3471 
 
i£x ICibrtB 
 
 SEYMOUR DURST 
 
 When you leave, please leave this book 
 
 Because it has been said 
 "£ver'thincj comes t' him who waits 
 
 Except a loaned book." 
 
 Avery Architectural and Fine Arts Library 
 Gift of Seymour B. Durst Old York Library 
 
Digitized by the Internet Archive 
 in 2014 
 
 http://archive.org/details/onmetamorphicstrOOmerr 
 
[Prom the American Journal of Science, Vol. XXXIX, May, 1890.] 
 
 ON THE 
 
 METAMORPHIC STRATA OF SOUTHEASTERN 
 NEW YORK. 
 
 By Frederick J. H. Merrill. 
 
 Of the strata comprised in the metamorphic terrane of 
 southeastern New York there are two principal divisions. 
 
 First, in Rockland, Orange, Putnam and Dutchess counties, 
 the granitoid gneisses and granulites of the Highlands, which, 
 with their continuation in New Jersey, can be shown strati- 
 graphically to be pre-Cambrian, since they underlie unconform- 
 able in southern Dutchess County, N. Y., and at several points 
 in New Jersey, a basal Paleozoic quartzite of Potsdam or 
 Lower Cambrian age. 
 
 Secondly, in Westchester and New York Counties the gneis- 
 soid quartzites and arkoses overlain by crystalline limestones and 
 mica schists, which border the rocks of the first division on the 
 southeast and have been regarded as altered Paleozoic rocks of 
 Ordovician or Cambrian age by Professors W. W. Mather and 
 J. D. Dana. 
 
 The metamorphic rocks of the New Jersey Highlands were 
 first definitely classified by Dr. N. L. Britton as a result of his 
 studies in 1885 and 1886.* In this terrane he identified three 
 groups : first ; a Massive Group, devoid of bedding planes, 
 
 * Annual Report G-eol. Survey of New Jersey, 1885-1886. 
 
384: Merrill — Metamorphic Strata of 8. K New York. 
 
 described by him as being the oldest formation exposed in that 
 state, and almost entirely composed of " hornblendic granite" : 
 second, an Iron-bearing Group, resting conformably on the pre- 
 ceding, well stratified, and comprising bedded granulites and 
 magnetite deposits, above the latter being crystalline limestones 
 containing magnesian silicates in various stages of alteration : 
 third and uppermost, a " Schistose Group," composed of mica- 
 ceous gneisses, of mica schists sometimes containing graphite, 
 and hornblendic and pyroxenic schists of varied composition. 
 The basal Paleozoic quartzite has not yet been found to rest 
 on strata of the third group, and the relative age of the latter 
 remains in doubt. 
 
 In the study of the metamorphic rocks of the New Jersey 
 Highlands and in a preliminary examination of the Hudson 
 River section, the writer was associated with Dr. Britton and 
 has since made a more extended study of the latter. The facts 
 there observed corroborate the general results of Dr. Britton's 
 stratigraphical work in New Jersey. 
 
 The basal member of the pre-Cambrian terrane of southeast- 
 ern New York and New Jersey, is a granitoid hornblende- 
 gneiss consisting chiefly of quartz, orthoclase, plagioclase and 
 hornblende. Magnetite and zircon sometimes occur as acces- 
 sories. As stated by Dr. Britton the rock is devoid of bedding 
 planes and has only a parallel arrangement of its minerals to 
 denote macroscopically its sedimentary origin. Its microscopic 
 structure, however, also suggests its detrital and metamorphic 
 character. 
 
 This rock is not a granite although usually called by that 
 name. In addition to the parallel arrangement of its minerals 
 which would differentiate it from the normal granites, its struc- 
 ture shows that it never attained that condition of molecular 
 freedom which would be necessary to enable its mineral com- 
 ponents to assume a crystalline form. The quartzes and feld- 
 spars occur in irregular masses interlocking at their margins 
 and in no case show crystalline boundaries. The hornblende, 
 which is allotriomorphie and tills the irregular interstices be- 
 tween the other mineral particles, has probably been developed 
 in the process of metamorphism from aluminous mud contain- 
 ing iron, lime and magnesia. As in rocks which have been 
 developed from the plastic or fluid state the magnesia-iron 
 silicates according to Rosenbusch, are idiomorphic with respect 
 to the feldspars and the latter are idiom orphic with respect to 
 the quartzes, the structure of the rock in question seems to 
 demonstrate that its component minerals are not developed 
 from a magma but are merely the fragmental particles of a sedi- 
 mentary rock and that during metamorphism only a slight de- 
 gree of molecular freedom was attained by them. 
 
Merrill — Metamorphic Strata of S. K New York. 385 
 
 The thickness of the basal member is at present indeter- 
 minate, as nothing has been found beneath it, and in the Hud- 
 son River section it is seen to descend below tide-level. The 
 maximum thickness exposed is in Breakneck Mountain, of 
 which the summit is said to be 1787 feet above tide. What- 
 ever be the true height of the mountain, its central mass from 
 tide-level to the summit consists of hornblende-gneiss, and 
 whether it be anticlinal or monoclinal in structure, an equiv- 
 alent thickness of this rock is exposed. 
 
 The second member or Iron-bearing Group, as Dr. Britton 
 has stated, consists essentially of granulites, this name being 
 used by him and the present writer to describe subcrystalline, 
 well stratified metamorphic rocks composed principally of de- 
 trital quartz and feldspar. They are highly metamorphosed 
 arkoses, and differ essentially from the under-lying formation 
 in the absence of magnesia-iron silicates such as hornblende and 
 biotite. The comparative absence of silicate minerals contain- 
 ing iron is indeed the conspicuous feature of this member taken 
 as a whole. In some places, as at Garrisons, a pale yellowish 
 mica is present, giving the rock the semblance of a granite, 
 under which name it is quarried, and in close proximity to some 
 of the magnetite beds in New Jersey, biotite and hornblende 
 are quite abundant. This formation is at least 500 feet thick. 
 
 The beds of magnetite appear to occur at various levels in 
 the second group and offer little evidence of their origin. If, 
 however, it were known that the rock which furnished the de- 
 tritus out of which these granulites were formed, contained 
 magnetite distributed through it in any appreciable quantity 
 there would be good reason for the conclusion that the magne- 
 tite beds of this horizon originated as beds of magnetic sand 
 concentrated by wave action. There is no indication that they 
 originated as bog ores. The} r are simply lenticular beds en- 
 closed by a stratified metamorphic rock formed of quartz and 
 feldspar sand and there is no adjoining rock of which the com- 
 position would suggest that it is a metamorphosed soil. There 
 is, besides, no indication of a former land surface on which 
 these beds might have been formed. The only suggestion of 
 organic agency is in the apatite which occurs with the iron ores 
 at some of the mines and this, if it were organic, could have 
 been derived from marine organisms as well as from terres- 
 trial. It is not, however, by any means certain that this apatite 
 is organic, since the mineral is a frequent accessory in various 
 eruptive rocks. 
 
 The strata of Dr. Britton's schistose group and the lime- 
 stones which, according to him, occur near the top of the sec- 
 ond member are found near the Hudson River, in Orange 
 County at Fort Montgomery, and in Putnam County between 
 
386 Merrill* Metamorphic Strata of 8. E. New York. 
 
 Highlands and Garrisons. It has not yet been determined 
 whether they should be regarded as part of the system to which 
 the two lowest groups or members belong, nor have their rela- 
 tions been determined with respect to the Manhattan Group 
 (hereafter defined). As before stated, no Ordovician or Cam- 
 brian Rocks have been found to rest upon them and hence 
 their pre-Cambrian age cannot, as yet, be predicated with cer- 
 tainty. 
 
 The thickness of the pre-Cambrian rocks in the Hudson River 
 Valley may, therefore, be stated as between 2300 and 2800 
 feet. As to their age, it is difficult to predicate anything on 
 account of their isolated position. The " upper member," or 
 granulitic group may or may not be the- equivalent of the 
 Huronian. There is nothing but its unconformable position 
 immediately below the Cambrian quartzite to suggest such 
 equivalence, and as it seems to be conformable to the " basal 
 member" that would have to be included in the same horizon. 
 
 The magnetite beds of this region have been compared by 
 Dr. Britton to those of the Grenville series in Canada, and 
 they may be equivalent but until the Laurentian of Eastern 
 Canada has been studied and classified with the aid of modern 
 methods of research any attempt at correlation will be unsatis- 
 factory. 
 
 The stratigraphy of the Highland region as displayed in the 
 section along the Hudson River is very simple. A small num- 
 ber of anticlinal ridges, 900 to 1700 feet in height, with a 
 northeasterly trend, are intersected by the Hudson River valley. 
 Along the lines of these the basal member of the Archaean is 
 exposed and resting on their flanks and in the synclinal troughs 
 the rocks of the Iron-bearing Group appear. 
 
 The most northern axis is that of the Fishkill range which is 
 probably a continuation of the Wawayanda Mountain axis of 
 Northern New Jersey, along which line in Orange Co. N. Y. 
 occur a number of isolated hills of gneissic rock known as 
 Sugar Loaf Mountain, Goose-Pond Mountain, Peddler Hill, 
 Rainer Hill, Mosquito Hill, Round Hill and Woodcock Hill. 
 
 Second is that of Storm King and Breakneck Mountains, 
 closely related orographically to the axis of Crow Nest Moun- 
 tain and Bull Hill or, as the latter is sometimes called, Mt. 
 Taurus. There is probably a fault line between these two axes 
 and nearly parallel to them, but the structural details have not 
 been accurately determined. A fourth axis is that which 
 crosses the Hudson at AVest Point. A group of axes crosses 
 the Hudson along the lines of Fort Hill, Sugar Loaf Moun- 
 tain, Anthony's Nose and Bear Hill and finally the anticlinal 
 of Manitou Mountain and the Dunderberg closes the succession. 
 These folds generally pitch steeply to the S.W. In this re- 
 
Merrill — Metamorphic Strata of 8. E. New York. 387 
 
 spect the most noticeable are those of Anthony's Nose and 
 Sugar Loaf, the latter of which exhibits this peculiarity by its 
 profile to the most careless observer. 
 
 On the southeastern flank of Dunderberg and Manitou 
 Mountains the stratified granulites again appear and these are 
 succeeded after a drift-covered interval near Peekskill and 
 Jones Pt. by slightly metamorphosed sandstones, limestones 
 and slate, presumably of Paleozoic age, the relation of which 
 to the sub-crystalline rocks of Westchester and New York 
 Counties is still an interesting problem. 
 
 Younger Hocks. — The metamorphic strata of New York 
 and Westchester Counties have long attracted the attention of 
 geologists and several attempts have been made to solve the 
 problems of their age and history. The extended researches 
 of Prof. Jas. D. Dana on the relations of the limestone belts 
 in the vicinity of Manhattan Island have furnished a clue to 
 the stratigraphy, and, after a careful study of a portion of the 
 terrane to which those limestones belong, the writer is enabled 
 to announce the following general results. As Prof. Dana has 
 noted (Am. Jour. Sci., Ill, vol. xxi, p. 439), the beds underlying 
 the limestones of New York County are highly quartzose, 
 while those overlying them are chiefly micaceous. Throughout 
 Westchester County south of the latitude of Sing Sing, an 
 area of about one hundred square miles, the writer has found 
 this lithological difference to prevail. The exact relation of 
 the lower beds to the granitoid gneisses and granulites of the 
 Highlands of Rockland, Orange and Putnam Counties has not 
 yet been determined by the writer, but his investigations have 
 satisfied him that the former, with the exception of the lime- 
 stones, are distinctly detrital rocks in which are preserved the 
 fragmental character of the quartz and feldspar which they 
 contain. The mica, chiefly biotite, is of metamorphic origin, 
 having been developed from aluminous mud rich in potash, 
 iron and magnesia. 
 
 As the limestones of the region under consideration contain 
 no organic traces, so far as we know, there is no direct clue to 
 their origin, but in the absence of evidence to the contrary we 
 may believe that the carbonate of lime was separated by organic 
 agencies from the sea water which held it in solution. The 
 well known chemical theory of Dr. Hunt accounts very satis- 
 factorily for the presence of bicarbonate of lime in sea water, 
 but it does not account so well for the separation of the car- 
 bonate from solution, and as this separation is effected very 
 extensively at the present time by various marine organisms 
 and as the cases of chemical separation of carbonate of lime 
 from solution in sea water are very few, we are justified in be- 
 lieving that the former process was in operation as far back in 
 
388 Merrill — Metamorphic Strata of S. F. New York. 
 
 the earth's history as the sea water was inhabited by creatures 
 possessing skeletons of carbonate of lime. The source of the 
 magnesia so abundant in the limestone is yet undetermined. 
 
 The lowest stratum yet recognized among the Westchester 
 County rocks is a reddish gneiss which forms the central mass 
 of some of the hills in Yonkers where it is well exposed, and 
 consists of small grains of detrital quartz with fragments of 
 reddish orthoclase and a few crystals of biotite which have 
 developed during the process of metamorphism. From its 
 macroscopical characters the rock would be called a gneiss. 
 From its microscopic structure it would be called a metamor- 
 phosed sandstone or arkose. As the name quartzite gneiss has 
 been applied in Germany to rocks of analogous structure, it is 
 proposed to designate the rock in question as an arkose gneiss. 
 It may be objected that all gneisses were quartzites or arkoses 
 previous to their metamorphism, but there is an evident neces- 
 sity for the use of some descriptive term which will convey 
 the idea of comparatively unaltered detrital structure and 
 differentiate such a rock as this from the pre-Cambrian gran- 
 itoid gneisses of Putnam County. 
 
 The thickness of the red gneiss referred to has not been 
 determined, as no section has been found which shows its rela- 
 tions to the formation beneath it, but it is believed to rest 
 upon the stratified granulites which form the second or upper 
 member of the pre-Cambrian formation. Since the best ex- 
 posures of the red gneiss have been found within the limits of 
 Yonkers and its relations to the overlying stratum are well 
 shown within the limits of that city, it is proposed to call it 
 the Yonkers gneiss. Outcrops of this rock are frequent along 
 the shores of the Hudson and good exposures may be found at 
 Hastings, on the property of Dr. Draper ; on the river near 
 the southern border of Tarrytown ; between Scarborough and 
 Sing Sing stations near the railroad ; along the southeast shore 
 of the mouth of Croton Bay on the property of Orlando B. 
 Potter, Esq. ; and a little south of Oscawana Station, on Osca- 
 wana Island : it also probably occurs in the ridge which sepa- 
 rates Annsville Cove from the valley of Peekskill village. 
 The best exposures are in Yonkers, on and near Jerome 
 avenue, a little north of the Xew York City line. 
 
 With regard to the variations in the composition and struc- 
 ture of the Yonkers gneiss, the most important are an increase 
 in the proportion and size of the feldspar fragments in ap- 
 proaching the Lauren tian Highlands. 
 
 Overlying the reddish Yonkers gneiss and beneath the crys- 
 talline limestone is a stratum of thinly bedded gray quartzite 
 gneiss. This contains but little feldspar and its component 
 beds vary in composition from almost pure quartz to a mixture 
 
Merrill — Metamorpkic Strata of S. K JVeiv York 389 
 
 of quartz and biotite or hornblende. Occasionally, layers of 
 pure biotite schist, an inch or two in thickness, are intercalated 
 with white, coarsely granular quartzite. This rock forms the 
 summit and eastern slope of the ridge which separates the 
 Sawmill River valley from that of the Hudson and also occurs 
 in that on the east side of the Hudson between Yonkers and 
 Spuyten Duyvil. It forms, as well, the anticlinal ridge of 
 Fordham Heights which borders the east shore of the Harlem 
 River and of which the southern extremity forms the long 
 narrow hill by the northern end of Seventh avenue and which 
 separates the latter from the Boulevard. 
 
 In general terms this quartzite gneiss is the gray rock used 
 for a building stone in southern Westchester County. The 
 Yonkers gneiss is also used for building, but not so extensively 
 and is characterized by its reddish color. ISo section has yet 
 been found which shows accurately the thickness of this gray 
 gneiss. It is at least two hundred feet thick and, in many ex- 
 posures, has an aggregate thickness of half a mile, but in these 
 it is apparently repeated many times by folding. As in most 
 cases the folds are isoclinal and their arches have been removed 
 by erosion, there is little stratigraphical evidence of the thick- 
 ness of the stratum thus folded. 
 
 Since this rock is well exposed and its stratigraphical rela- 
 tions are well shown in the Fordham Heights it is proposed to 
 call it the Fordham gneiss. 
 
 The Fordham gneiss varies widely from the normal type, in 
 places, through the presence of hornblende and garnet and an 
 increase in the amount of feldspar and mica, but the localities 
 of variation are comparatively few. 
 
 Between the Fordham gneiss and the crystalline limestone 
 of the Hastings Quarry and in Yonkers at the south end of 
 the railroad trestle near the Lowerre Race Track, a stratum of 
 thinly bedded quartzite from five to ten feet thick is found. 
 It seems probable that this is of wide distribution but on ac- 
 count of its slight thickness it is very rarely preserved when 
 the limestone which rested upon it has been removed by 
 erosion. 
 
 At Tuckahoe this quartzose stratum next beneath the lime- 
 stone contains numerous flakes of biotite. 
 
 The position and stratigraphy of the limestone areas of 
 Westchester County have been carefully studied by Professor 
 Dana. My only contribution to the geology of these beds has 
 been to determine the relations and character of the associated 
 rocks, to note a few unrecorded outcrops and in a few cases to 
 extend the limits of those previously known. 
 
 Professor Dana has estimated the thickness of the bed oc- 
 curring in Tremont and the Harlem River valley at from six 
 
390 Merrill— Metamorpkie Strata of & K New York. 
 
 hundred to seven hundred and fifty feet. The measurements 
 of the writer would indicate that the thickness varies from six 
 hundred to eight hundred feet, it being apparently greater on 
 New York Island than in Morrisania. The eastern bed at 
 Tuckahoe is but one hundred and fifty feet thick. For this 
 rock I propose the name of Inwood limestone, from the locality 
 on New York Island in the vicinity of which it is well exposed. 
 
 The rocks which overlie the limestone are highly schistose 
 and consist largely of mica with a small proportion of quartz 
 and usually little or no feldspar. Garnet, staurolite, fibrolite 
 and cyanite are the chief accessories. There are some beds of 
 gneiss among them, but these are very small and the studies of 
 the writer enable him to state positively that mica preponder- 
 ates in the rocks above the limestone beds. Xo sections have 
 yet been found which would warrant an expression of opinion 
 as to the exact thickness of these schists, but it probably ex- 
 ceeds one thousand feet. 
 
 The mica schist formation which belongs above the lime- 
 stone is of very limited extent in Westchester County. A 
 synclinal ridge of this rock extends from Park Hill, in Yon- 
 kers, northward along the east bank of the Saw Mill River 
 and has been traced to Elmsford. 
 
 North of Croton landing the mica schist containing garnet 
 and staurolite extends along the bank of the Hudson for about 
 a mile and east to the norite area of the Cortlandt Series. 
 Xear Crugers the schists have been described by Professor 
 Dana 
 
 Between the Bronx River and Long Island Sound, in South- 
 ern Westchester County, there is a considerable extent of 
 mica schist but its limits are not determined. In Eastchester 
 village, on the west shore of Eastchester Creek the rock is 
 a gneissoid quartzite. At New Rochelle the rock along the 
 shore of the Sound probably belongs below the limestone.. 
 The same rock, essentially a gneissoid quartzite, occurs on the 
 shore of Mamaroneck Harbor, while Milton Point in Rye 
 township seems to be composed of the mica schists. On the 
 Hudson River shore, in general, the limestone areas are suc- 
 ceeded to the north by mica schists and to the south by the 
 arkose gneisses. 
 
 As these uppermost beds are well exposed on Manhattan 
 Island of which they constitute the principal rock formation 
 they may well be called the Manhattan schists. 
 
 The name Manhattan Group was proposed in 1868 by R. P. 
 Stevens, Esq., to include the rocks of New York Island and it 
 seems proper that it should, for the present, be retained, in- 
 cluding in it. with the Manhattan schists, the Inwood limestone 
 and the Fordham gneiss, the Yonkers gneiss which though not 
 
Merrill — Metamorphic Strata of 8. E. New York. 391 
 
 found on Manhattan Island, is evidently a part of the same 
 littoral deposit to which the Fordham gneiss belongs. The 
 Manhattan schists are the only beds now to be found on New 
 York Island with the exception of the limestone areas de- 
 scribed and mapped by Professor Dana and the small area of 
 Fordham gneiss at the north end of Seventh avenue. 
 
 If in time this group be correlated with some other which 
 has been previously described, the names here suggested may 
 be unnecessary, but until the question is indisputably settled 
 they are of use in referring to the formation and its sub- 
 divisions. 
 
 Intercalated with the Manhattan schists and also with the 
 beds of the Fordham gneiss we find at a great number of 
 localities, hornblendic and augitic strata of limited thickness, 
 usually only a few feet. In composition these rocks resemble 
 diorites and diabases, and in structure they are granular, but 
 their present well stratified condition renders it difficult to say 
 whether they are originally eruptive rocks or not. Whatever 
 their origin they are now metamorphic rocks and as such may 
 be called amphibolites and pyroxenites according to the ter- 
 minology of Kalkowsky. It is probable that to rocks of this 
 character we are indebted for some of our serpentines, notably 
 that of 60th street near 10th avenue, New York City, for as 
 originally suggested by Dana and lately demonstrated by 
 Gratacap it is derived from the alteration of amphibole, and 
 on 61st street near 11th avenue, a bed of amphibolite occurs, 
 of which the line of strike passes through the well known and 
 interesting serpentine above mentioned and which lies about 
 250 feet southeast. 
 
 Age of the Manhattan Group. — It is not yet in the power 
 of the writer to contribute any positive information on this 
 important question He has not yet found any decisive evi 
 dence of the age of the rocks in question. All the suggestive 
 evidence, however, favors the view taken by Professor W. W. 
 Mather and subsequently elaborated by Professor J. D. Dana, 
 viz : that the rocks of the Manhattan Group are the metamor- 
 phosed equivalents of the Paleozoic beds of Southern Dutchess 
 County. 
 
 After a careful study of the stratigraphy in the vicinity of 
 Peekskill which seems to be the index of this geological 
 chapter, and at other points along the northern margin of the 
 Manhattan terrane, the writer concludes that if this group is 
 pre-Cambrihn, its identity as such has been obscured by a series 
 of stratigraphic vicissitudes so complicated that it is beyond his 
 powers, at present, to conceive them. 
 
392 Mrr rill — Metamorphu: St nit a of 8. K New York. 
 
 A fact which may be of much significance, is that, the Pots- 
 dam or Lower Cambrian sandstone of Southern Dutchess 
 County lies unconformably on the second or granulitic member 
 of the pre-Cambrian formation, and it is upon the same second 
 member that the basal beds of the Manhattan Group rest. 
 No unconformability has yet been found between the Manhat • 
 tan Group and the underlying pre-Cambrian beds, and it is 
 chiefly this lack of positive evidence that leaves the writer in 
 doubt as to the geological equivalence of the former. Of 
 equal significance, however, is the lack of unconformability 
 between the Lower Silurian strata of Peekskill Hollow, Tomp- 
 kins Cove, and Verplank's Point, which are but partially 
 metamorphosed, and the metamorphic beds of the Manhattan 
 Group which adjoins them. 
 
 The crystalline limestones of the Manhattan Group are, as 
 already stated, highly magnesian and in this respect they cor- 
 respond in composition to the Calciferous limestones of New 
 Jersey which according to Professor Geo. H. Cook (Geology 
 of New Jersey 1868) contain from seventeen to twenty per 
 cent of magnesia. 
 
 In this abstract it has not been possible to give in detail the 
 evidence upon which the writer's conclusions are based. A 
 complete discussion of the evidence must therefore be reserved 
 for future publication.