, is also much modified drift ; but the retreat of the ice margin was, appar- ently, too rapid to permit the 'development of a normal delta plain in water from fifty to over a hundred feet deep. About one fourth of a mile northeast of Hillside Pond a small brook flows northward across Hillside Street from the pass between OCCAS. PAPERS B. S. JT. H. IV. 36. 562 Boyce and Tucker Hills ; and farther to the northeast there are onlv slio;ht mdications of modified drift over the o-ently undu- lating surface of till until we have crossed the valley of Blue- berry Swamp. This intervening ground, it may be noted, is mainly above the level of Braintree Pass, which was probably opened soon after the ice retreated from the vicinity of Hillside Pond. The only noteworthy development of modified drift which may be correlated with the Braintree Pass outlet is that on the line of Randolph Avenue near Hillside Street ; and the most notable feature of this occurrence is the completely isolated mound of gravel rising twenty-five to thirty feet above the general level between the avenue and the street. The origin •of this solitary hillock is by no, means clear. Its slopes are steep ; it is elongated in an east-west direction ; and it is com- posed in part of quite coarse, angular and ill-assorted gravel, with an occasional large stone. The elevation of its summit must be about 220 feet, agreeing not only with Braintree Pass, but also with the narrow development of modified drift bordering the higher till slope on the east side of Bandolph Avenue. The hillock might be regarded as an erosion outlier of this plain or terrace ; or as a remnant of an esker tributary to the plain ; or, perhaps, better still, as a deposit formed in such a basin or hole in the stagnant margin of the ice-sheet as Russell has described in the stagnant margin of the Mala- spina glacier in Alaska.^ The comparative meagerness of the deposits in the eastern part of the basin, while possibly attributable to the absence of glacial streams of adequate power, is, I think, more satis- factorily accounted for on the supposition that, soon after the opening of the Braintree Pass outlet, the ice completely relaxed its hold on the western end of the range and the im- pounded waters subsided to the level of Lake Neponset, which was then discharging eastward across the Hoosicwhisick plain 1 Jouni. geol., vol. l.pp. 230-233. 563 at an elevation of about 170 feet. This abandonment of its ■site was the closing episode in the histoiy of the glacial lakelet of the Blue Hills. Although we seem to have in the broad and gently sloping plain northwest of Hancock and Hemenway Hills a consider- able surface development of modified drift, referable, perhaps, to the Hoosicwhisick stage of Lake Neponset, it is probably, in the main, of slight volume, a thin deposit of sand and gravel over a plain of till. And the almost complete absence of modified drift from the upper valley of Pine Tree Brook, seems, at first, to show that the ice retired rather rapidly from this part of the hills. It is more probable, however, that at this stage of the glacial retreat the glacial streams, upon which the formation of the modified drift was mainly depend- ent, were diverted to the east or west by the high relief of the Blue Hills. In accordance with this suggestion we find that the eskers along the northern base of the hills agree closely in trend with the maro-in of the northern range, the clearest example being the beautiful esker on the Cunningham estate south of Edge Hill Road in Milton. These eskers and the plains accompanying them may be most conveniently described in connection with the history of Lake Neponset in Part 4. 564 LAKE BOUVE, AN EXTINCT GLACIAL LAKE IN THE SOUTH- ERN PART OE THE BOSTON BASIN.' Br AMADEUS W. GEABAU. Introduction, Among the glacial deposits of eastern Massachusetts, the frontal accumulations of washed sands and gravels hold a pi-e- eminent position, not alone on account of their extensive development, but also because they furnish the best criteria by which the succession of events in the glacial history of this region may be determined. J. B. Woodworth, in a recent paper, ^ has given a comprehensive general account of these de- posits in southern New England, and has applied to them the name of glacial wash-plains. He recognizes four types of glacial wash-plains, dependent for distinction on their relation to the ice-margin. These he describes as follows : (a) " Fron- tal moraine terraces, with an ice-contact slope, charged with till and bowlders, a true morainal deposit." (b) "Frontal ter- races, like the preceding but lacking the till-coating along the ice contact." (c) " Esker-fans, small plains of gravel and sand built at the mouth of subglacial tunnels and channels in the ice ; associated with an esker or esker-like chain of deposits made in the ice-sheet at the same time." (d) "Wash-cones, steeply sloping deposits, with ice-contact slope on the iceward 1 The existence of the glacial lake now known as Lake Bo^uve was clearly recog- nized by Professor Crosby in Part 2 of this work; and he has, on pages 274 to 286, given a general account of its features — eskers, delta-plains and outlets — so far as they are developed in the area covered by that Part. My purpose, therefore, has been to extend this stiidy, in greater detail and in the light of the later advances of glaciology, over the entire basin of the lake. It affords me pleasure to acknowledge the cordial cooperation of Professor Crosby in the prosecution of this work, and the valuable assistance of Mr. M. L. Fuller in the study of the Monaticinot Valley. "- Bull. Essex inst., vol. 29, 1899, pp. 71-119. 565 side culminating in a liigh point, with gentler slope outward, in the manner of alluvial cones." Of these four types the frontal terraces are by far the most important^ because most frequent and most extensive. They naturally fall into two groups, the glacial sand-plain or delta, and the frontal apron plain, the former a subaqueous, the latter a subaerial deposit. Esker-fans, in so far as they fall in either the one or the other of these groups, are best con- sidered with the frontal terraces. When most typically developed, the glacial sand-plain pre- sents three types of topographic elements which are strongly contrasted, and which in the detail of their features, and in their combination, serve to distinguish the glacial sand-plain from sand and gravel deposits of similar character though different origin. These three elements are the top, front, and back slopes. In all cases, these slopes are of constructional origin, i. e., their characteristics are due to the deposition of materials derived from the ice sheet. The top slope is characterized by a low gradient, which, though variable, probably never exceeds the limit of a few degrees. It may be a nearly plane surface descending in one direction, or, what is perhaps more frequent, it may have the form of an exceedingly flat semi-cone, whose apex forms the head of the delta. The frontal slope normally bounds the glacial delta on three sides, describing a curve, varying from semicircular to elon- gately semi-oval in outline, with the longer axis parallel to the ice-front. This slope may have a vei\y moderate gradient, or it may be as high as twenty degrees or even more. In all cases, however, it has a much steeper gradient than the top slope, audit usually joins that slope so as to form with it a pronounced obtuse angle. In all but the very smallest sand- plains the outline of the front is not a continuous curve, but rather a combination of curves, giving a lobate character to the outer margin. (Plate 30.) 566 The back slope, when typically developed, is distinguished from the frontal slope by its steeper gradient, this being nor- mally the maximum angle of repose of the unconsolidated material, and by the absence of convex lobation, there being often instead a pronounced concave lobation, M^ith more or less prominent cusps between the concavities. The line of junction between the back and top slopes is typically a sharp one, the angle between these slopes being usually much less obtuse than that between the top and frontal slopes. A perfect development of the three types of slopes in a given glacial sand-plain is relatively uncommon, the form being modified to a greater or less extent. The most frequent modi- fication of the top slope consists in the presence of kettle holes, which, when perfectly developed,, are sub-circular or elongate depressions, with steep slopes comparable to the back slopes of the sand-plains. These depressions must not be confounded with the elongated, often narrow and shallow depressions or. fosses, which mark the imperfect junction of two successive sand-plains. In these latter depressions, the slopes are not similar, but present on one side the characters of a frontal, and on the other those of a back slope. Other modifications of the noi'mal top slope consist in longitudinal creases which are merely the backward extensions of the interlobate depres- sions of the frontal slopes, and may commonly be regarded as due to subsequent erosion. The modifications of the frontal slopes, unless affected by erosion, are chiefly in the angle, which, as already stated, may vary within certain limits. The back slopes are most subject to variation. This often aflPects the outline, which may range from straight to cuspate or very irregular; or it may aflPect the inclination, which may be steep, or gentle and more or less uniform, or very irregular, the surface becoming hummocky or kame-like. Finally, an esker, or a series of eskers, may join the back slope, as is habit- ually the case in esker fans. A not infrequent though acci- dental modification of the slopes, and one which affects all 5G7 three types, is found in pre-existing i-ock or till masses against or around which the plain has been built. These masses some- times entirely replace one or the other of the side slopes, but they rai'ely cause more tha« an interruption of the top slope. In structure and in the character of the component materials, the glacial sand-plains exhibit significant features. The char- acter of the o-ravel is that of sub-ano-uhir or well-rounded water- worn and water-laid material, sharply angular pebbles and those exhibiting glacial markings being generally absent except where the former are produced by the subsequent splitting of water- Avorn pebbles. The stratification is usually well marked and is divisible into topset beds, or those approximately parallel to the top slope, and foreset beds, or those approximately parallelto the frontal slope. Truncation of the foreset beds, and deposition of the topset beds upon the truncated edges of the foreset beds, is a feature of common occurrence. Irreg- ular bedding with thinning and thickening of layers, and minor cross-bedding, and ripple marks, are among the other structural features usually found. What have been denom- inated backset beds, parallel to the back slope have been described,^ but they are of slight thickness and in most cases differ genetically from the topset and foreset beds. In grade of texture the material of a sand-plain varies considerably, but in practically all cases a decrease in coarseness is observable from the head to the front of the deposit. The topset beds commonly exhibit a coarser texture than the foresets. The glacial origin of the type of sand-plain above described is today probably unquestioned, and their interpretation as deltas built into a body of standing water against a temporarily stationary ice-front, will probably be questioned by few. That they are subaerial deposits is negatived by the gentle gradient of the top slope, the abrupt change from this to the frontal slope, the lobate outline of the latter, and the relation of top- set and foreset beds. A subaerial fan of glacial origin may 1 Davis, Bull. geol. soc. Amer., vol. 1, p. 197. 568 exhibit the steep back slopes of the normal sand-plain, but the frontal slope is absent, the top slope being continuous from the head to the foot of the plain. In like manner the combination of the topographical and structural features of the sand-plains serves to distinguish them from any other known type of glacial deposits. These deltas are most nearly related to normal stream deltas built in a standing body of water. They share the combination of top and frontal slopes and internal structure, but the former are at once distinguished by the characteristic back slopes, which, in the case of the normal stream deltas, are replaced by the shore of the body of water into which the delta is built. This characteristic back slope is one of the most significant topographical features. It marks the contact with the front of a now vanished ice-sheet, against which the deposit was banked and on the removal of which, through melting, the characteristic slope was assumed by a slumping of the unsup- ported material, this at the same time producing the appear- ance of backset beds, as suggested by Upham. This ice- contact slope is not confined to this type of deposit, but is normally characteristic of all ice-bound extra-glacial deposits. (See Woodworth : The ice contact in the classification of gla- cial deposits. Amer. geol., vol. 23, 1899, pp. 80-86.) It therefore becomes a criterion by which we can determine the successive temporary stages of the ice-front, when it remained stationary long enough to permit of a frontal accumulation of "•lacial detritus. While the delta character of the type of sand-plain here xiescribed, and its intimate relation to the front of the waning ice-sheet, admit of little doubt, opinions differ as to the char- acter of the body of water in which the delta was formed. We cannot conceive that deltas of this class can form on the o[)en sea- coast, because any deposits there made will be con- tinually modified by wave action, and the perfect constructional slopes, so characteristic of these deltas, will not be retained. 569 We might, of course, assume with Simler ^ that these deposits were formed at a depth below the surface to wliich wave action does not reach, but in that case we could not exj)ect to obtain the characteristic angle between the top and frontal slopes, which, as far as observation and experiment extend, marks the average level of the water in which the delta was forming. Nor can we expect the characteristic lobate front, indicative of a shifting or forking stream mouth with distrib- utaries spreading out in all directions, each building its own independent lobe. Even if the delta could be formed without being modified by wave action while building, it could not survive the attack of the waves during the process of uplifting, unless Ave postulate a catastrophic elevation, as Professor Shaler has done in the case of the Martha's Vineyard and Nantucket plains, vv'hich are now generally regarded as subaerial over-wash or apron plains, but which he supposed to have been formed at a considerable depth below sea-level. As Professor Shaler says : " We may be sure that if they [the sand-plains] were exposed even for a few days to the action of the Atlantic surf and tides they would bear the unmistakable marks of water action" (loc. cit., p. 45). ~ , It thus appears that the formation of perfect glacial deltas on an exposed shore, and their subsequent preservation, is a mat- ter of extreme improbabili-ty. Most of the objections urged against their formation on open shores do not, however, hold in partially enclosed and protected embayments ; and in such basins sand-plains may be formed, provided tidal action is counterbalanced by a sufficient influx of fresh water from the melting ice. Fuller^ has endeavored to explain the origin of the Great Barrington sand-plain in the Narragansett Basin in this way; but Woodworth^ holds that a flooded estuary, kept 1 Geology of Nantucket, Bull. 53, U. S. geol. surv., p. 44, 2 Amer. geol., vol. 21, 1899, pp. 310-321. s Amer. geol., vol. 18, 1896, pp, 150-168, 391, 392. 570 full by excessive melting of the ice, and probably blocked by residual ice masses in its lower course, produced a fresh-water basin in which the plains were deposited above the present sea- level. Recognizing the possibility of the formation of glacial deltas in protected basins at sea-level, the probability of their being thus formed becomes very slight when we deal with a series of plains of different altitudes. For in that case we must postu- late differential elevation or subsidence, to account for the varying heights of the plains. If, as is usually the case with serial plains, successively lower series are met with as Ave pro- ceed northward, we must, if we hold to the submarine view, postulate successive slight and rapid elevations, interrupted by periods of relative stability ; and if, as is also often the case, such series of deltas in distinct, though not widely se|)arated, areas along the ice-front show no correspondence in elevation, we must further postulate a differential movement for each period, between the two localities. Thus the successive series of plains of the Weymouth-Hingham area bear no relation to the series in the closely adjoining, though separated, IS^eponset Valley, such as we should expect to find were the elevations of their top slopes determined by a temporarily higher sea-level. If, finally, as can be shown in the majority of cases in eastern Massachusetts, the area in which well-developed sand-plains occur is rimmed around by high land of rock and till, except where, following the indication of the ice-contact slope Ave can readily postulate the existence of an ice dam, and if, further- more,, notches in this rim can be correlated Avith the successive plains, the necessity for the considei'ation of the sea-level and its relation to the basin thus enclosed falls away entirely, and the theory of a fresh-water lake, ponded behind an ice dam, and of successive levels, as determined by the uncovering of succes- sively lower outlets in the rim, becomes the only one Avhich Avill account satisfactorily for the facts in question. Lake Bouve is the first of such vanished crlacial lakes in 571 Massachusetts to which a name has been applied, and of which the outline and extent and the succession of events in its his- toiy have been established. (See Part 2, p. 274 et seq.)^ The name, as stated by Professor Crosby, was chosen in honor of the late Mr. T. T. Bouve, a former President of the Boston Society of Natui'al History, a life-long resident of Hingham, and an enthusiastic student of the glacial geology of that region. Extent and Outline of Lake Bouve. Lake Bouve was situated in what are now Plymouth and Norfolk Counties, Massachusetts, on the southern watershed of Boston Bay. The area covered at one time or another by the waters of this lake comprises nearly the whole of the present townships of Hingham, Weymouth and Braintree, the eastern half of Quincy, and the adjoining lower portions of Randolph, Holbrook and Rockland. This area forms a basin open on the north, but surrounded on the east, south and west by a rim of rock and till. Following this rim from its northwestern end in the Blue Hills at West Quincy, to a point a mile or more north of Prospect Hill in Hingham, a distance of nearly thirty miles, we find no pass below the 130 to 140 foot contours. About a mile north of Prospect Hill, however, the rim becomes de- pressed to 95 feet above sea-level, and from that point north- ward to Turkey Hill in Hingham it is very irregular, becoming at one point depressed to the 60 foot contour. North of Tur- key Hill the rim of the basin quickly descends to the level of the salt marshes. The basin of Lake Bouve drains northward into Boston Bay ; and the principal streams of this watershed, beginning on the west, are : first, Monatiquot River, which reaches the bay through the estuary known as Weymouth Fore River ; second, and third. Mill River and Old Swamp River, both of which 1 Amer. geol., vol. 17, 1896, pp. 128-130; Science, n. s., vol. 3, 1896, pp. 212, 213. 572 are tributary to Whitman Pond and through that to the estuary of Weymouth Back Eiver ; and, fourth and fifth, Plymouth River and Beechwood River, the two principal branches of Weir River, which was tributary to Hingham Harbor in pre- fflacial times, but now reaches the bav through Nantasket Harbor. The upper valley of the Monatiquot River is largely of a swampy character. The upper end of the valley of Mill River is occupied by Weymouth Great Pond, out of which the river flows. The valley of Old Swamp River, again, as the name implies, is of a decidedly swampy character. The valley of Plymouth River is clogged and truncated by drift ; but its original upper course is probably represented by the basin of Accord Pond, which is now, through the accident of drift accumulation, tributary to Beechwood River. And on the water-parting, in the direct line of Beechwood River, lies Val- ley Swamp. At the heads of these streams we find, naturally, the lowest depressions or passes of the southern water-parting or rim of the basin. The western valleys are, in their upper courses, the most deeply incised ; but this more strongly ac- centuated topographic character is compensated by the increased elevations of the water-parting westward ; so that, in spite of their varying depths, we find a surprising uniformity in the elevations of the passes. The first, third and fifth passes, or those at the head, respectively, of Monatiquot River, Old Swamp River and the direct line of Beechwood River (Valley Swamp Pass) , are tl>e lowest. They vary but little in eleva- tion from 140 feet, and agree so closely that, apparently, tliey served simultaneously as outlets of Lake Bouve. The some- what greater elevations of tlie Weymouth Great Pond and Accord Pond passes, is probably due in part, at least, to greater depths of drift. The elevation of the water surface at Weymouth Great Pond is slightlv aljove 150 feet.^ The pass at the head of tlie ' Mr. Frank A. Bates, Secretary of the Boston Scientific Society, kindly furnished me with the levels of the pipe-line from Weymouth Great Pond to Weymouth town hall. The water-level of the l'on-o d - o Tli d (?q c; d t^ 5; d -■ ^ 00 + d c- ad + t- UO ;^ c^ d o d t^ ^ d -^ o CEPHALA (MEASURE lO CO 4^ 0a o O la o lO 0^ CO CO CJ m ^ o o o IM S ^ o cl U5 iCl o in C-. CO '^ CO o o ■Cj s ^ o ■a ci rH ^ "^1 l^ C5 >> S lO o CI fc "^ "^ S^ * o o la •* o IM 00 ^ CJ 05 lO ^. CO 0-. ■* n ■ 01 '-0 "^ ^ 'X lO 00 >x « C) >o ^ »-1 ^ o '^ ^ lO '^ 00 00 ^« 5 '■ ci X ■ o a rtSic: a ^5 4J^ 2a M '3 : T5 : OJ 33 2 02 • 3 ^53 f^y. X bt? : a g tjj tH O c« eS S 0) 05 - c« tea 4J^ ^^ o Cm * ^ aj.G o "rt oEc i^ a 0) ^: -u 2 u s'S -■=ss C H; c3 OJ — ^ c S'w 3 Vj H hj c h-I c C a O J^ o s ^ ti So cv 00 cc "-^ "* ^ '^ o J CO (M ^ s Ol !^ ^ "O §5 c 05 CO CO Ol 1-1 CTl 1 o <» 00 =0 g o Ifl 8 o 05 t- 00 f. ^ & o t- '"' (M ■^ o — ■* o ■* c^ lO s o CO Oi o iM 1 8 IM 00 '^ o OS CO o t- f « 5 OS o i=s t- 00 i 8 ■s s g cs 1 CO t- i? s s § o 5 CO o o c^ ■* _, Oi S3 "2 o eo s "fl _ ^ !M o t) o u. t™ 00 I^ o '^ o 00 CO tP , ^ ^ ^ Si S 9 en a "ii !t-l o '-l-i ; ® III o " S C S<^ -53 ;^rr p. c« Mo S-M ,"?=« ■S Co m g u O S o S o ■-.=. hJ C5 J O cS 692 8. Nearly complete small specimen, showing genal spines, hypostoma, etc., and 17 segments in the thorax. Coll. Bridge- water normal school (Plate 37). 9. Immature individual, with 15 thoracic segments, and with cephalon and pygidium displaced. Free cheeks wanting. Coll. mus. comp. zool. 10. Central portion of a large cephalon. Coll. mus. comp. zool. 11. Cephalon, exclusive of free cheeks. Coll. mus. comp. zool. 12. Imperfect cephalon, exclusive of free cheeks. Coll. mus. comp. zool. 13. Nearly perfect specimen, except free cheeks. Shows tubercle on occipital ring. Coll. Boston soc. nat. hist., cat. no., 2753. 14. Nearly perfect specimen, parts restored. Coll. Boston soc. nat. hist., cat. no., 2754 (Plate 39). Counterpart, cat. no., 6265. 15. Specimen, with free cheeks and great part of thorax wanting. Coll. mus. comp. zool. 16. The type specimen. Coll. Boston soc. nat. hist., cat. no., 6264 (Plate 36). 17. Individual showing part of head and anterior thoracic segments. Stud. pal. coll. Harv. univ., cat. no., 324. 18. Large specimen ; anterior portion of cephalon, one side of cephalon and thorax, and posterior part of thorax, bi'oken away. Coll. mus. comp. zool., exhibition. 19. Head shield with anterior thoracic segments and short spines. Coll. Boston soc. nat. hist., cat. no., 12040. 20. Showing cephalon and part of thorax. Coll. Boston soc. nat. hist., cat. no., 4573. 21. Fragmentar}' specimen, showing 17 thoracic segments and a fragment of the 18th, also very long, slender, palpebral lobes. One side of thorax, free cheeks, and part of head, broken away. Stud. pal. coll. Harv. univ., cat. no., 325. 693 22. Imperfect specimen showing part of head and part of thorax (with counterpart). Coll. Boston soc. nat. hist., cat. nos., 2 and 3. 23. Showing free cheek and spine. Part of head and thorax broken away. Stud. pal. coll. Harv. univ., cat. no., 556. 24. Large specimen with counterpart, the two showing nearly all the features. Coll. Boston soc. nat. hist., cat. no., 2756. 25. Large specimen with imperfect glabella and thorax. Cheek figured by Walcott. (See Plate 35, fig. 3). Coll. Boston soc. nat. hist., cat. no., 1. 26. Large specimen, showing head with part of the free cheeks, and a part of the thorax. Stud. pal. coll. Harv. univ., cat. no., 330. 27. Showing fragment of cephalon and part of thorax. Free cheeks and part of fixed cheeks wanting. Coll. Mass. inst. technology. Also other fragmentary specimens in that collection. 28. Paradoxides regina Matthew. From a cast of the type specimen. Stud. pal. coll. Harv. univ. Horizon and Locality. — In the siliceous argillites of Mid- dle Cambrian age at Hay ward Creek, near South Quincy, township of Braintree, Massachusetts, numerous specimens. Also in similar rock on the land of Mr. T. A. Watson in Weymouth, Mass., and in shaly beds at East Braintree. Thanks are due to the following gentlemen for the use of specimens of fossils : To Prof. Alpheus Hyatt for access to the specimens in the collections of the Boston Society of Natural History, and the Museum of Comparative Zoology. To Prof. R. T. Jackson for access to the specimens in the Student Palaeontological Collection of Harvard University. To Mr. Charles D. Walcott and Mr. Chas. Schuchert for access to the collections of the U. S. Geological Survey and of the U. S. 694 National Museum. To Mr. H. T. Burr for the use of his collection of North Weymouth fossils. To Mr. Boyden of the Bridgewater Normal School for the loan of the specimen of Paradoxides illustrated on Plate 37. To Mr. John H. Sears I am greatly indebted for the loan of the extensive col- lection of Nahant fossils which he has made for the Peabody Academy of Sciences. Mr. T. A. Watson has supplied much new material, and has also aided greatly in the exploration of the ledges on Pearl Street, North Weymouth. To Mr. W. W. Dodge I am much indebted for the use of his collection of Braintree fossils. Other acknowledgments are due Mr. J. B. Woodworth, Prof. W. H. Niles, Dr. E. O. Hovey, Mr. W. E. Hobbs, and Mr. Robert Burke. To Miss Elvira Wood and Miss L. R,. Martin thanks are due for the care and labor bestowed on the illustrations of the fossils. The U. S. Geological Survey granted the privilege of making electrotype copies from their original plates of Brain- tree fossils. ERRATA. Page 299, fourth line from bottom, for PL 16 read PL 20. " 300, first and second lines from top, for PL 17 read PL 21 ; for PL 18 read PL 22. " 300, fifth line from bottom, for PL 15 read PL 19. " 301, eleventh line from bottom, for PL 19 read PL 26. " 340, ninth line from the top, for PL 14 read PL 16. " 610, in the table, before "31. Aristozoe?" insert: '''' Leperditia^ cf. L. solitaria (?)"; and oppo- site it, in column 10, insert ; " r." Occas. Papers, Vol. IV. PLATE 31. (Figures 3, 4, 6 a-c, 9 a-f, 10 c, 11 b, and 12 a were drawn by Miss Elvira Wood, Instructor in Palaeontology in the Massachusetts Institute of Tech- nology ; the remainder were drawn by Miss L. R. Martin. ) Eigs. 1 a-d. Acrothele gamagei (Hobbs.) P. 615. 1 a. Pedicle valve showing a large foramen just posterior to the beak. The area is not well shown. X 2. 1 b. Pedicle valve showing a ratlier abrupt flattening of the posterior portion.' There is a suggestion of a faint median pseudo-deltidium. No foramen is shown. X 2. 1 c. Pedicle valve with the sides merging posteriorly into the flattened area. No foramen is shown. X 2. 1 d. Exterior of a brachial valve showing the hinge-line and the con- cavity on either side of tlie beak. Paradoxides beds, Hayward's Quarry. Coll. W. W. Dodge. Eig. 2. Iphidea bella Billings (?). P. 617. Ventral valve, showing area and pseudo-deltidium. X 1^. Nahant limestone, Nahant, Mass. Coll. Peabody acad. sci., cat. no., 705. Eig 3. Obollelacrassa (Jisll) {?). P. 619. Surface and lateral views. X 2. Nahant limestone, Nahant, Mass. Coll. Boston soc. nat. hist., cat. no., 11,970. Eig. 4. Lingulella rogersi Walcott. P. 624. Valve from the original specimen described by Professor Rogers. (From a gutta percha cast.) X 2. Quartzite pebble, from conglomerate, Fall River, Mass. Coll. Boston soc. nat. hist., cat. no., 2750. Eig. 5. Parmophorella acadica (Hartt). P. 625. A crushed and broken shell. Paradoxides beds, Hayward Creek. Stud. pal. coll. Harv. Univ., cat. no., 503. Eigs. 6 ar-c. Scenella (?) sp. P. 627. Summit, lateral and posterior view of a specimen from which the shell has been removed. X 8. Red Lower Cambrian limestone bowlder. Pleasant Beach, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,971. PLATE ol (continued). Figs. 7 a, b. Platyceras prlmaevum Billings. P. 028. 7 a. The shell from the apical end of the spire. The centre is broken away. X 12. 7 b. Dorsal view of the same, showing the obliquity of the body whorl, its angulation along the dorsum, and the rapid widening toward tlae aperture. X 12. Red Lower Cambrian limestone bowlder, Sandy Cove, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,960. Figs. 8 a, b. Platyceras cleflectum Grabau. P. 630. Top and side view of the type. X 16. Lower Cambrian limestone bowlder, Sandy Cove, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,965. Figs. 9 a-f. Watsonella crosbyi Grabau. P. 632. 9 a. Right side of a specimen with the angulation somewhat rounded and the lines of growth arching far up on the dorsum. X 5. Pleasant Beach, Cohasset. 9 b. Right side of another individual of normal form. X 5. Sandy Cove, Cohasset. 9 c. Right side of a larger individual with broken beak, and faint radiating striae. X 5. Sandy Cove, Cohasset. 9 d. Left side of a fragment. X 5. 9 e. Apertural section of same, shown on a. broken surface below the centre. X 5. 9 f. View of dorsum near the aperture, showing keel. The specimen is compressed near the aperture, hence the dorsum is narrower below. X 10. Red Lower Cambrian limestone bowlders from Cohasset. Coll. Boston soc. nat. hist., cat. nos., 11,952, 11,951, 11,953, 11,954. Fig. 10 a-c. Eaphistoma attleborensis Shaler and Foerste. P. 633. 10 a. Base of a specimen showing dee,p and wide umbilicus, and a notch in the outer lip. X 5. 10 b. Side view of same, showing fiat base, compressed whorls, and pro- jecting aperture. X 5. Pleasant Beach, Cohasset. 10 c. Lateral view of a specimen with the whorls less compressed than in the preceding, and rather loosely superimposed. X 7. Sandy Cove, Cohasset. Red Lower Cambrian limestone bowlders from Cohasset. Coll. Boston soc. nat. hist., cat. nos., 11,955 and 11,957. Figs. 11 a, b. Straparollina remota Billings. P. 635. 11 a. Top view of a nearly complete specimen with the tip broken. X 4. PLATE 31 (continued). 11 b. Profile of same, showing flat base and rounded upper surface of the whorls. X 4. Ked Lower Cambrian limestone bowlder, Pleasant Beach, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,958. Figs. 12 a-c. Stenotheca abrupta Shaler and Foerste. P. 637. 12 a. Longitudinal section of a specimen probably belonging to this species. X 8. White Nahant limestone, Nahant, Mass. Coll. Peabody acad. sci. 12 b. Mould of the interior, showing ten corrugations. The basal corru- gation is broken away, and the apex is imperfect. X 8. 12 c. Interior of the shell of the same specimen, adhering to the rock matrix. The fragment shows the thickened margin on the concave ventral side ; also faint vertical striae. X 8. Red Lower Cambrian limestone bowlder, Bass Point, Nahant. Coll.. Boston soc. nat. hist., cat. no., 11,962. Fig. 13. Stenotheca curvirostra Shaler and Foerste. P. 638. Lateral view of a specimen witli strong corrugations over the dorsum, fading ventrally. X 12. Red Lower Cambrian limestone bowlder, Sandy Cove, Cohasset. CoU.- Boston soc. nat. hist, cat. no., 11,964. Fig. 14. Stenotheca pauper Billings. P. 639. A small specimen witli coarse corrugations across tlie dorsum, fading ventrally. X 12. The shell is broken along the dorsum. Red Lower Cambrian limestone bowlder, Sandy Cove, Cohasset. Coll. Boston soc. nat. hist, cat no., 11,963. Fig. 15. Stenotheca levis Walcott. P. 641. A specimen with strong rounded corrugations near tlie venter only. X 12. Red Lower Cambrian limestone bowlder, Pleasant Beach, Cohasset. ColL Boston soc. nat. hist, cat. no., 11,961. Occas. Papers, Boston Soc. Nai.Hisi..Vol.IV. Plate 31 '■fill Cambrian Fossils of the Boston Basin. HcHotpc Prirrtinfl Ci-j,Bonior., Occas. Papers, Vol. IV. PLATE 32. (Figures 1 a, b, 7 c, 8 e, f and 9 c, were drawn by Miss L. R. Martin. The remainder were drawn by Miss Elvira Wood.) Figs. 1 a, b. Hyoliihes shaleri Walcott. P. 642. 1 a. View of ventral side of a fragment, with cross-section. Nat. size- 1 b. Dorsal face of a fragment embedded in rock, showing lip. Nat. size. Paradoxides beds, Hayward Creek. Coll. W. W. Dodge. Figs. 2 a-h. HyoUthes pr biceps Billings. P. 643. 2 a. Dorsal side of a young individual showing forward-curving striae and lip. X 1^- 2 b. Fragment of a specimen showing dorsal side and cross-section. X If- 2 c. Cross-section referred to this species, obliquely cut. Enlarged. . 2 d, e. Sections of large specimens referred to this species. X — . 2 f-h. Various cross-sections, the variations in outline being due chiefly to obliquity of cut. Enlarged. White limestone, Nahant. Coll. Peabody acad. sci., cat. nos. 698, 695, 708, 709, 694, 700, 694 a. Figs. 3 a-c. HyoUthes excellens Billings. P. 646. 3 a Dorsal side of specimen described. Nat. size. 3 b. Cross-section of same. 3 c. Cross-section of a specimen of this species nat. size. White lime- stone, Nahant. Coll. Peabody acad. sci., cat. nos., 708 a, 708 b. Figs. 4 a, b. HyoUthes americanus Billings. P. 647. 4 a. Dorsal view of the specimen described ; the upper part is broken away. X 1^. 4 b. Cross-section of same. White limestone at Nahant. Coll. Peabody acad. sci., cat., no., 708. Figs. 5 a, b. HyoUthes searsi Grabau. P. 649. 5 a. Cross-section of the largest specimen known. Nat. size. White limestone. East Point, Nahant. Coll. Peabody acad. sci., cat. no. 710. 5 b. A group of cross-sectioixs. Nat. size. White limestone, north side of Nahant. Coll. Boston soc. nat. hist., cat. no., 11,967. PLATE 32 (continued). Fig 6. Hyolithes communis Billings. P. 651. Cross-section, showing normal form. White limestone, Nahant. Coll. Peabociy acad. sci., cat. no., 708 c. Figs. 7 a-c. Hyolithes impar Ford. P. 652. 7 a. A normal shell with oval cross-section. X 1^- 7 b. Fragment (restored), witli cross-section, x 2. 7 c. A cross-section. Enlarged. White limestone, Nahant. Coll. Peabody acad. sci., cat. nos., 700, 698 a, 702. Figs. 8 a-g. Orthotheca cylindrica Grabau. P. 654. 8 a, A fragment of a characteristic small specimen, showmg regular concentric striae and a constriction at one end. x 3. Red limestone bowlder, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,963. 8 b. Fragment of a larger specimen. X 1 ^ White limestone, Nahant. Coll. Peabody acad. sci., cat. no., 693 a. 8 c. Cross and longitudinal sections, the latter showing a septate apical portion. Nat. size. 8 d. Cross-section of two invaginated specimens. Enlarged. 8 e. Fragment with septate apical portion. X 3J. 8 f. Two invaginated specimens, the inner one showang concentric striae. X 3^. Associated with the preceding. 8 c-f. From red limestone bowlders from Cohasset and Bass Point. Coll. Boston soc. nat. hist., cat. nos., 11,963, 11,959, etc. 8 g. A small curved specimen which may be of this species. X 1^. White limestone, Nahant. Coll. Peabody acad. sci., cat. no., 693. Figs. 9 a-d. Orthotheca emmonsi (Ford). P. 655. 9 a. Dorsal view of a specimen showing faint concavity. Nat. size. 9 b. A characteristic ci'oss-section. 9 c. Cross-sections of invaginated shells. 9 d. Ventral view of a specimen with cross-section. White limestone Nahant. Coll. Peabody acad. sci., cat. nos., 703, 697, 703 a, 703 b. Fig. 10. Orthotheca (?) foerstei Grabau. P. 657. Cross-section of a group of invaginated shells referred to this species. Enlarged. Red limestone, Pleasant Beach, Cohasset. Coll. Boston soc. nat. hist., cat. no., 11,973. Fig. 11. Hyolithes (?) haywardensis Grabau. P. 653. Type specimen, showing several layers. X 1\. Paradoxides beds, Hayward Creek. Coll. Boston soc. nat. hist., cat. no., 11,974. PLATE 32 (Continued). Fig. 12. Hyolithellus micans Billings. P. 658. A fragment referred to this species. X 3. Ked limestone bowlder, Nahant. Coll. Boston soc. nat. hist., cat. no. 11,966. Fig. 13. Longitudinal section of a hyolithid (?), showing two septa. X 10. White limestone, Jeffrey's Ledge. Coll. Peabody acad. sci., cat. no., 714. Fig. 14. Longitudinal section of an undetermined shell. X 10. White limestone, Jeffrey's Ledge. Coll. Peabody acad. sci., cat. no., 712. Fig. 15. Longitudinal section of a Salterella (?). X 4. White limestone, Jeffrey's Ledge. Coll. Peabody acad. sci., cat. no., 712 a. Occas. Papers, Boston Soc.Nat.Hist„Vol.Iv: Plate 32. Cambrian Fossils of the Boston Basin. Occas. Papers, Vol. IV. PLATE 33. (All the figures were drawn by Miss Elvira Wood, Instructor in Palaeon- tology in the Massachusetts Institute of Technology.) Figs. 1 a-j. Olenellus {Holinia) broggeri Walcott. P. 662. la. A young head, crushed and broken, but showmg the characteristic features. X 2. 1 b. Fragment of cephalon. 1 c. A large fragmentary c phalon, much crushed. 1 d. Central portion of ce^ alon, laterally compressed. (From a gutta- percha cast.) 1 e. Part of lateral cheek, showing interocular spine. 1 f . Hypostoma without ihe lateral wings. 1 g. Ends of pleura, referred to this species. 1 h. A pleuron, showing the characteristic groove. 1 i. Left cheek of a yorng individual, showing interocular s^Dine. 1 j. Doublure, referred to this species. (Figs. 1 b to 1 j are natural size.) Fig. 2. Urotheca pervetus Matthew. P. 6-59. A fragment. Enlarged. Fig. 3. Microdisms sp. P. 671. A pygidium. X 4. Lower Cambrian Slates, Pearl St., North Weymouth. Coll. H. T. Burr, etc. Occas. Papers, Boston Soc. Nat,Hist.,Vo] IV. Plate 33, '-if Id ,1' i !!•/ jh ig Carnbrian Fossils of the Boston Basin. Hcliotype Rindjis Co.BosUm. Occas. Papers, Vol. IV. PLATE 34. (Figs. 8-10 were drawn by Miss L. K. Martin, the others by Miss Elvira Wood, Instructor in Palaeontology in the Massachusetts Institute of Tech- nology.) 1 a,'b. Olenellus sp. P. 66-5. la. A nearly perfect cephalon. Nat. size. lb. A fragment of the head of a young individual. Nat. size. 2 a, b. Olenellus {3Iesonacis) asaphoides (Emmons). P. 667. 2 a. A nearly perfect cephalon of a young individual. X 2. 2 b. Fragment of the cephalon of an older individual. Nat. size. 3 a, b. Obolella atlantica Walcott. P. 620. Two valves of different sizes, .showing external characteristics. X 3. 4-6. Metadoxides 7nagnijicus M^tthevf (?). P. 670. Fragments of genal spines, referred to this species with reservation. Nat. size. 7 a-c. Strenuella strenua (Billings). P. 672. 7 a. A small head exclusive of free cheeks. X 2. 7 b. A larger head. Nat. size. 7 c. Fragment of a cephalon. X 2. All the above are from the purplish Lower Cambrian slates of North Wey- mouth. Collected by Mr. Burr, with exception of specimen, fig. 2 a, which was collected by the writer. Fig. 8. Strenuella strenua (Billings) . P. 672. A fragmentary cephalon. X Ij- Eed slate pebble, Lower Cambrian, from Nahant. Coll. Boston soc. nat. hist., cat. no., 11,968. Fig. 9. Agraulos quadrangularis (Whitfield). P. 674. Cephalon, showing spine and palpebral lobes, also slight glabellar grooves. X 1^. Paradoxides beds Hayward Creek. Stud. pal. coll. Harv. univ., cat. no., 333. Fig. 10. Agraulos quadrangularis (Whitfield). P. 674. A large and broad cephalon, showing spine. X 1^. Paradoxides beds, Hayward Creek. Collected by Dr. D. F. Lincoln. Coll. Boston soc. nat. hist., cat. no., 11.975. Occas. Papers, Boston Soc. Nai.Hisi., Vol .IV. Plate Cambrian Fossils of the Boston Basin. Heliotype hintiij Co.B«3lon. Occas. Papers, Vol. IV. PLATE 35. (Keproduced from plate 7, Bull. 10, U. S. geol. surv., by courtesy of the Director.) I'ig. 1. Agraulos quadrangular is (Whitfield). P. 674. Head, exclusive of the free cheeks. Natural size. Paradoxides beds, Hayward Creek. Coll. mus. comp. zool. Fig. 2. Ptychoparia rogersi Walcott. P. 679. The type specimen. Natural size. Paradoxides beds, Hayward Creek. Coll. mus. comp. zool. Fig. 3. Paradoxides harlani Green. P. 681. Large free cheek flattened by compression. Paradoxides beds, Hayward Creek. Coll. Boston soc. nat. hist., cat. no., 1. Fig. 4, 4 a-c. Hyolithes shaleri "Walcott. P. 642. 4, 4 a, b. Dorsal, ventral and lateral views of the type specimen. 4 c. Cross-section of same. Paradoxides beds, Hayward Creek. Coll. mus. comp. zool. Occas. Papers, Boston Soc. Nat. Ilisl., Vol. IV. Half 35. CAMBRIAN FOSSILS OF THE BOSTON BASIN (Paradoxides Beds.) Occas. Papers, Vol. IV. PLATE 36. Paradoxides harlani Green. P. 681. The type specimen. Two thirds natural size. Presumably from the Paradoxides beds of Hay ward Creek. Coll. Bos- ton soc. nat. hist, cat. no., 6264. Drawn by Miss L. E. Martin. ^ Occas. Papers, Vol. IV. PLATE 37. Paradoxides harlani Green. P. 681. A small nearly complete specimen of the broad type. Natural size. Paradoxides beds, Hayward Creek. Coll. Bridgewater Normal School. Drawn by Miss L. R. Martm. Occas. Papers, Boston Soc. Nai.Hisl.,Vol .IV. lUr '•/ Paradoxides l^arlani Green Natural size Heliolvpe Co,Boston. Occas. Papers, Vol. IT. PLATE 38. (Eeproduced from Plate 8 of Bull. 10, IT. S. geol. siirv., by courtesy of the Dii'ector.) ■ Figs. 1 to 1 e. Paradozides harlani Green. P. 681. 1, 1 a. Two medium-sized heads, showing variations in the frontal border. Natural size. 1 b, c. Broad and elongate forms of the pygidium. 1 d. View of the four posterior thoracic segments and pygidium ; the latter is crowded up beneath the segments. 1 e. Large hypostoma attached to the frontal doublure of the head. Paradoxides beds, Hayward Creek, Braintree, Mass. Coll. mus. comp. zool. Occas. Papers, Boston Soc. Nat. Hist, Vol. IV. 1 Hlalc j«. lii le .7 Id PARADOXIDES HARLAN! GREEN. Occas. Papers, Yol. IV. PLATE 39. (Eeproduced from plate 9, Bull. 10, U. S. geol. surv., by coartesy of the Director.) Fig. 1. Paradoxides harlani Green. P. 681. A large individual, preserving the body and parts of the head. The light-colored portions are restored ; and the pygidiuni, vrhich in the specimen is pushed a little out of position, is replaced. Nat. size. Paradoxides beds, Hayward Creek, Braintree, Mass. Coll. Boston soc. nat. hist., cat. no., 2754. Occas. Papers, Boston Soc, Nat. Hist., Vol. IV. Plate 39. PARADOXIDES HARLANl GREEN- Table of tiie Ilaritasket ;^itrstn. rfi»an5.t©, box^cterliig tm& iw%dorlvlng the b@«^M©d a^cks. ;'lj*st COKglor^er^t© . teasel. r-nni-Xnti; on gi^Atiite ..,,.., E5*S0 fe©t» lioekj? Ifecfeir GrajTsitte IPlatemj., Cliff >M«tosts,, ©-■ s First Melaphyr, ct»^;>act and .!aspei*:y ««..«.•..'••. 2&-40 feet* rOcky ileck, last ami l/©st Forpfi3n''lt© llille ar^ CllfT l-latoau, i@cond €:as3t sM i;tost Fot^hr^rrSt© iftlls, Oreat Uili, etc* i-'Oi*^srlfee «.•..-..•••«..«••»«.•..*.. fjO-lOO feot. B«st 5«i?i V^est .l?oii>h5^1t© liiils, Blaefe |i:oeiE, citc« f55lrd Gor.glorK3rat< .,,.♦,..,,•. » • 60-10© foet, CemglOKKEfBte ..^xi -^^iiU »^i*»t*5au, Fol;30&>*a IsIjj*-'.^, ,< v^uirid ili,ll sBd Oi»e©n villi* •>o©ond M©laph;ji», gs»©