Q E 12.0 UC-NRLF B M 512 fl?D x r LIBRARY UNIVERSITY OF CALIFORNIA. GIFT OF ^ \ Class THE GEOLOGY OF THE FOX ISLANDS, MAINE, A Contribution to the Study of Old Volcanics. A DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY GEORGE OTIS SMITH. THE GEOLOGY OF THE FOX ISLANDS, MAINE, A Contribution to the Study of Old Volcanics. A DISSERTATION SUBMITTED TO THE BOARD OF UNIVERSITY STUDIES OF THE JOHNS HOPKINS UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY GEORGE OTIS SMITH OF THE UNIVERSITY FO PUBLISHED BY THE AUTHOR, SKOWJ4EGAN, MAINE. Q PRESS OF J. 0. SMITH & Co. SKOWHEGAN, ME. CONTENTS. PAGE INTRODUCTORY: Geographical, ' Topographical, 7 Geological, Nature of Problem, Areal Distribution of Rock-types, 11 NORTH HAVEN GREENSTONE COMPLEX: Distribution, Description of Rock-types, Diabases, Amyg-daloids, Tuffs, - IS Columnar Structure, - 16 Age, 19 NIAGARA SEDIMENTS: Distribution, Southern Harbor Section, 20 Fauna, Conditions of Deposition, CALDERWOOD'S NECK SCHISTS: Description and Distribution, Relations, Age, - - - 29 THOROUGHFARE VOLCANICS: Distribution, Description of Rock-types, Andesites, Porphyries, Amygdaloids, - 36 CONTENTS. Pyroclastics, 37 Sequence, - 40 Ag-e, 42 Conditions of Volcanic Activity, 43 VINAL HAVEN: ACID VOLCANICS: Distribution, - 46 Description of Rock-types, 46 Aporhyolites, 46 Flow-breccias, 51 Tuffs, ..* 51 Sequence, - 52 Conditions of Volcanic Activity, - 52 DIKES: Description, - 56 Distribution, - 57 DIABASIC AND GRANITIC INTRUSIVES: Areal Distribution, 59 Granite, 60 Diabase and Diorite, - 61 Relative Ag-e, 63 Relations to Dikes, 64 Time and Conditions of Intrusion, 64 ALTERATION : Character of Alteration, 66 Structural and Mineralog-ical, - 66 Devitrification, 69 Nomenclature of Altered Ig-neous Rocks, 70 GEOLOGIC HISTORY: Acadian Troug-h, 72 Pre-Niag-ara Volcanic Activity, 72 Niag-ara Sedimentation, 73 Post-Niag-ara Volcanic Activity, ,- 74 Devonian Intrusions, 75 PLATES. I. Photo-micrographs. II. Geolog-ical Map. PREFACE. The following" geological study of the Fox Islands is presented as a contribution to th^ solution of the problem of the identifica- tion and interpretation of ancient volcanic rocks. The writer owes his interest in this subject to the late Professor Georg-e H. Williams, whose enthusiasm in petrographical research was so in- spiring* to his students. The Fox Islands area was visited and studied in 1893 by Dr. W. S. Bayley of Colby University, and the writer wishes to acknowl- edge the kindness of Dr. Bayley in granting- him the privilege of studying" the rocks collected at that time, as well as the accom- panying- full suite of thin-sections. The area was studied and the different series mapped by the writer in the spring- of 1895, and a subsequent visit was made in the spring- of 1896. In the preparation of this paper, assistance has been received from Prof. Clark, Dr. Mathews, and Messrs. Gilbert and Willis of the Geological Department of this University, and Dr. Bayley of Colby University, to all of whom the writer desires to express his indebtedness. He also wishes to acknowledge the courtesy of Pro- fessors Remsen and Morse of the Chemical Department, under whose direction rock-analyses were kindly made by Messrs. K. W. Mag-ruder and W. A. Jones. INTRODUCTORY. GEOGRAPHICAL. The rock-bound coast of Maine offers special attractions to the geologist. It is characterized by many deep in- dentations, the largest being Penobscot Bay, in which are situated the Fox Islands. These islands, a geological study of which is here presented, are slightly over eight miles due east from the city of Rockland and about twenty-five southwest from Mount Desert. Like the latter island, the Fox Islands lie just within the general coast-line which has a southwest and northeast trend. Unlike Mount Desert, however, they are at some distance from the main- land, being nearly in the middle of the bay, which is twenty-eight miles wide, from Owls Head to Naskeag Point. To the southeast and southwest, these islands are exposed to the open Atlantic, but to the south about ten miles distant is an outer fringe of small islands, the Matinicus group. North Haven and Vinal Haven, the principal members of the Fox Islands, are among the larger of the many rocky islands on the Maine coast. Surrounding these are many smaller islands, and the group extends ten and three-fourths miles from north to south and eight and one-fourth from east to west 1 . TOPOGRAPHICAL. Topographically, these islands are well di- versified. Although at no point is there any great elevation, yet many hills rise from one to two hundred feet above the sea-level, and afford pleasing variety of relief. The most striking topo- graphical feature, however, is the extremely irregular coast-line, which is to be regarded simply as an unmodified contour of the rocky relief. Peninsulas, necks and points, bays, harbors and coves are the elements, which constitute the greatest charm of the Fox Islands. Although one might sail around the islands in a dis- i. Maps of the Fox Islands are sheets 309 and 310, East and West Penobscot Bay scale i 140,000, and 31 la, Fox Islands Thoroughfare, I : 20,000, of the U. S. Coast and Qeodetic Survey. Sheets 309 and 310 were used to furnish the base of the geological map herewith presented, Plate II. 8 GEOLOGY OF THE FOX ISLANDS. tance of thirty miles, the coast-line of the two larger islands alone measures nearly one hundred and fifty miles. Indeed, so deeply indented are the shores, and so low the divides, that a submerg-- ence of less than twenty feet would divide North Haven and Vinal Haven into scores of smaller islands. In connection with the present topography it is of interest to note its origin, as well as the [modifications, which are at present being- made in it. Both the surface features of these islands and the deeply incised coast-line have had their origin in the erosion processes of the past. Penobscot Bay is one of the many drowned river valleys on the New England coast; and the islands which now comprise so large a part of the area between the shores of the Bay, simply represent the tops of the hills in the old Penobscot valley. Before the depression of the land, the time of which Prof. Shaler, in his study of the geology of Mount Desert, has fixed as about the close of the Glacial period, what we may term the Fox Hills rose within the valley of the Penobscot just as the more prominent Camden Hills still rise on the western side of the Bay. The Penobscot, which now reaches tide water forty miles to the north, then had its mouth perhaps as many miles to the south. The Thoroughfare, now separating- the two islands, was then a transverse valley extending- down on either side to the floor of the larg-er valley and occupied by side streams. In a similar way, the harbors and coves mark the drainag-e lines of the old Fox Hills. Such topography naturally expresses the different degrees of rock resistance to erosion, and the best example of this is the cutting- of the softer sedimentary rocks on North Haven almost down to what is now sea-level, while the dense volcanic rocks immediately to the south rise abruptly as Ames Knob. This topography, then, was the product of atmospheric ag-encies, and differs in no essential re- spect from that which is being- produced to-day under similar con- ditions. During- the ice invasion of Glacial time, some modification of the older topography may have resulted. It is to be noted, how- ever, that the characteristic features of Mount Desert, interpreted by Shaler and Davis as due to glacial erosion, are not represented on these islands. The more prominent inlets, especially the Thoroug-hfare, might be compared to the remarkable transverse INTRODUCTORY. 9 valleys of the Mount Desert rang-e, except that the former have a g-eneral trend more nearly east-west, while the glacial striae show a north-south direction. The mantle of g-lacial drift, consisting- of clay and sand, with some gravel, is comparatively thin on the Fox Islands, yet drain- ag-e may have been embarassed somewhat by local deposits. Fresh Pond on North Haven, originated, doubtless, in the dam- ming-, at that point, of a stream with its several tributaries. In the main, however, the pre-Glacial topography was only slig-htly modified by the ice-cap which covered the Fox Hills. Just as the topography of the present has had its origin in the past, so that of the future is being- determined by the ag-encies and processes of to-day. The complex succession of inlets and pro- montories sig-nifies a young- coast-line. With lapse of time, as maturity and old ag-e are attained, such a coast-line would become simpler throug-h the action of the waves. Continued erosion, in the same way, tends to reduce the relief of the land; and it is on the shore of these islands, that the beginning's can be noted of the chang-e from the present to the future topographical features of the Fox Islands. This work in which the waves are eng-ag-ed is simply that of cut- ting- back the salient and filling- in the re-entrant features of the coast. The former process can be observed at every exposed point on the islands, and the latter is no less apparent in many of the coves between these points. Thus, at Oak Hill, the extreme northern point on North Haven, the one process is to be seen, while on either side, beaches and sea-walls of shingle and sand have been thrown across the coves and the streams dammed, so that marshes now mark what was once the upper part of the coves. In many localities similar relations can be observed, and indeed it may be possible that on the long- southwestern peninsula of North Haven, barrier beaches of this character may have connected an old smaller island, about a half mile long-, to the larg-e island, forming- what is Crabtree Point. Another constructive process allied to the one just mentioned is that of sedimentation at the upper part of the long- coves and harbors. The fine sediments are broug-ht in by the entering- streams and by the tides from other parts of the shore, while the 10 GEOLOGY OF THE FOX ISLANDS. outflowing- tides are not strong- enough to sweep out the mud and sand, except in the channels where the scouring- action is more complete. Thus, observation of what is being- done on the coast at present warrants the assertion that a continuance of these processes must result finally in a coast line which will g-ain simplicity, but at the same time lose its present charm of variety. GEOLOGICAL. An oceanic island may be treated as a geologic unit; but lying- as they do within the g-eneral coast-line, the Fox Islands are to be reg-arded simply as detached portions of the main- land. Geologically, these islands are to be connected with the other areas of the region, just as if they were yet what we have termed the Fox Hills. Correlation with the rocks of other parts of this coast is necessary to a full understanding of the geologic problem. It is, therefore, unfortunate that the amount of geological work done on the coast of Maine has not been greater. The earlier State surveys of Jackson 1 and of Hitchcock 2 were only reconnais- sances of the mineral and agricultural resources. Later works by Professors Shaler 3 , Davis 4 , and Crosby 5 , however, furnish detail studies of limited areas to the east of Penobscot Bay, while that of Dodge and Beecher 6 is invaluable as determining the age of the sedimentary series on North Haven, and calling attention to the presence of associated volcanic rocks. The volcanic areas on the Maine coast were also mentioned by the late Professor Williams in his inspiring essay on the volcanic rocks of eastern North Ameri- ca 7 ; and the importance of the volcanic series of the Fox Islands has been noted by Dr. W. S. Bayley 8 and the writer 9 . 1. Annual Reports on the Geology of the State of Maine, 1837, l8 3, 1839. 2. Reports on the Geology of Maine, 1861, 1862. 3. Geology of Mount Desert. 8th. Annual Report, U. S. G. S. Am. Jour, of Science (3rd ser.) Vol. 32, p. 35. 4. An Outline of the Geology of Mt. Desert. Introd. to Flora of Mount Desert Isl. Me. 1894. 5. Proc. Boston Soc. Nat. Hist. 1880. p. 109. Am. Jour, of Science (3rd ser.) Vol. 23, p. 64. 1882. 6. Am. Jour, of Science (3rd ser.) vol. 43,^). 412. 1892. 7. Journal of Geology, vol. 2, p. i. 1894. 8. Bull. Geol. Soc. Am., vol. 6, p. 474. 1894. 9. Johns Hopkins Univ, Circular No. 121. 1895. INTRODUCTORY. 11 In its broader geologic relations, this group of islands is to be considered as forming a part of the Acadian province, which has been recognized as an area of sedimentation throughout Paleozoic time; and the position of the Fox Islands on the northern side of this great trough must be kept in mind as the geologic history is traced out. The work of the Canadian geologists may assist, since the periods of sedimentation and of disturbance were pro- bably the same throughout the whole geological province. NATURE OF THE PROBLEM. The problem is three-fold: To in- vestigate the geological relations of the different members of the rock series on the Fox Islands; to determine the original nature of the rock-types represented; and to trace the changes which they have subsequently undergone. The study is, thus, both geotec- tonic and petrographic. Its object is to obtain such data as to suggest an outline, at least, of the geologic history of the Fox Islands, and, perhaps, to contribute to the determination of those conditions which prevailed in the Penobscot Bay area at a time when it was the scene of long continued and widely extended volcanic action. AREAL DISTRIBUTION OF ROCK-TYPES. In their whole extent, the two islands of North Haven and Vinal Haven present consider- able variety in the rocks which compose them. The bold character of the coast and its great extent, due to the many indentations, afford opportunity to realize the complexity of these rock masses. In this complexity, however, there is system, and in the areal distri- bution of the rock-types represented is to be found the first clue to the solution of the geologic problem which is the subject of the present study. The delimitation of the different areas must be preparatory to the study of the rock-types and their relations. The island of North Haven is composed of greenstone schists, with the exception of the two peninsulas which extend southwest into the Thoroughfare, Indian Point and the larger one between Waterman's Cove and Southern Harbor. Since these schists pre- sent considerable variation, the area may be termed the North Haven Greenstone Complex. Immediately south of the green schists, and extending eastward from the shores of Southern Harbor, is a narrow strip of sand- 12 GEOLOGY OF THE FOX ISLANDS. stones, shales, conglomerates, and limestones, a fossiliferous series of Niagara Sediments. Other sedimentary rocks, quartzitic slates, wholly different from those of Niagara age, are found on the northeastern part of Vinal Haven; these will be called the CalderwoocTs Neck Series. Occupying the greater part of the two southern peninsulas of North Haven, already mentioned, are the red, gray and purple volcanic rocks. These continue across the Thoroughfare to the northeastern shore of Vinal Haven and the northwestern part of Calderwood's Neck, and the series being typically developed on the shores of the Thoroughfare may be termed the Thoroughfare Volcanics. The northwestern part of Vinal Haven, with the exception of the strip of volcanics just mentioned, is characterized by a differ- ent series of volcanic rocks, more acid in composition. These will be described as the Vinal Haven Acid Volcanics. Throughout these various areas are found dikes of igneous rocks, both acid and basic, which deserve mention. . The southern and greater part of the island of Vinal Haven is made up of holocrystalline rocks, granite, diabase, and diorite, the Intrusives. In brief, such are the areas, which will be considered in the fol- lowing pages. In the main, this order of discussion of the areas is that from north to south, a succession, however, which will be seen to be geological as well as geographical. NORTH HAVEN GREENSTONE COMPLEX. DISTRIBUTION. The greenstone area, as already denned, includes the greater part of North Haven. Of this triangular island, the northwestern coast, almost all of the eastern, and the greater part of the southern are composed of greenstone schists, which are also found outcropping- on the low hills of the interior. The topog- raphy of the region is without conspicuous features and the northern coast-line is rather exceptional in its simplicity. The general trend of the hills seems to be determined by that of the rock structure or by intrusive dikes. Of the whole area, that por- tion bordering- upon Southern Harbor presents the most variety in rock-types and structures. Here also, the relations of the greenstones to the rocks which lie to the south can be observ- ed, and a more careful study was made, therefore, of the rocks of this portion of the greenstone area. DESCRIPTION OF ROCK-TYPES. The old field term, "green- stone," has been employed for this area, since, although not exact, it is descriptive. The metamorphosed character of the rocks is evident, and the difficulties attending- the determination of their original nature justify the use of the name. The greenstones, in the main, are aphanitic and of a light yellow to grayish green color. In portions of the area the rocks are quite schistose, while more generally they show a columnar structure which will l?e con- sidered in a following- paragraph. Study of the shores of Southern Harbor shows the presence of large masses of amygdaloidal greenstone, as well as smaller amounts of a rock, which is rather tuffaceous in texture and at one point even breccia-like. The greenstone is locally also quite plain- ly crystalline, and it is often difficult to distinguish between this greenstone complex, to be regarded as a geologic unit, and the later intrusives. The number of dikes, both acid and basic, is greater here than in any other equal area on the two islands. 13 14 GEOLOGY OF THE FOX ISLANDS. DIABASES. The compact aphanitic rock found just west of the head of Southern Harbor is the least altered of the specimens ex- amined microscopically and thus favors the determination of the original nature of these greenstones. The structure is plainly diabasic with feldspar laths mostly small, but often approaching- a porphyritic development. Between the feldspar individuals, which have a composition about that of andesine, there are wedges of chlorite, nearly isotropic, which may represent the pyroxene constituent or possibly some amount of glassy base. The feld- spars show a considerable amount of alteration both at their edges and in the interior of the larger laths. Secondary minerals con- stitute a large part of the rock section and include clear grains of zoisite, small amounts of pyrite and calcite, spherulitic areas of chlorite, with stronger double refraction than the other variety, and water-clear grains of secondary feldspar, which appear in marked contrast to the cloudy original feldspar. Another second- ary mineral even more abundant than the zoisite occurs in very small yellowish grains, highly refractive and appearing quite white in incident light. The bright interference colors and im- perfect extinction make it probable that this is either epidote or titanite, very finely granular. Comparison with masses of the latter mineral resulting from alteration of ilmenite, favors the de- termination as titanite. Some of the sections contain considerable quantities of fine needles .005 to .01 mm. in length, golden yellow in color, with high refractive index, strong double refraction, parallel extinction, and positive optical character, all of which characters belong to rutile, thus showing the presence of titanium in these greenstones. The texture of these rocks and their composition, as shown by the remnants of original minerals and the nature of the secondary minerals, point to their original nature having been that of fine grained diabases or possibly diabase porphyries. The porphyritic phase of these diabases occurs at Websters Head on the northern side of the island, where in a rock, similar to the Southern Harbor greenstone in megascopic appearance, the diabase texture is less evident, the feldspar being more prominent and in larger individuals. The rock has been subjected to crush- ing and the feldspar fractured, one twinned individual having its NORTH HAVEN GREENSTONE COMPLEX. 15 fragments separated by quite wide intervals. The light green chlorite shows the Berlin-blue interference tint, characteristic of pennine. Brightly polarizing fibers of a light green mineral also occur in connection with the altered feldspar, and from its pleochroism and optical orientation, this mineral is believed to be an amphibole, probably actinolite. AMYGDALOIDS. The amygdaloidal type occupies the more south- ern parts of the area, immediately north of the Niagara sediments. On the weathered surface, the grayish greenstone is pitted, while a fresh fracture is spotted with shining calcite grains. In the thin section, the calcite areas are found to be surrounded by a ground- mass with feldspar laths and the general characters of that of the rock already described. The calcite amygdules are well defined, for the most part, but in some of the greenstones which show evidences of the most dynamic action, there are areas in which the calcite has replaced parts of the rock, thus enclosing feldspar laths. TUFFS. The best development of the tuffaceous type is at the head of Southern Harbor, although the tuffs also occur on Crab- tree Point and the southwestern shore of the island. In the most aphanitic rock of this type, the diabasic structure is wholly want- ing, and the rock is composed of an aggregate of secondary minerals similar to those found in the other types. These charact- ers may express a complete alteration of the diabasic rock, but more probably indicate an original difference. Other specimens are plainly tuffaceous, even megascopically, and the coarse breccia occurring just south of where the second stream enters South- ern Harbor from the east, is doubtless a tuff breccia. The greenstone complex thus resolves itself into a series of di- abase flows with accompanying pyroclastics. In places, these di- abases may have been porphyritic, while in the southern part of the area, they were distinctly characterized by gas cavities,' now filled with amygdules of calcite. The general strike of the series, as inferred from the succession of the different types, is somewhat north of east. The diabasic or ophitic structure has been thought to often ex- press crystallization under surface conditions. Lowinson-Less- 16 GEOLOGY OF THE FOX ISLANDS. ing* 1 considers diabases to have been submarine effusives at consid- erable depth, and augite-porphyrites the terrestrial flows from a magma of similar composition. Such a relation is thought bj- Barrois 2 to have held in the Silurian volcanic series of Brittany. This view interprets the diabasic structure as expressive of con- ditions of crystallization intermediate between those obtaining in intrusive masses and those of sub-aerial effusives. Nevertheless, the tuffaceous and amygdaloidal phases of the North Haven di- abases show that the conditions were essentially superficial and truly volcanic. The structure of these rocks as. described in the following pages also adds to the weight of evidence as to the vol- canic character of this greenstone complex. COLUMNAR STRUCTURE. Throughout the greater part of the area, the aphanitic and amygdaloidal diabases are characterized by a columnar structure. Seen in cross section only, as on a glaciated surface or low cliff on the shore, the rock appears to be divided into irregular ellipsoidal masses or lenses, from a few inches to two feet in major diameter. These masses of compact rock are embedded in a matrix, which is a schistose phase of the same rock, and the less resistance of the matrix to weathering agencies gives prominence to the structure, the oval sections often being surrounded by narrow crevices. There is considerable variation in the perfection of the structure. On especially favor- able exposures, the best one being a vertical cliff caused by a local fault, the weathering- has proceeded to such an extent that the masses of more compact greenstone project, and remind one of stumps of small trees in an inclined position. In short, the structure is seen to be that of irregular cylinders of indeterminate length. Somewhat similar structures have been observed in the green- stone schists of Michigan 3 , of Canada 4 , and of Saxony 5 , and in the diabase amygdaloids of Vogtland 6 . In all of the occurrences, how- 1. Bull. Soc, Beige, de. Geol. Vol. 2, 1888, p. 86. 2. Bull. geol. carte France. No. 7, 1889, p. 61. 3. Williams: Bull. U S. Geol. Survey, 62, p. 166. 4. Lawson: Report Geol. and Nat. Hist. Survey of Canada, 1885, C C. p. 52. 5. Rothpletz : Zeitschr. Deutsch, geol. Gesell., vol. 31, pp 374-397. 1879. 6. Dathe: Jahrbuch K. Preuss. geol. Landesanstalt, 1883, p. 432. NORTH HAVEN GREENSTONE COMPLEX. 17 ever, the structure is usually ellipsoidal rather than cylindrical; and where irregular masses are present they interlock as in a mosaic, which is not characteristic to any extent of the North Haven diabases. The origin of the structure in the case of the Vogtland amygda- loid is believed by Dathe to be contraction during the consolida- tion of the diabase flow and in support of this view, a concentric arrangement of the anrygdules parallel with the surface of the ellip- soids is cited. No such arrangement was observed in the diabases of North Haven, which otherwise have many points of similarity with the Vogtland rock. Such an origin, moreover, would fail to account for the schistose matrix between the columns. A dynamic origin, that of brecciation in situ, has been consider- ed a more natural explanation for the other occurrences mentioned above. In the present case, however, the origin seems to have been compound, a true contraction parting, modified by dynamic action. At one point on the north shore of Southern Harbor, a few typical pentagonal prisms are to be seen in the low cliff, and on either side, the transition into the phase already described is so gradual that no sharp line of limitation for the prismatic part- ing could be drawn. Columnar parting of this type is very common in effusive sheets, and might be expected in large masses of amygdaloidal diabase. The more or less vertical cracks which divide the mass into prisms are due to contraction after consolidation, under the influence of surface conditions. The cooling plane is the surface, approximately horizontal, and the hexagonal prism is the normal form, resulting from the contraction in two directions chiefly 1 . Usually, however, the number of sides of the prisms varies from three to nine. Such parting is functional upon both quantity of contraction and rate of cooling subsequent to consolidation; thus, it is most prominent in comparatively crystalline basic lavas, and near the surface of the lava mass, preferably when rough and slaggy. All of these conditions are seen to have existed in these amygdaloidal diabases. In such a surface flow considerable i. Iddings: Am. Jour. Sci. Vol. 31, 1886, p. 321. Other literature on the subject is, Bonney: Quart Journ. Geol. Soc , 1876, p. 140, and Reyer: Theoretiscbe Geologic, p. 90. 18 GEOLOGY OF THE FOX ISLANDS. irregularity in the form and arrangement of the prisms is to be ex- pected. Prismatic parting- is believed by Walther 1 to be due to cooling of a lava stream. in contact with water, and thus to mark shore conditions. Such a view might be in harmony with the hypothesis of a submarine origin for diabase, already cited; but the hypoth- esis seems hardly necessary. If a lava mass, parted into these prismatic blocks, be subjected subsequently to even a slight amount of differential movement the prisms wuld be more or less modified in form. The prismatic angles would be rounded, and the prisms thus become rude cylin- ders, in a schistose matrix. Such action might be favored by an ellipsoidal parting subsequent to the prismatic, with the long axis of the ellipsoid parallel to that of the prism. This position of the ellipsoid would follow from the fact that in the vertical direction, the resistance to contraction would be counteracted by the weight of the superincumbent mass 2 . This hypothesis of subsequent dynamic action as applied to the North Haven occurrence is supported not only by the field rela- tions, but by the microscopical features as well. Both the com- pact columns and the interstitial schistose portions are seen to have suffered dynamic action, but in different degrees. In the former, the calcite filling the old gas cavities shows a slight amount of deformation marked by curved cleavage cracks, while the feldspar laths are always bent. In the schistose matrix, the calcite amygdules have been shattered, a mosaic replacing the large calcite units, and the feldspars are broken. The amygdules are also less sharply defined and the amount of calcite is relatively greater in the schistose phase. Such dynamic action, then, was not only subsequent to the formation of the feldspar laths and the consolidation of the rock, but later than the filling of the amygda- loidal cavities of the porous lava. In many cases, the structure of the diabase may be ellipsoidal, rather than columnar, but the latter is the more characteristic for the area. The former could result from a similar dynamic modifi- 1. Walther: Jahrbuch. K. K, geol. Reichanstalt vol. 36, 1886, p. 295. 2. Iddings: op. cit. NORTH HAVEN GREENSTONE COMPLEX. 19 cation of a mass of lava parted into shorter prisms or even ellip- soids. At one locality these structures can be compared with the concentric weathering* in a dike, and the contrast is such as to allow no confusion of the three different structures. AGE. The age of these diabasic lavas is determined as pre- Niagara from their relations to the conglomerate at the base of the Niagara sediments. The contact is well exposed on the shore of Southern Harbor, where the amygdaloidal diabase underlies the conglomerate which dips to the south at an angle of about 60. The extremely disintegrated character of the diabase at this point and the apparent enclosing of large lenses of this rock by clastic material may indicate superficial disintegration before the deposition of the Niagara. In fact such action would be expected in the case of these vesicular lavas. NIAGARA SEDIMENTS. DISTRIBUTION. The area occupied by these sedimentary rocks is smaller than that of any other series on the Fox Islands. The principal exposure of the Niag-ara series is a wedg-e-like area ex- tending- east from the shore of Southern Harbor, where it is less than a quarter-mile in width, across the Ames Knob peninsula, with the apex of the wedge east of the mud flat and near the road from North Haven villag-e to Pulpit Harbor. Its full extent is thus less than a mile in leng-th, while the greatest width is ex- posed on the shore of Southern Harbor. These fossiliferous sedi- ments are also found in a narrow strip on the western end of Stimpson's Island. Throug-hout both areas, the surface is low and the rocks are mostly concealed, except on the shores. SOUTHERN HARBOR SECTION. The best section of the Niag-- ara sediments is found on the shore of Southern Harbor, just northwest of Ames Knob. Here the basal member is in immedi- ate contact with the greenstone to the north. The upper members are about seventy feet from the overlying- volcanics, but this break is in part represented a short distance further east by red shales with interbedded conglomerate. The section was measured at low tide, and is much more complete than that exposed above the hig-h water mark. The section, in descending- order, is as follows: Sandy shale and quartzite, 65 Concealed, 70 Gray shale, 49 Conglomerate and shale, 12 Red shale, 55 Conglomerate, 1 Red shale, 10 Gray shale, 37 Concealed, 17 20 NIAGARA SEDIMENTS. 21 Gray shale and limestone, 71 Quartz-porphyry, intrusive sheet, 10 Concealed, 16 Gray limestone, weathering- brown, and shale, 104 Grit, coral fragments, 1/i Sandy limestone, 4 Concealed, 42 Basal conglomerate, 16 580X feet. The general strike of the members of the series varies from N. 85 E 1 . near the base to ^". 40 E. in the upper members in the shore section. Higher beds, exposed on the north slope of Ames Knob, however, give a strike similar to that of the basal mem- bers. The dip is from 50 to 60 to the south and even steeper at a few points. To the east, on the west shore of the mud flat, the lower beds of the series have much the same sequence and strike, but the upper portions show some slight folding and displacement, so that the strike and dips vary somewhat from those of the more regular section, already given, where these minor folds are less apparent. FAUNA. The paleontology of this series of sediments has been carefully studied by Prof. C. E. Beecher of Yale, and the follow- ing full quotation is made from his article already referred to 2 . "The paleontological evidence furnished by the specimens from North Haven apparently does not admit of a distinct sub-division of the rock series into recognized periods. A large proportion of the species clearly points to a correlation with the Niagara of New York, and to this period they are here referred. The presence of quite a number of characteristic Clinton species is a disturbing element in any attempt to draw exact parallelisms with the New York sub-divisions of the Upper Silurian. From present data, however, it seems justifiable to consider the fossiliferous rocks at North Haven as representing a faunal equivalent to the Clinton and Niagara, with a decidedly strong Niagara facies. Therefore, 1. Referred to the true meridian, as the magnetic variation is over i6 c W. 2. Am. J. Sc., vol. 43, pp. 416-418. 1892. 22 GEOLOGY OF THE FOX ISLANDS. the broader term Niagara will more correctly express the chrono- logical relations of these strata. "The Clinton species are mostly confined to the lower half of the series. This is of considerable significance, but, at the same time, nearly all the typical Niagara forms are associated with them in abundance. Another fact to be noted is the conspicuous absence of some species from the entire series: as, Caryocrinus ornatus, Spir if er niagarensis, Rhynchotreta cuneata, and Pentamerus oblongus. Similar discrepances occur in the Niagara fauna at Waldron, Indiana, and elsewhere, and merely serve to give dis- tinctive features to such faunas." On account of subsequent collection, the list of fossils has been revised by Prof. Beecher, and through his kindness, the more com- plete fauna is here given. Monograptus clintonensis Hall. Coenostroma, 2 species. Streptelasma calyculum HalL Cyathophyllum sp. Chonophyllum niagarense HalL Favosites venustus HalL Favosites niagarensis HalL Favosites f avosus HalL Cladopora sp. Coenites ramulosus HalL Syringopora sp. Monticulipora sp. Halysites catenulatus Linne. Heliolites spiniporus HalL Crinoid fragments. Ichthyocrinus laevis. Kucalyptocrinus caelatus. Cornulites sp. Tentaculites sp. Encrinurus punctatus Wahl. Calymene niagarensis HalL Homalonotus delphinacephalus Gr. Dalmanites limulurus Green. NIAGARA SEDIMENTS. 23 Ceraurus niag-arensis Hall. Illaenus ioxus Hall. Proetus Stokesi Hall. Proetus sp. Beyrichia, 2 species. Leperditia sp. Fenestella sp. Stictopora sp. Ling-ula lamellosa Hall. Orbiculoidea sp. Orthis eleg-antula Dal. Orthis hybrida Soiv. Orthis, 2 species. Plectambonites transversalis Wahl. Plectambonites (cf.) sericea Sow. Leptaena rhomboidalis Wile. Stropheodonta profunda Hall. Orthothetes subplanum Con. Chonetes cornutus Hall. Pentamerus occidentalis Hall. Whitfieldella nitida Hall. Meristina sp. Nucleospira pisum Hall. Atrypina disparilis Hall. Spirifer crispus His. Spirifer sulcatus His. Spirifer radiatus Soiv. Cyrtina pyramidalis Hall. Atrypa reticularis Linne. Atrypa nodostriata Hall. Rhynchonella neg-lecta Hall. Rhynchonella obtusiplicata Hall. Rhynchonella sp. Rhynchonella (Wilsonia) sp. Nucula sp. Tellinomya sp. Avicula demissa Con. ? Avicula subplana Hall. 24 GEOLOGY OF THE FOX ISLANDS. Hydrozoa, Actinozoa, Crinoidea, Annelida, Trilobita, Ostracoda, Avicula sp. Cypricardinia sp. Platyostoma niagarense Hall. Ivoxonema sp. Pleurotomaria sp. Bellerophon sp. Cyrtolites sp. Murchisonia sp. Hyolithes sp. Orthoceras annulatum Sow. Orthoceras subcancellatum Hall. Orthoceras virgulatum Hall. Orthoceras (annulated), 2 species. Oncoceras sp. Gomphoceras sp. Cytoceras subcancellatum Hall. SUMMARY. 3 species. Bryozoa, 12 2 2 8 3 2 species. 27 6 " 6 Brachiopoda, Pelecypoda, * Gastropoda, * Pteropoda, 1 " Cephalopoda, 8 " Total 80 species. Prof. Beecher states that nearly all the corals are confined to the lower beds, being- found in a conglomerate of coral fragments together with quartz and hydromica schist pebbles, and in a lime- stone. The greater number of species was found in a shale near the middle of the series, on the shore northeast of Ames Knob. He concludes: "The fauna is evidently a rich one, and several of the unidentified species are probably new to science. Many of them show distinctive characters but the majority are too frag- mentary and poorly preserved for accurate description or determi- nation." CONDITIONS OF DEPOSITION. A series of old sediments is of the greatest value to the geologist, since it affords an expression of past conditions, both terrestrial and submarine. Land topogra- NIAGARA SEDIMENTS. 25 phy and configuration of the sea-floor have conditioned the processes of sedimentation, but the determination of these con- ditions, through study of a single section of sedimentary rocks, is not a simple task. In the series of about six hundred feet of Niagara sediments, as exposed on North Haven, the general sequence is that of basal conglomeratic and arenaceous beds, fol- lowed by limestones, which grade upward into shales and coarser sediments. The basal member of the series is a conglomerate which includes in its lower portion larger well-rounded masses of the subjacent greenstone, but in its upper part, well-rounded pebbles of black and green slates, quartzite, and vein-quartz, materials unlike the under- lying rock. Such a basal member may be considered to have been deposited in a sea transgressing over a land surface of rock more or less disintegrated, thus accounting for the residual boulders of greenstone. In the later stages of the deposition of the con- glomerate at this point, the processes of sorting became more efficient and the material was derived from a greater distance. The limestone-shale series shows changing conditions, which, however, were most constant during the deposition of the hundred feet of limestone with slight amounts of shale, overlying the grit. This latter bed, from its evenness of grain affords evidence of perfect sorting, and the presence of coral fragments may be con- sidered to indicate breaker action. These calcareous deposits are quite fossilferous and the conditions, therefore, were such as were favorable to the lime secreting organisms. Even within the shaly bands, nodules of limestone occur and some of the masses are plainly coralline. Increasing amounts of silt being contributed to the sea terminat- ed the deposition of limestone, and the upper portions of the section are essentially argillaceous. In these shales the remains of the Niagara organisms are even better preserved than in the lime- stone. For the most part, the increased supply of sediments probably originated in greater efficiency of stream transportation, which, in turn, was brought about by uplift of the drainage area, tributary to this portion of the Niagara sea. That there were changes in wave and current action is shown by the presence of thin conglomerates, resting upon well-washed surfaces of the 26 GEOLOGY OF THK FOX ISLANDS. shale, but grading- upward into the shale. At one locality, traces of ripple-marks can be detected in the quartzite associated with the shales. The red shales which occur as two important members in the shore section and also higher up in the series exposed a short dis- tance to the east, deserve special mention. They are of interest both from their relations in the section, being 1 at one point in direct contact with the superjacent volcanics, and also from their close resemblance to certain fine-grained tuifs within the volcanic series. This similarity is primarily one of color, for none of the tuffs quite approach the fineness of grain that characterizes these shales. Two hypotheses may be presented to account for the red color so noticeable in these shales. One is that proposed by Russell 1 as a general explanation of the red color of certain formations. The material from which these shales are derived has resulted from subaerial decay of crystalline rocks, and in the process of deposi- tion has not been subjected to attrition sufficient to separate the ferric oxide from the silt or fine sand grains. Such an origin in- volves definite land conditions and special shore conditions. Of these the latter would probably vary most in the deposition of a series of sediments like the one under consideration; and the red shales would indicate the absence of prolonged wave action to which the associated gray shales and clean quartzites must have been subjected. The abrupt change to the well-rounded conglomer- ate, which is found interbedded with the red shales in both occur- rences, is difficult to explain by this hypothesis. The origin suggested by the field relations is that the red color is due to fine ash from neighboring volcanic explosions. Such volcanic dust could easily supply the iron to color these sediments, and this hypothesis is supported by the resemblance to the true tuffs in the overlying volcanic series. Prof. Emerson 2 , as quoted by Russell, gives a similar explanation for the red color of some of the sandstones of the Connecticut Valley. Thin sections of this shale examined microscopically show the 1. Bull. 52, U. S. Geol. Survey, p. 44. 2. Gazeteer of Hampshire County, Mass. 1888, p. 18. NIAGARA SEDIMENTS. 27 rock to be composed of angular fragments of colorless minerals, .01 to .05 mm. in diameter, enclosed in a matrix nearly opaque even in the thin section. The mineral fragments seem to be quartz or feldspar for the most part, with some shreds of musco- vite, and small amounts of chlorite and calcite. The red color is due to a fine pigment, some oxide of iron, which not only coats the fragments but constitutes a large part of the rock, the dark red material appearing to be plentiful in fragments as well as forming the cement. The thin sections were treated with acids but with no appreciable bleaching effect. The similarity to the finer parts of the Hypical tuffs, occurring within the volcanic series appears under the microscope as well as in the field. The evidence, therefore, seems to favor the latter hypothesis, namely: that the red color is not due to a cement of residual clay, but to finely comminuted material, ejected from a volcano and deposited in the shallow water close to the old shore. Thus the red shales would present the earliest evidences of the volcanic activity of the region, an activity which later resulted in the red and gray lavas, which overlie the Niagara sediments. CALDERWOOD'S NECK SCHISTS. DESCRIPTION AND DISTRIBUTION. The rocks included in this series are dark quartzitic slates, often rusty on the weathered surface, banded schists, quartzose but varying in color and grain, and quite massive quartzites. These rocks occupy the northern half of Calderwood's Neck, with the exception of the narrow part west of Carver's Cove. Felsitic rocks occur in close proximity to the granite and volcanic rocks and undoubtedly represent contact phases of members of this Calderwood's Neck series. These are mottled or variegated, the general color being a light green. In the thin section, they are seen to be composed of nearly pure epidote with some zoisite, in the lighter areas, and of green hornblende in the darker. Connected with this larger area is a narrow strip of similar rocks along the western shore of Seal Cove. Here the rocks are somewhat more massive, being both quartzitic and slightly cal- careous. Epidote is quite prominent along the bedding planes, giving a banded appearance in many places. Rocks resembling these in general appearance occur elsewhere on Vinal Haven, on the shore of Leadbetter's Narrows, and on the extreme eastern side of the island near Coombs' Hill. In the latter occurence the schists are actinolitic in appearance, except where in contact with a granite dike. Here there is a marked bleaching of the schist similar to that observed at several localities in the Calderwood's Neck area. There are also indications of small masses of these schists within the diabase area of Barley Hill, and the volcanic area on the western side of Vinal Haven, but these areas are too small to be mapped. On North Haven, the presence of this series is not indicated at any point. RELATIONS. The series in its principal development distinctly shows sedimentary characters. The bedding can be observed even at some distance and seems quite constant, and the banding of the schists agrees with it, the strike, for the Calderwood's Neck area 28 CALDERWOOD'S NECK SCHISTS. 29 being" usually within a few degrees of north-south, and the dip 30 to 40 to the west. On the shores of Seal Cove, however, the rocks dip more to the northwest. At no point do these schists come in contact with the greenstone or Niagara sediments already considered. East of Seal Cove and north of Mill River, the schists are interbedded with quartz- porphyry, with the relations apparently those of intrusive sheets of porphyry in the sedimentary rocks. Considerable metamor- phism has taken place in the latter series, and epidote is present along- the joint planes. On the opposite shore, south of Perry's Cove, diabase sheets and dikes mask the contact with the over- lying- volcanics. AGE. The contact relations prove that these schists are older than the granite and diabase intrusions and the volcanic series. In their metamorphic condition and in the absence of all traces of fossils, these old sediments are in marked contrast to the Niagara shales. Their practically uniform composition for a distance of nearly a mile across the strike, on Calderwood's Neck, show the condition of their deposition to be wholly different from those of Niagara time, since in the Niagara section of six hundred feet, the sediments show great variety. It seems, probable, therefore, that these schists may represent an older sedimentary series, and with the North Haven diabase flows constitute the oldest part of the Fox Islands. THOROUGHFARE VOLCANICS. DISTRIBUTION. Anyone visiting- the village of North Haven, which is one of the many Maine summer resorts, or even sailing- through the Fox Islands Thoroughfare, cannot fail to notice the red, yellow and purple rocks that form the bold shores. These are the volcanic rocks, most typically developed on North Haven, but also extending to the northern shores of the southern island. The two hills northwest and northeast from the village, mark the northern limit of this volcanic series in its chief development; and throughout, these more resistant rocks determine a bolder topogra- phy than that seen to the north. On the southern side of the Thoroughfare, the series continues to the smaller portion of Cal- derwood's Neck, while the extreme southwestern limit of the area is at Brown's Head. This main area, then, is crescent-shaped, concave southward. Similar volcanics form the peninsula ending in Indian Point, as well as the five islands that lie to the east and south. DESCRIPTION OF ROCK-TYPES. The area just defined has been termed one of volcanic rocks. However, the determination of the volcanic nature of these rocks, so different in appearance from the lavas and other products of recent volcanic action, depends upon a careful investigation of their many characters. Therefore, a description of their structure and composition is necessary, and this, together with a statement of the field relations, will con- tribute to the determination of the conditions of the volcanic activity. The two great classes of volcanic rocks are the lavas and the pyroclastics, and the usual gradation from the one type to the other is found among the Thoroughfare volcanics. The lavas of this series include several distinct types and deserve careful study, since they furnish what evidence there is as to the nature of the magma from which these volcanics originated. Among these lavas are found old andesites, porphyries, and amygdaloids. 30 THOROUGHFARE VOLCANICS. 31 ANDESITKS. A rock very prominently exposed on Ames Knob, the bold hill northwest of North Haven village, is one of a dark gray color, often with a slight purple tinge. It is porphyritic with well defined phenocrysts of feldspar, yellow to greenish white, and at one locality on the north slope of the Knob, the rock is quite beautiful in appearance. The abundant feldspar pheno- crj^sts are in striking- contrast to the almost black groundmass of the rock, which much resembles ohe variety of the "porfido antico". Typically, this Ames Knob rock is compact, but amygda- loidal and even drusy phases also occur. In the thin section, the phenocrysts and ground-mass appear well differentiated. The feldspar crystals, with the extinction angles of labradorite approaching- andesine, are tabular and of good size, with albite-twinning- and zonal structure. Although the feldspar phenocrysts are plentiful, few are unaltered. Aggre- gates of epidote and pennine, with some clear grains of secondary feldspar, have replaced the original feldspar without affecting the crystal outlines. The ferro-magnesian constituents have been completely altered and phenocrysts that appear well defined in ordinary light, are seen in polarized light to be composed of chlorite and calcite. The outlines suggest both hornblende and pyroxene. These latter phenocrysts are older than certain of the feldspar crystals into which they project. The ground-mass is composed of feldspar laths, with a notice- able hyalopilitic or felted texture, as well as a fluidal arrangement near some of the phenocrysts. The light brown base is isotropic in places and much darker where included between several feld- spar crystals. Magnetite occurs throughout the ground-mass in small grains and larger masses are associated with the ferro- magnesian phenocrysts. In addition to the altered crystals of feldspar, there are areas of epidote, roughly spherulitic, with some zoisite, and areas of chlorite, both of which by their outlines suggest irregular amyg- dules. However, the laths of the ground-mass show no parallel- ism with the borders of these areas; and in a few cases, the secondary minerals surely represent an alteration which has ex- tended beyond the outlines of the original crystals. This porphyritic rock presents the characters of an andesite> 32 GEOLOGY OF THE FOX ISLANDS. especially in the texture of its ground-mass. Upon the determina- tion of the darker phenocn r sts would depend whether it is to be considered as originally a hornblende- or pyroxene-andesite. Com- parison, however, with types to be described later favors the de- termination as a pyroxene-andesite 1 . The rock well exposed on the southern part of the hill just north of North Haven village is compact, very dark although of a purple tone, and quite, fresh in appearance. The feldspar crystals are apparent only by reason of their bright cleavage faces, and dark green phenocrysts are also noticeable. In the thin sec- tion, this rock proves to be of special interest from the opportuni- ties it affords for the study of the processes of alteration 2 . The feldspar phenocrysts exhibit the albite-lamellation, have the extinction angles of labradorite, and are beautifully zonal, being slightly more basic toward the center. Many of the crys- tals have an altered core, but are perfectly idiomorphic, showing that the clear outer zone is original. (Plate I, Fig. 3.) The light green pyroxene phenocrysts have the high extinction angle of augite, and orthopinacoidal twinning is the rule, being often re- peated in the same crystal. The augite, in part, is altered to pen- nine. Both feldspar and augite phenocrysts have a zonal arrange- ment of inclusions, showing changing rates of crystallization in the later stages. The ground-mass is less andesitic in character than that of the rock described above and is full of minute mag- netite grains. Some interstitial areas of glass appear within groups of feldspars, while elsewhere, faint micropoikilitic areas are common in the ground-mass. Other andesites are found which are of a dark red color, both on the weathered surface and the fresh fracture. The bright feld- spar phenocrysts are also often tinged with red, and although this 1. The use of the term andesite for these Paleozoic lavas is not in accord with the petrographical nomenclature of the German school. Geological facts, however, justify this disregard of the age element in rock classification, and the usage adopted by the writer is that of many American as well as English petrographers. vid. Iddings : I2th Ann. Rep. U. S. G. S., p. 582; and Harker : Petrology for Students, p. 128. Rosenbusch also favors this reform in petrographical terms, vid. Microskopische Physiographic der Massigen Gesteime, 1895, PP- 4~7- 2. The results gained from the study of the alteration of all the Fox Islands rocks will be discussed in a subsequent section. THOROUGHFARE VOLCANICS. 33 rock lacks somewhat of the brilliancy, it suggests the "porfidp rossa antico" of Egypt. One specimen has a ground-mass charact- erized by minute, irregular particles of red to brown opaque material. The ground-mass itself now consists of micropoikilitic areas, but, doubtless, was originally very glassy. The feldspar phenocrysts are less numerous than in the other andesites and are badly altered, but from the amount of calcite present, a rather basic character may be inferred. A few phenocrysts older than the feldspar, as seen in ordinary light have the characteristics of olivine. (Plate I, Fig. 1.) The association with magnetite, the dark border of iron oxide, the irregular transverse cracks, and the badly corroded outlines of the prisms constitute the points of re- semblance. In polarized light it is seen that none of the olivine remains, but that these phenocrysts consist of fibrous serpentine, with fibers normal to the transverse cracks, which are bordered by the opaque oxide. Small grains of quartz plainly secondary occur within the serpentine mass. Not only is magnetite included within these olivine phenocrysts, but also large hexagonal prisms of apatite, the oldest constituent of the rock. From the former presence of olivine as an original constituent, this rock would naturally be considered the most basic of the North Haven volcanics, and, therefore, a chemical analysis has been made by Messrs. K. W. Magruder and W. A. Jones. SiO 2 . ' * . 63.25 Ti O2 .... . . . trace A1 2 O 3 . 14.89 Fe 2 O 3 . . . . . '..". . 6.54 Fe O . . . . . . . none MgO . IS ; . . ; . 0.82 CaO - '.. 'v . . . / 0.59 K 2 O ... . . . . . 4.78 Na 2 O . ., ~. . . . . ; 4.47 P 2 O 5 . . . .' . . . . 0.61 H 2 O (by Pb O) . . . . . 2.67 C O 2 . . . . . ... 0.78 Loss by ignition, not accounted for by H 2 O and C O 2 , and including small amounts of Cl & F of apatite .. .. 0.53 99.93 34 GEOLOGY OF THE FOX ISLANDS. In its mineralogical composition, this rock approaches the basaltic type, but as the analysis shows is somewhat too acid. The olivine phenocrysts, moreover, are not very numerous and there is reason to regard this as simply an olivine-bearing phase of the andesite. Another specimen of the red andesite, collected on the same slope of Ames Knob, has the large feldspar phenocrysts quite un- altered, and with extinction angles indicating basic andesine. The ferro-magnesian constituent is not present except as well de- fined prismatic areas, now composed of chlorite and calcite. There is much magnetite and some ilmenite, and apatite is associated with the magnetite as inclusions in the feldspars, and also as long prisms in the ground-mass. One crystal of titanite or sphene was observed, altered at the edges into calcite and an opaque oxide. The ground-mass contains feldspar laths more acid than the phenocrysts, usually simple but sometimes twinned, and often irregularly terminated. There are also present laths and trichitic masses of a brown color. The flow structure is well developed near the phenocrysts (Plate I, Fig. 2.) but elsewhere the parallel- ism of the elements of the ground-mass is less apparent. With crossed nicols, the micropoikilitic structure of the ground-mass can be seen, but this seems to be due to secondary crystallization. Another of the dark gray andesites is of much coarser texture, with larger phenocrysts of yellowish white feldspar. In the thin section, these feldspars are seen to have suffered the usual altera- tion to epidote and zoisite, and the porphyritic darker constituents are also wholly altered. They have the opaque resorption borders and, therefore, probably represent either hornblende or mica, some cross-sections approximating the angles of hornblende. Magnetite is very abundant and apatite occurs in some of the more basic areas. The ground-mass is lighter colored, although containing minute grains of magnetite. Microlitic laths of feldspar constitute a thin felt, and the clear base has a very slight effect upon polarized light, even appearing isotropic in great measure. It is probable that this rock is essentially a hornblende-andesite, rather more acid than many of the other lavas here described. THOROUGHFARE VOI.CANICS. 35 PORPHYRIES. A dark green rock associated with these ande- sites has many of the characters, both megascopic and microscopic, similar to those described above, but structural differences in the ground-mass seem to warrant a separation of this type from the andesitic. The feldspar phenocrysts are almost wholly altered, but the twinned crystals of monoclinic pyroxene are quite fresh, some chlorite being- present as a product of the partial alteration. The pyroxene crystals are nearly colorless in the thin section and have the optical orientation of augite. The ground-mass is holo- crystalline, but very fine-grained and approaches the ophitic structure of diabase. The rock, therefore, may be called a di- abase-porphyry 1 . At the base of Ames Knob on the south side occurs another porphyritic rock of a green color. The yellow feldspars are very numerous, the porphyritic character of the rock becoming- some- what masked by their abundance and the presence of black dots of a secondary mineral. This rock is essentially holocrystalline and since the feldspar laths of the ground-mass show some fluidal arrang-ement around the phenocrysts, this type may be termed an andesite-porphyry. The feldspar phenocrysts are labradorite, and augite is represented sparingly, mostly altered to chlorite, but giving- the hig-h extinction ang-le where unaffected. Chlorite also forms the matrix of the ground-mass. The black spots seen in the hand-specimen are found to be areas of very finely divided zoisite of a lig-ht brown color in the thin section. One of these ag-gre- g-ates is at the center of an altered feldspar crystal. Kpidote and secondary feldspar are other alteration products in this andesite- porphyry. Other rocks within this area appear quite holocrystalline. The darker silicates' are replaced by secondary minerals, chlorite and calcite, and the feldspars are badly altered. Much ilmenite is present, with associated leucoxene, which, in part, has the char- acters of sphene, and the rhombohedral parting- of the ilmenite is very finely exhibited throug-h this alteration. The ground-mass is i. Porphyry is used here purely as a term indicative of structure. Its retention in this sense throughout all the rock families, thus abandoning the term porphyrite, is the usage favored by Cross, Iddings, and Pirsson. 36 GEOLOGY OF THE FOX ISLANDS. microgranitic with no evidence of any flow structure, such as characterizes most of the other rocks described. This rock type corresponds to the porphyrite of many petrographers and may be termed either a diabasic or a dioritic porphyry. In one diabase-porphyry, with labradorite phenocrysts, there are granuar patches of more acid feldspar. The interlocking* grains are much larger than the laths of the ground-mass and contain relatively more magnetite and hematite than the rest of the rock. This striking- variation in composition and texture doubtless represents a primary segregation within the crystallizing- lava. Within this North Haven area, there also occur quartz- porphyries, which from their association with equivalent pyro- clastics are to be considered as belonging- to the volcanic series. The meg-ascopic determination of these yellow porphyries is not difficult, since the quartz phenocrysts are plainly seen. These quartzes are prominent in the thin section, and furnish beautiful examples of well defined phenocrysts, the sharp outlines of the bi- pyramids being interrupted only by the deep embayments. In some instances, this effect of mag-matic corrosion has been in part remedied by the subsequent addition of quartz, which seems to be of the same period of crystallization as the ground-mass. The feldspar phenocrysts have altered to calcite and white mica, and secondary muscovite also clouds the interlocking- grains of the ground-mass. In a few cases, these grains are of twinned feld- spar, and in others, they have small areas of quartz at the center surrounded by micropeg-matitic inter growths. Similar intergrowths constitute a marked feature of another quartz-porphyry, the volcanic character of which is less certain. Here, however, the micropeg-matitic areas envelop the quartz phenocrysts and are in crystalline continuity with them, forming- wide borders, which often reach from one crystal to another. AMYGDALOIDS. Meg-ascopically, one of the most noticeable rocks on North Haven is an amygdaloid which occurs on the shore of the Thoroughfare. The ground-mass of the rock is gray, and the elongated amygdules are very plentiful, being composed of white and pink quartz. On weathered surfaces, the rock often ap- pears very vesicular. THOROUGHFARE VOLCANICS. 37 The ground-mass when studied microscopically resolves itself into short feldspar laths, badly altered, in what may have been originally a glassy base, quite dark in color. The feldspar laths show some parallelism to the edges of the amygdules, and also a certain amount of flow-structure. The few phenocrysts of feldspar have characteristic outlines but are now wholly altered to chlorite,, calcite, and quartz-feldspar aggregates. The amygdules constitute one-fourth of the rock section. The structure is concentric, quartz grains and chlorite forming the outer zone, from which quartz grains or perfect hexagonal prisms project inward. The central portion is filled with calcite or quartz, the latter, in a few cases, being in roughly spherulitic plumes. This rock may be termed an andesitic amygdaloid. A dense fine-grained red rock occurs with the gray andesites on Ames Knob, and differs from them in the small number of pheno- chrysts. Examined microscopically, it is found to represent an ex- tremely vesicular lava, the irregular cavities having been filled with secondary minerals. The base is of a red color but wholly opaque and thus unlike that of the other lavas studied. One feld- spar crystal was observed, but the degree of alteration prevented any determination of its composition. The lavas which have been described above are the andesites, pyroxene-, hornblende-, and basaltic, the andesite- and diabase- porph}^ries, and the amygdaloids, comprising a series varying quite widely in structure and mineralogical composition but pro- bably much less in chemical composition. Associated with these lavas of intermediate composition is the quartz-porphyry, a much more acid type. PYROCXASTICS. Associated with the lavas are the volcanic fragmental rocks, the pyroclastics. They are igneous in origin, but clastic in structure and thus connect the compact lavas with the ordinary sediments. So close, indeed, is the relationship and the resemblance between these pyroclastics that it often is difficult to determine to which class some of the members of the North Haven series belong. On the other hand, it becomes a problem where to draw the line of separation between the pyroclastics and the more normal sediments, to which, moreover, the volcanic de- 38 GEOLOGY OF THE FOX ISLANDS. tritus has contributed so much. It is necessary to realize that the gradation throughout is complete and sharp distinctions are, to a great degree, artificial. The most important of the pyroclastics represented on North Haven is the volcanic conglomerate on the slopes of Ames Knob. The more or less rounded fragments of the different andesites vary in size from that of a boulder several feet in diameter to that of coarse sand. The cementing material is small in amount and its character difficult of determination; but, when studied micro- scopically, this matrix is seen to be clastic, consisting of angular crystal fragments and particles of opaque material similar to that forming the ground-mass of certain of the lavas. Slight traces of flowage may indicate that the cementing material was of the character of a volcanic mud, when it filled the spaces in this mass of coarser ejectamenta. The large fragments show the andesite structure in a marked degree, and some closely resemble the dark vesicular lava last described. A volcanic conglomerate somewhat finer-grained occurs on Iron Point, and the fine dark red material which cements the fragments of andesites again suggests an original volcanic mud. In both oc- currences, the volcanic conglomerate grades into a breccia, which is usually of the character of a lava including blocks of older lavas. Such a breccia would be a primary or flow-breccia. In other cases, however, the breccia seems to differ from the volcanic conglom- erate only in the degree of rounding of the fragments and, there- fore, may be termed a tuff-breccia. The pyroclastics of the North Haven series afford examples of the gradation in size of grain, from the coarse conglomerate to the tuff, composed of fine volcanic dust. The coarser tuffs are identi- fied most easily, since the angular andesite fragments can be seen on the weathered surface of these tuffs, which are usually dark colored and compact, although often quite resembling a sandy shale. As in the conglomerates, the included fragments have been derived from the various andesitic lavas, the more basic, the vesicular, and those containing the altered hornblende phenocrysts with resorption rims. Mineral fragments also occur, magnetite and ilmenite being most abundant, and the former characterized by apatite inclusions. These mineral grains are extremely angu- THOROUGHFARE VOLCANICS. 39 lar and cannot be considered as having- resulted from disintegra- tion of the andesites through atmospheric agencies. The cement- ing- material is larg-ely calcite. A series of tuffs much finer in grain are the compact pink rocks on Ames Point, which on the weathered surface are lighter colored and plainly tuffaceous. In the thin section, they show very few larger fragments of minerals or of lavas, but consist of a mass of almost colorless angular particles. These minute frag- ments seem to represent glass sherds, with conca've triangular out- lines for the most part, often crescent or sickle-shaped. The structure determined by this material is that described as the "aschen-structur" by Mugge 1 , and considered by him to be char- acteristic for deposits of the most finely comminuted g-lass from a volcanic explosion. This ash-structure somewhat resembles the rhyolitic or fluidal structure but may be distinguished, as in the present case, by the fact that the angular areas are distinct from the matrix even in polarized lig-ht. The fragments although originally glass, are now cryptocrystalline, with a somewhat zonal arrangement which suggests that these triangular areas represent cavities between rounded grains rather than the g-lass fragments themselves. However, such zonal crystallization has been ob- served by Arnold-Bemrose 2 in the altered lapilli of Carboniferous tuffs and Mugge 3 shows that it is to be expected in the devitrifi- cation process, since the pseudomorphism would proceed exactly parallel with the outer boundary in an amorphous glass. Further- more, a study of the arrangement of the triangular, crescent, and filament-like areas readily shows that the complexity can hardly be explained as due to interstices between grains. One section ex- hibits an included fragment of a lava with a quartz phenocryst, associated with this typical ash-structure. (Plate I, Fig. 4). These fine tuffs, therefore, are to be considered as derived from a rhyolitic lava. Associated with the quartz-porphyry are tuffs, yellow to light gray in color, and very felsitic but including fragments of slate 1. Neues-Jahrbuch f. Min. etc. Beil-B. 8., p. 648. 2. Quart. Journ. Geol. Soc. London, vol. 50., p. 60. 3. Neues-Jahrbuch f. Min. etc. 1896, vol. I., p. 79. 40 GEOLOGY OF THE FOX ISLANDS. and dark hornblendic or chloritic material. Quartz and feldspar grains can be observed and portions of the surface appear sericitic. The clastic character is even more evident when the rock is examined microscopically. Corroded quartzes with embayments, angular pieces of clear feldspar, both orthoclase and twinned plagioclase, short prismatic zircons, and shreds of muscovite are included, but the fragments that constitute the greater part of the tuff are different in character. These are very angular even approaching the concave forms of the smaller glass particles de- scribed above, and consist of quartz or feldspar with undulatory or roughly spherulitic extinction. A marked difference can be ob- served also in ordinary light between these fragments and the others which are plainly fractured phenocrysts, and secondary crystallization seems to be indicated. Chlorite and calcite are prominent constituents of the matrix and the latter mineral occurs also as an alteration product of some of the feldspars. This rock, then, may be termed a quartz-porphyry tuff and the angular in- clusions of quartzite and slate, doubtless also represent the ex- plosive action at the volcanic vent. Certain of the purple tuffs on Iron Point are quartzitic in appear- ance and from their perfect banding suggest sedimentation where wave-action could sort the material. The sub-angular fragments are wholly of volcanic origin, as in the other tuffs, but the rock represents a phase more nearly allied to the more common classes of sediments. An overlying conglomerate so closely resembles well-washed conglomerates of common occurrence as not to de- serve to be considered a pyroclastic. SEQUENCE. As data for the determination of the history of the vulcanism of this area, observations on the stratigraphical se- quence of the different volcanic products are necessary. A uni- form succession comparable with that of sedimentary rocks in a limited area is hardly to be expected in a volcanic series, and on comparing several geologic sections of the North Haven volcanics, differences in the sequence will be noted. While, however, detail- ed areal study emphasizes these local differences, a general se- quence is observed. The lowest members of the series, the dark gray pyroxene- THOROUGHFARE VOLCANICS. 41 andesite and andesite-porphyry, are most prominently exposed on Ames Knob, and to the south of this hill the sequence can be traced without difficulty. Above the andesitic lavas, which are locally very vesicular, occurs the volcanic conglomerate, a well de- fined deposit with a general strike of N. 65 E. The included fragments of the dark gray pyroxene-andesite and porphyry which constitute the greater part of the volcanic conglomerate afford proof of the relative age of the members of the series. To the south and west, this conglomerate grades into breccias, which show both the tuffaceous and flow characters. The red basaltic andesites overlie the conglomerate and breccia and are in turn followed by the more acid hornblende-andesite. At this point in the series the lavas become less prominent and on the shores of the Thoroughfare the tuffaceous rocks are exposed, acid and basic tuffs being interbedded. The best-bedded of these tuffs strike N. 85 W. and dip 62 to the south. A short distance to the east, the amygdaloidal flows, with amygdules giving a strike of N. 80 W., occur above the hornblende-andesite and are often interbedded with the tuffs. Just west of North Haven village, occur the more acid mem- bers, the quartz-porphyry tuffs and small amounts of the quartz- porphyry, while on the shores of the Thoroughfare, at the village, the yellow quartz-porphyry is prominently exposed. This more acid lava is not as widely extended as the andesitic types. In the eastern part of the main area of the volcanics, on the shores of Waterman's Cove, much the same succession of red and gray andesites, andesite-porphyry, and breccia is found, with some quartz-porphyry, which here is doubtless instrusive. On Iron Point, portions of the series are exposed which would overlie the rocks of the Ames Knob district. These are well bedded tuffs, striking N. 80 K. and dipping 58-61 S. With these tuffs occurs the medium-grained conglomerate with well-rounded pebbles. It is to be noted that throughout the principal area on North Haven, the bedding of the pyroclastics is quite constant. Across the Thoroughfare on the extreme northern portion of Calderwood's Neck, the same sequence is found, the breccias over- laid by tuffs and the conglomerate, the series striking N. 40 W. and dipping 82 S. At this locality the section is continued even 42 GEOLOGY OF THE FOX ISLANDS. * higher than on Iron Point, for breccias overlie these bedded rocks, and at the entrance of Seal Cove, well-bedded tuffs occur above the breccias with approximately the same strike, and a dip of about 50 S. This change in strike from that found to the north- west on Iron Point must be accounted for by folding, with possibly some faulting. To the east on the shore of Carver's Cove the lower of the bedded tuffs and the conglomerate are found, striking N. 5-25 E. and dipping about 35 to the west. Again folding and faulting are necessary structural elements in the explanation of the distribution of these beds. A short distance to the south, two parallel faults can be seen in the shore cliff. The strike of these vertical faults is N. 45 W. and the displacement, which is rela- tively downward on the southern side, amounts to somewhat over fifty feet in one of the faults, while the amount of throw in the other cannot be measured, since the lavas and breccias, which overlie the bedded tuffs, afford no datum-plane. The actual ob- servation of displacement at this point strengthens the probability of unexposed faults existing to the north. On the western side of Vinal Haven Island, the upper part of the series is again exposed, consisting of andesitic lavas, with tuffs and breccias. The well-bedded tuffs in this part of the area strike N. 18-25 E. and dip 62-67 to the southeast. At Brown's Head, the upper part of the series is in contact with the Vinal Haven acid volcanics to be discussed later. The crescent shape thus expresses the structure of the North Haven series in its principal development as well as the areal dis- tribution. In the area of similar volcanics to the east, along the shores of the Little Thoroughfare, the most prominent rock-type is a flow- breccia, with which occur tuffaceous rocks of varying character. There is nothing in the sequence, however, to warrant more than a correlation in general character, with the rocks of the principal area. AGE. The age of the North Haven volcanics can be deter- mined only by their relation to the Niagara sediments. In the Ames Knob locality the volcanic rocks overlie the upper members of the Niagara series with about the same strike and dip. The THOROUGHFARE VOLCANICS. 43 lack of any marked unconformity must be considered as showing- that the volcanic outbreaks followed the deposition of Niag-ara sediments, before the tilting- of those sediments to their present position. The red shales of the Niag-ara series quite probably represent old muds, which received considerable amounts of vol- canic dust, thus marking- the commencement of, the intermittent volcanic activity. On Stimpson's Island, however, there is geological evidence which bears more directly upon the ag-e problem. A strip of lime- stones and quartzites less than a hundred feet in width occurs within the area of volcanic rocks. The strike and dip of these rocks, N. 60 W. and 30 to the northeast, and the thinness of the section prevents exact correlation with the Niagara series on North Haven, but the rocks very closely resemble the latter sedi- ments and the only fossils found, traces of brachiopod shells and a crinoid stem, are also similar in appearance to the more plentiful org-anic remains of the other series. These sediments, therefore, were deposited under similar conditions and, most probably, in the same period as those which have been proved to be Niag-ara, althoug-h not necessarily ever closely connected with the other area. Volcanic rocks, gray lavas, underlie these quartzites and limestones, and tuffs and breccia overlie them, so that they must be reg-arded as having- been deposited during- a time of cessation in volcanic activity, a break not recorded in this manner at any other locality on the Fox Islands. It seems justifiable, therefore, to believe that the volcanic activity beg-an in Niag-ara time. CONDITIONS OF VOLCANIC ACTIVITY. . In any attempt to picture the nature of the vulcanism that characterized this area in Niag-ara time, the position of the volcanic center should be determined. It would be of interest to know the vent through which the lavas of this series were poured forth and the f rag-mental material ejected. However, it is probable that the area of North Haven volcanics de- scribed above is not more than a frag-ment of the original area and that through the processes of erosion, many times repeated, .the greater part of the volcano has been cut away. In the study of this limited area, no definite comparisons of the relative thicknesses of the volcanic series can be made, such as would point to the posi- 44 GEOLOGY OF THE FOX ISLANDS. tion oi the volcanic center. The breccias so prominent among- the North Haven rocks are somewhat agglomeratic in character, but their distribution is too general to indicate accumulation within a crater. Topographically, Ames Knob is suggestive of an old vol- canic plug, but the relations of the rocks at this locality do not justify such a view. Thus, the evidence obtained in the study of these old volcanics does not warrant even conjectures as to the site of the vent. The rock-characters of these volcanics are somewhat indicative of the conditions under which these rocks had their origin. The character of the crystallization in the andesites is expressive of consolidation at the surface and the flow structure shows that these lavas consolidated while in motion. The amygdaloidal character of several of the lava-flows resulted from the expansion of escaping gases in the molten mass, while the intercalation of beds of tuffs with the amygdaloids shows that, at times, these gases became explosive and the eruption of lavas was interrupted by the ejection of lapilli and volcanic dust, which were deposited upon the flanks of the volcano. The flow-breccias which have been described, may be considered as having their origin in the mingling of two lava-streams of different eruptions. The cooling crust of a lava-flow becomes broken into angular blocks and these may be caught up in a later stream of lava, which in its lower portions may be composed almost wholly of these angular frag- ments. In these flow-breccias, the included blocks and the matrix may be quite different in character, masses of the gray andesite being usually embedded in the red andesite, which often shows flowage around and between these included blocks. In other cases, the lava blocks have been cemented by tuff material or by volcanic mud. The volcanic conglomerate is composed of less angular fragments and more varieties of lava are represented within a given mass of the conglomerate than in the breccia. These differences may result from the transportation of the fragments by running water for limited distances down the slopes of the volcano. Just as the tuffs furnish the most indubitable criteria for establishing the volcanic nature of the series, so they also afford the best evidence with regard to the topographical features of this part, at least, of the old volcano. Many of the tuffs seem to have THOROUGHFARE VOLCANICS. 45 originated simply as sub-aerial deposits upon the gentle slopes of older lava-flows, but in other cases, the deposition may have been in a body of water. The sorting- of the different sizes of grain and the banding 1 of successive beds is as perfect as in any thin- bedded shale or sandstone. However, the fragments of minerals and lavas are as angular as in the sub-aerial deposits, so that they could not have been subjected to long continued wave-action. On Iron Point, where the bedding of the tuffs is best exhibited, the beds of finer tuffs are often eroded on their upper surfaces, the succeeding beds showing slight local unconformities. Such erosion-breaks in the beds are only a few feet in extent and have resulted probably from wave-action along the tide-flats. The con- glomerate which was later deposited has well-rounded pebbles and is of the character of an ordinary sediment which thus marks an interlude in volcanic activity, although the conditions do not seem to have been favorable for marine life, as was the case when the sediments occurring within the volcanic rocks on Stimpson's Island were deposited. The pyroclastics and other sediments show that this volcanic area was near the sea-level and that oscillations of level were more or less frequent in the period during which the volcano was in eruption. The activity did not close with the eruption of the rocks which have been considered as the North Haven series; but in the later phases of the vulcanism, the products were of a some- what different character and will be discussed in the following section. VINAL HAVEN ACID VOLCANICS. DISTRIBUTION. The area of the acid volcanics may be describ- ed as a rough circle fitting- into the crescent of more basic vol- canics, the horns of which extend southward, the diameter of this nearly circular area being- about two miles. Extremely g-ood sections of the series are afforded by Perry's and Crockett's Coves which extend well into the center of the area. The topography, like that of the other volcanics, is quite bold, especially in the south- ern portion, where the elevation of 210 feet, the greatest on the Fox Islands, is attained. DESCRIPTION OF ROCK-TYPES. The volcanic rocks of this area exhibit less variety than those of the area to the north. Fewer types are to be distinguished and described, yet these are of special petrographic interest. They rank among- the most beauti- ful of volcanic rocks and cannot fail to attract attention as exposed on the weathered ledg-es and sea-cliffs. Among- these acid vol- canics, both lavas and pyroclastics are ag-ain represented and four rock-types will be described: the taxitic and the spherulitic apo- rhyolites, the flow-breccias, and the tuffs. APORHYOLITES. The acid lavas which form the greater part of this Vinal Haven series are characterized by flow-structures. No phenocrysts, other than small mag-netite and zircon crystals, ap- pear in any of the specimens examined, and the rock is commonly very felsitic in appearance. In the thin section, the texture is found to be cryptocrystalline in part, and elsewhere to exhibit traces of glassy structures, now more or less obliterated by de- vitrification. In short, the rock is rhyolitic and was originally somewhat glassy, and, therefore, the term aporhyolite has been applied to the type. The two characteristic structural types of these aporhyolites, which may be distinguished meg-ascopically, will be considered separately. The taxitic aporhyolites are those in which the fluxion charact- 46 VINAL HAVEN ACID VOLCANICS. 47 er is most marked. Flow-f elsite was the field-term employed in de- scribing- these rocks, and megascopically as well as microscopical- ly, this lava is striking- in appearance. It is very compact and composed of lig-ht flesh-colored and dark gray or purple layers of var}dng width. Where this banding- is regular, the rock often shows a parallel parting-. There are many cases in which the flow-structure is very irregular, and on the weathered surface of the taxtites the structure is most beautifully exhibited, the flowag-e-lines being- extremely delicate. In the thin section, the taxitic structure is equally apparent. The darker lines are seen to owe their color to masses of black cumulitic grains, marg-arites, and trichites, minute crystallites, all of which, doubtless, have the composition of magnetite, being- often connected with grains of that mineral. The detail of the fluctuation as presented by the arrang-ement of these crystallites is much more delicate than that seen megascopically. The parallelism in the more regularly banded taxites is also well shown in polarized lig-ht, by the varying- degree of crystalliza- tion. The bands in which the trichitic flow-lines are so promi- nent often contain chains of minute spherulites, which will be described later with the other spherulites. More commonly, how- ever, these bands are characterized by a peculiar intergrowth of fine fibers of what seem to be quartz and feldspar, althoug-h the con- stituents are too minute to be determined. This intergrowth may be described most accurately as a web-structure, the long- delicate threads being more continuous in the direction parallel to the megascopical banding. In the lighter band and lenses, the crystal- lization is coarser and less definite in structure, although spheru- Jitic in part. Some areas of quartz and epidote occur, and are doubtless of secondary origin. The spherulites and the web-structure seem to represent inter- growths of mineral fibers which crystallized within the lava, later than the trichites and other crystallites, but before consolidation. Their parallel arrangement may have been determined by chemi- cal or physical differences along definite planes, while the more intricate flow-lines traced by the crystallites are the result of movement in the lava. In certain areas in these rocks, the crystallization is of a more granular character and these rather in- 48 GEOLOGY OF THE FOX ISLANDS. definite mosaics of quartz and feldspar may have originated through devitrification of original glassy portions of the rock. The spherulitic aporhyolites are closely related to the taxitic, and gradations are found between the two types. The spherulites vary in abundance and in size, a few small ones, the size of a pin- head or of a pea, occurring- disseminated through the variegated purple or dark gray rock, or, on the other hand, the spherulites being- in great abundance, often completely covering- the exposed surfaces, and attaining- a size of several inches in diameter. Such a rock is of so unusual an appearance, that it would readily be noticed by even a casual observer. The spherulites vary greatly in form, some being- perfect spheres, others ellipsoids, and often several are united to form irreg-ular nodules. They are usually compact and felsitic, more rarely hollow and lined with quartz. The hollow forms may owe their condition to subsequent solution, and no undoubted lithophysal structures were observed in these old rocks which have so many characters in common with later rhyolitic lavas. Examined microscopically, the spherulitic aporhyolite is found to be characterized by even more delicate arrangements of the trichites and other crystallites. The spherulites, which may be described as complex intergrowths of crystal-fibers, radially ar- ranged, vary in outward form, the sections being circles, segments of circles, and plumes. The radiate arrangement of the fibers is the common characteristic and the different outward forms taken by such intergrowths are equally expressive of the spherulitic crystallization, as has been shown by Prof. Iddings 1 . These spherulites, especially the smaller, are not always to be seen in ordinary light, as they are often wholly independent of the flow- structure, the lines of trichites passing through the spherulites without interruption. In other cases, the spherulites may be readily seen, through a concentric arrangement of bands rich and poor in trichites; or again, the spherulitic forms may be slightly different in color, the finely divided pigment being of a brown tint rather than black, a feature noted in some of the spherulites of Obsidian Cliff 2 . The micro-spherulites and the 1. Spherulitic Crystallization. Bull. Phil. Soc. of Washington, vol. n, p. 459. 2. Iddings: 7th Ann. Report, U. S. G. S., p. 277. VINAL HAVEN ACID VOLCANICS. 49 chain-spherulites found both in this rock and in the taxitic apo- rhyolite are not apparent except in polarized light. In the chain- spherulites there is an approach to the axiolitic type, and it is noted that a group of three black crosses, as seen when the axis of the chain is in parallel position with one of the nicols, resolves itself into two crosses, rather than three, when the section is re- volved 45. These micro-spherulites often compose considerable portions of these aporhyolites, producing- an effect similar to that pictured by Iddings 1 . The constitution of these spherulitic intergrowths is a problem for the solution of which these old rocks do not furnish as satis- factory material as the younger lavas. Yet, some of the charact- eristics of these structural forms can be observed in these apo- rhyolites and deserve description as affording- some degree of proof of the essential similarity of the old lavas to their later equiva- lents. The smaller spherulites are optically neg-ative; while, in the larg-e spherulites, the radiate structure can be resolved into more or less distinct fibers, the optical character of which could not be determined, owing- to the fact that several are usually superimposed, even in the thin section. In the best examples of the spherulitic crystallization on the larg-er scale, these radiating fibers are seen to be embedded in interlocking- grains, relatively larg-e; and the radiate structure can be observed, even in ordinary- lig-ht, owing- to the slig-ht difference in refraction and in color be- tween the fibers and the grains. These grains appear to be uni- axial, in part at least, and most probably are quartz, while the arborescent fibers are feldspar. Such an intergrowth is a form of the micropegmatitic structure, and is undoubtedly the original form of these larger spherulites. There are, however, indications that in some cases, subsequent alteration has replaced the feldspar fibers with quartz, thus changing the micropegmatitic structure to a purely granular mosaic. In one spherulite, this change, al- though only partial, seemed more complete in the outer grains, where only mere traces of the fibers remained, while nearer the center, the arborescent growth was continuous. (Plate I, Fig. 6.) Where the change has been complete, the spherulite is identified simply from the circular outline of the fine mosaic. I. yth. Ann. Report, Plate 17, Fig. i. 50 GEOLOGY OF THE FOX ISLANDS. Different generations of the spherulitic crystallization are in- dicated in some of the sections examined. The spherulitic plumes, where they occur in connection with the more perfect spherulites, are of a darker color and appear to be older, being 1 sometimes en- closed in the later spherulitic growth, or conditioning- the outward form of the larger and lighter spherulite. In one section, all of these plumes are seen to have grown in the same direction, parallel with the flow, and also to have determined sharp eddy-like curves in the flow-lines of trichites. One of the most beautiful of these spherulitic structures is where a long- plume suddenly spreads out into three quadrants of a circle, as seen in section. The matrix in which these larg-er spherulites occur is usually cryptocrystalline, although there are areas of coarser grains of quartz and feldspar, around which there seems to be some flowag-e. These may represent vesicles in the rock, subsequently filled. Allied with these are smaller miarolitic areas of quartz into which project fine laths, optically continuous with the fibers of small spherulites. That the rock was in part g-lassy, is proven by the perlitic cracks which are beautifully shown in some of the sections. These are outlined as lig-hter lines in the ground-mass of a lig-ht brown color, and in polarized lig-ht are seen to have been filled with quartz. In these portions of the rock, the very delicate trichites do not appear to have been affected by the subsequent devitrification of the glass. The above description of the structures which are so character- istic of the taxitic and spherulitic aporhyolites of Vinal Haven can express only imperfectly the fine details which afford to the observer the most conclusive evidence of the practical identity of these rocks with recent acid lavas. An analysis of one of the most typical and seeming-ly least altered spherulitic aporhyolites was made by Messrs. K. W. Ma- gruder and W. A. Jones, and it affords the best evidence obtain- able, as to the rlryolitic character of the magma from which these spherulites crystallized. VINAL HAVEN ACID VOLCANICS. 51 Si O 2 ,'.'-'. -'v : ; ,',Y :.>;* . . 77.28 Ti O 2 .... . . . . . trace A1 2 O 3 \ >. .:' 11.24 Fe 2 O 3 >':, . _''*.;; ' .; . 1.74 Fe O . . . .... 'v >, Y '..- . none MgO . ". ' . . . . 0.21 Ca O . . . .. . . - . . trace K 2 O . -..'.'.- . . , \ . 4.55 Na 2 O . . . .... . . 3.10 p 2 o 5 . , . , <^ v . : . . 0.022 H 2 O (by Pb O) . V . .- . . 1.16 CO 2 . . . . . . 0.17 99.472 FLOW-BRECCIAS. Many of the aporhyolites present traces of brecciation. Meg^ascopically, the rock is similar to the spherulitic lava, but on examination is found to lack homogeneity. In some cases, fragments of the older andesitic lavas are included with those of the acid g-lass, but the matrix is seen to have been a molten lava, thus making- the type a flow-breccia. Under the microscope, the mass of the breccia is observed to be characterized by a flow- structure, much more irregular than in the taxites. One specimen affords a beautiful example of the structures seen in obsidians, ribbons and filaments of the trichitic g-auze being- twisted into com- plex patterns. (Plate I, Fig-. 5.) Traces of perlitic cracking- also give evidence of the originally glassy state of the rock. The fragments which are similar to the enclosing- material usually appear to have been brecciated in situ. Spherulitic plumes are broken across and the parts slig-htly displaced, and in all cases the spherulitic crystallization preceded the brecciation. Other frag-ments possess a flow-structure sharply contrasted with that of the enclosing- lava, and in one, there has been a marked shearing without rupture 1 . Devitrification has made these glassy rocks quite holocrystalline, but in the rock with the delicate obsidian structures, there has been very slight devitrification. TUFFS. The tuffs of this series are in the main, much more i. Figured by Dr. Williams, op. cit, p. 23. 52 GEOLOGY OF THE FOX ISLANDS. compact than those of the North Haven series, but like them vary in color from dark purple to light pink. In grain, they present the same variation from a coarse breccia to a rock composed of the finest particles and showing* the ash-structure. The larger frag- ments are mostly of the banded and spherulitic lavas, but in some instances, andesite fragments are very plentiful. Crystal frag- ments also occur in the finer tuffs. Epidote occurs in these tuffs, partly in the role of cement, and partly as an alteration product. This mineral was also found in the lavas of this series, but in less amount. Chlorite and calcite also occur as secondary minerals, the latter only in tuffs with andesite fragments. SEQUENCE. The structure of the Vinal Haven series is com- paratively simple, the lavas and tuffs forming a basin, which over- lies the tilted andesitic volcanics. Where their bedding can be determined, these acid volcanics dip at an angle of 45 or over toward the center of the area, except in the southwestern part where the dip is more to the southeast. On the northwestern side of the area, the lower members of the Vinal Haven series are found in contact with the more basic series, and the relations seem to be wholly comformable. To the northeast, the contact is hid- den, for the most part, by the dike of diabase. Tuffs constitute the basal members of the series and the separa- tion of many of these tuffs from those of the underlying series is difficult. Above these tuffs the general sequence is taxitic apo- rhyolite, overlaid by the spherulitic aporhyolites and flow-breccias, with the other tuffaceous beds as probably the highest members of the series represented. In the taxitic lavas, the lamination gives approximately the same strike and dip as the underlying tuffs. The spherulitic aporhyolites are best exposed in the central por- tion of the area, on the shores of Perry's and Crockett's Coves, and especially along the shore south of Brown's Head. CONDITIONS OF VOLCANIC ACTIVITY. The volcanic eruption of these acid lavas and tuffs immediately succeeded that of the North Haven series. Indeed, it seems probable that the ejection of the acid lava began before that of the andesitic closed. Within the VINAL HAVEN ACID VOLCANICS. 53 North Haven series, acid members were found, overlaid by more basic types of lavas and pyroclastics, and again in the upper tuffaceous beds of what has been termed the North Haven series, there is a mixture of material. Thus, at Brown's Head, a coarse tuff-breccia composed of large fragments of andesites, was found to contain a few angular pieces of the taxitic and spherulitic apo- rhyolites. These must have been derived from some lava-mass older than the similar lavas now exposed to the east. As far as a difference of age is concerned, the separation of the North Haven and Vinal Haven series is purely arbitrary. However, it seems probable that the two types of lavas were erupted from separate although neighboring vents, the rhyolitic eruptions beginning while the andesitic were in their later stages. The taxites which directly overlie the deposits of fragmental volcanic material, are a striking type of lavas. The lamination which characterizes them is expressive of the conditions of their eruption and consolidation. These layers differing in color and in degree of crystallization have resulted from the flow of the viscous lava, spreading over a surface horizontal or only slightly inclined. Where this lamination is most regular and the layers quite thin, a comparatively greater distance from the point of origin may be inferred 1 . Loewenson-Lessing in calling attention to this taxitic type of lavas has considered them as more closely allied to pyroclastics 2 . The fact of the existence of a gradual passage between the lavas and the pyroclastics may well be emphasized in view of their common origin, yet the explanation of the taxites as clastic in character seems unwarranted. A simultaneous disintegration and cementation within a molten rock-mass may be conceivable, but is hardly sufficient ground for fixing the clastic character of a banded lava. Brecciation can take place in an already consolidat- ed rock, but the term does not seem applicable in describing differentiation of a magma into layers of slightly different chemi- cal composition and physical condition. 1. Iddings: yth Ann. Report U. S. G. S., p. 260. 2. Tschermak's Min. u. Petrog. Mitth. 1888, p. 528. Bull. Soc Beige cl. Geol., vol. 5, p. 103, 1891. 54 GEOLOGY OF THE FOX ISLANDS. The macro-spherulites first appear in the upper part of this taxitic aporhyolite, where they are confined to the lighter layers. The conditions determining- this kind of crystallization are not easily understood. The arrangement of the spherulitic growths in these old rhyolitic lavas point to the fact that whatever agent con- tributed to the crystallization, it was often confined to quite definite planes, and that, in some cases, the conditions favoring the process were not limited to a single period in the consolidation of the lava. Some of the earlier plumes seem to have been formed before the lava had ceased flowing-, while, in other cases, the delicate trichitic lines of flow maintain the same course within the perfect spherulites as in the rest of the rock. These arborescent intergrowths of .quartz and feldspar seem to have crystallized very rapidly in a glassy lava, which had come to rest and become nearly solid, superheated water and steam, doubtless, being im- portant factors 1 . The flow-breccias appear to be locally developed within the central area of spherulitic aporhyolites. In part, these breccias probably originated as lava flows including fragments of older lavas, as was the case in the breccia so prominently developed in the North Haven series. In other cases, the brecciation seems to have been confined to the lava itself, certain strains to which the cooling mass was subjected being sufficient to break up those parts which had consolidated. This process seems to have been of a different character from the breaking up of lenticles of flow which would give an ataxitic or rhyolitic structure, since here the breccia- tion took place when the cooling glass was sufficiently stiff to break into extremely angular fragments. This brecciation also followed the spherulitic crystallization and affords the best proof of the original nature of these spherulites. There are cases, how- ever, in which the dislocation may have taken place in the la\*a when in a less viscous state, as in the case cited above, where the flow-structure shows the shearing. A similar brecciation in a very stiff lava has been observed in the spherulitic rock of Silver Cliff, Colorado, .where the spherulitic crystallization was subse- I. Cross: Constitution and Origin of Spherulites. Bull. Phil. Soc. of Washington, Aol. ii, p 433- THOROUGHFARE VOLCANICS. 55 quent to the dislocation. Cross 1 considers the lava to have been broken by sudden and violent volcanic shocks, the pieces possess- ing- sufficient plasticity to perfectly coalesce again. In the Vinal Haven occurrence, a greater degree of plasticity must be inferred in many cases, since the separated fragments are cemented by glassy material, which shows fluidal structures. These flow- breccias are especially indicative of the vicissitudes in the process of consolidation of a lava-flow, and therefore possess more than mere petrographical interest. They seem to be a type of volcanic products, which much more truly occupy the middle ground be- tween lavas and pyroclastics than do the taxites, to which refer- ence was made above. The tuffs occurring- in the upper part of the Vinal Haven series differ from those in the basal portions in that they contain no frag- ments of the andesitic lavas. Their purely acid composition in- dicates a cessation of the andesitic eruptions, which opened the Niag-ara vulcanism. How much more of rhyolitic lavas and ejecta- menta were erupted cannot be determined owing- to the granitic in- trusion to the south of this area and to the subsequent erosion of any hig-her members of the series, that might have been left to the north of the granite contact. :. Op. cit., p. 433. DIKES. DESCRIPTION. As one walks along- the shores of North Haven and of the northern part of Vinal Haven, numerous instrusive dikes attract attention. These may stand out as walls from the low cliffs of less resistant rock, or more often, they are marked by deep cuts extending- into the bank. In the interior, the dikes are always resistant to weathering- processes, and thus form the axes of the hills and ridges. The relative resistance of country rock and dike thus varies, but whether on the shore or inland, these dikes form conspicuous topographic features. Of these dikes, two types are distinguished, the one dark and usually finely crystalline, the other lig-ht colored and felsitic as a rule, althoug-h in a few cases plainly crystalline. This division will be seen to be an important one and the two types will be de- scribed separately. The acid dikes rang-e from microgranites or quartz-porphyries, in which the porphyritic constituents may be easily distinguished, to very fine-grained rocks as felsitic in appearance as the acid lavas of Vinal Haven. The coarser of the quartz-porphyries might be mistaken for granites, but on close examination it is noticed that the quartz and feldspar occur in well defined crystals. Under the microscope these phenocrysts are seen to be beautifully idiomorphic, the quartz showing- fine examples of magmatic cor- rosion. The orthoclase is intergrown with albite, and plagioclase also occurs both in tabular crystals and in the ground-mass. Muscovite is the other principal constituent, and the porphyry seems rather acid, only a few shreds of hornblende being- noted. The ground-mass in the coarser type is microgranitic, but in other of the quartz-porphyries, it shows a tendency to micropeg-matitic structure, especially as borders to the quartz phenocrysts. Pe- culiar sheaf -like feldspars occurring- in the ground-mass also serve as centers for such intergrowths. The felsitic rock is of a light green color and is found to have a 56 DIKES. 57 cryptocrystalline ground-mass, containing* a few very small pheno- crysts of feldspar, mostly altered to calcite. This rock occurs in narrow dikes and is probably the equivalent of the coarser por- phyry of the larger dikes. All of these acid dike-rocks are of a light red or yellow color on the weathered surface. The basic dike-rocks also show considerable range in grain. The more common type is plainly diabasic, in one case being* very coarse-grained, but other rocks are found which are of the same dark gray color and wholly aphanitic. Examined microscopically, the former type is found to be composed of large plates of augite, colorless or slightly tinted in the thin section, and of feldspar laths of an earlier crystallization than the augite. The typical diabase structure is modified somewhat in a few specimens by a porphy- ritic development of the feldspar. Biotite was observed in some specimens, but only in small amounts. These diabases all show more or less alteration, calcite and chlorite being- the principal products. * t The finer-grained of these basic rocks are to be grouped with the diabases from their similarity of mineralogical composition. The diabase-structure, however, is not as marked, since the augite occurs in smaller grains while the mass of the rock is a felt of feld- spar laths. A few larger feldspars occur but these are full of in- clusions and badly altered. One characteristic of this type is the presence of quartz, in rounded grains relatively very large as com- pared with the other constituents, and abundant enough to be noticed in the hand-specimen. These quartz grains are surround- ed by resorption-rims of augite in short radiating prisms, partially altered to chlorite. As shown by Diller in his study of the quartz- basalt of California 1 , the presence of this envelope of pyroxene ex- presses only the composition of the basic magma which corroded the quartz. That the quartz is indigenous to the diabase which forms these dikes seems more probable also from the wide and uni- form distribution of the quartz grains. DISTRIBUTION. Throughout these islands, the basic dikes are more numerous than those of acid composition, except possibly to Bull. 79, U. S. G. S., p. 28. 58 GEOLOGY OF THE FOX ISLANDS. the north of Mill River, where the interbedded sheets of quart- porphyry occur. In the southern part of North Haven, small dikes of the diabasic rocks are especially common, occurring" as single dikes which can be traced for long 1 distances, or as groups of smaller branching- dikes. No system could be traced in these dikes, but the strike is commonly between northeast and north- west. The acid dikes occurring- on North Haven are also usually small, the widest being- twenty feet in width. On Vinal Haven, the dikes are much larg-er, those of the diabasic character being- sufficiently wide to be mapped in two cases. One of these is the dike which extends from near the head of Seal Cove north to Perry's Cove and thence probably across to Hopkins' Point and along- the southern shore of the Thoroug-hfare. The intrusive nature of this mass of diabase is proven by the frag-- ments of the overlying- acid volcanics included at one locality and by the numerous apophyses which cut the quartzitic schists to the east. In its northern and western extension, the relations are not so plain, since the contacts are concealed, but the manner in which this rock cuts across the strike of the well-bedded tuffs seems to indicate a true intrusion. The rock in this dike is in great part that characterized by the quartz grains, althoug-h to the south near the contact with the granite, it is a typical diabase. It is in- teresting- to note that the same quartz-bearing- phase occurs in the smaller dikes which cut the volcanics on the opposite side of the Thoroughfare. DIABASIC AND GRANITIC INTRUSIVES. AREAI, DISTRIBUTION. In addition to the dikes of igneous rock which have been injected into the older rocks, there are larger in- trusive masses of granite and other holocrystalline rocks. These lie to the south of all of the areas discussed thus far, and consti- tute the greater part of the island of Vinal Haven. Of these rocks the granite is the more important in areal extent, and de- serves special mention also from the fact that it is the basis of the most of the industry of the southern island. The northern ^ boundary of the granite area is a line running from about midway on the eastern shore of Calderwood's Neck due west to Seal Cove and thence southwest to the northern point of Leadbetter's Island. To the south of this contact of the granite with the quartzitic schists and acid volcanics, Vinal Haven is com- posed wholly of granite, with the exception of two quite large areas of darker rock, the "black granite" of the quarry-men. One of these areas comprises the southern extremities of the main island as well as Lane Island and part of Green Island. The other and the larger of the two is on the eastern side of Vinal Haven and may be called the Barley Hill area, from the more prominent hill within its limits. The western boundary of this area is in the low plain occupying the central portion of the island, and thus the contact with the granite is hidden. 1 Topographically, the area of these holocrystalline intrusives is characterized by the same rugged shores as the volcanic areas. The hills are mostly bare and rounded knobs and have gentle slopes, except on the western side of the island, where hills of light granite rise from the shore that are quite imposing as one sees i. In this geological study of the Fox Islands, the time given to the field-work did not permit a detailed study of the granite area, and other small areas of the diabase or diorite may remain unmapped. It is believed, however, that such possible omissions will not affect the geologic relations, as here presented. 59 60 GEOLOGY OF THE FOX ISLANDS. them from the Bay. The central portion of the area is a large plain, swampy in part, where the glacial drift is doubtless deeper than over any other portion of the Fox Islands. GRANITE. The principal type of the Vinal Haven granite is of a light gray color, with a slight pink tint, and of a rather coarse texture, but very uniform both in color and in grain. The con- stituents, which are prominent in the hand-specimen, are the clear quartz, in good sized areas, the feldspars, white and flesh colored, the latter being more abundant, and the black flakes of biotite. Near its contact with the "black granite" the granite is often rather porphyritic, the feldspars being larger and more idiomor- phic. Another type of the granite, which is found on the eastern side of the island; is fine-grained, although otherwise similar to the principal type. In the thin-section, these rocks are found to be typical biotite- granites, with the normal sequence of crystallization of their con- stituents. The orthoclase is more abundant than the plagioclase, shows Carlsbad twinning, and is often in perthitic intergrowths with albite. The quartz usually occurs in large allotriomorphic plates, except in the porphyritic phase, where the quartz blebs are rudely idiomorphic,- showing an interruption in their later stages by the crystallization of the ground-mass. Biotite occurs in large plates in the coarser variety and in small shreds in the finer-grained. Hornblende is an accessory constituent in vary- ing amounts, although in the porphyritic variety, it is in almost equal amounts with the biotite and is more idiomorphic, with the usual twinning. Magnetite and apatite are the other constituents. The feldspar and biotite show slight traces of alteration, but generally the rock appears very fresh and shows no traces of hav- ing suffered any dynamic action. The granite is comparatively free from the segregation patches of dark constituents; but within the granite mass, there are numerous aplitic veins, the "salt-rock" or "salt-horse" of the quarrymen. The coarser of these rocks much resembles the finer- grained granite in general appearance, and is found to show the same texture and composition except that hornblende is equally prominent with the biotite, intergrowths of the minerals occurring DIABASIC AND GRANITIC INTRUSIVES. 61 with the prism of the former parallel with the basal pinacoid of the latter. Sphene is an accessory constituent of this rock. The finest-grained of these granitic veins is a rock, composed of a mosaic of rounded grains of feldspar and quartz, the plagioclase in a few cases being- somewhat idiomorphic; and muscovite occurs as the other principal constituent. These "veins" are of the class termed "contemporaneous", and are to be regarded as belonging- to the same general intrusion as the rest of the granite mass, rep- resenting material injected from the molten portions below into fissures in the already consolidated granite. This granite has been quarried for nearly a half-century; and within the past few years, the Vinal Haven quarries have been the most extensive in operation in America. The principal quarry is the Sands Quarry where the coarse-grained granite occurs. This granite is also quarried at numerous points on Vinal Haven and the smaller islands, the most important of which is Hurricane Island, where the rock is especially well adapted for economical quarrying. The finer-grained stone is quarried at the East Bos- ton and Roberts Harbor quarries. Paving-blocks can be easily quarried at almost any point in the granite area, and numerous smaller quarries have been opened for this purpose. The value of this granite lies in the easiness with which it is quarried and worked, the fine polish which it takes, and the proximity of the quarries to good harbors. Mr. G. P. Merrill, Curator in the National Museum, illustrates the capabilities of the Vinal Haven quarries by the statement that on a visit to these quarries in 1883, "he was shown the remains of a huge block of granite 300 feet long, 20 feet wide, and varying from 6 to 10 feet in thickness, that had been loosened from the quarry in a single piece and afterward broken up. The largest block ever quarried and dressed was the General Wool monument, now in Troy, New York, which measured when finished, 60 feet in height by 5^ feet square at the base, or only 6 feet 7 inches shorter than the Egyptian obelisk now in Central Park 1 ." DIABASK AND DiORiTK. The "black granite" of the first area mentioned, that forming the southern part of Vinal Haven, is an i Stones for Building and Decoration, 1891, p. 189. 62 GEOLOGY OF THK FOX ISLANDS. olivine-diabase. In a few localities, the rock is coarse and almost gabbro-like, the plates of purple augite being* equally prominent with the feldspars. Usually, however, the diabase is fine-grained, dark gray when fractured, but black on a polished surface. Laths of feldspar can be distinguished although quite small. Micro- scopically, the rock is seen to have the characteristic diabase-tex- ture. The feldspars are of a tabular habit, with polysynthetic twinning* by the albite law often combined with the pericline, and the tendency to idiomorphism marked. In composition, the feld- spar is generally an acid labradorite. Augite of a faint pink tint is the constituent second in importance, occurring- in rather large plates enclosing- the feldspars. Olivine is abundant in colorless, rounded grains, only slig-htly altered and is the oldest of the es- sential constituents. Mag-netite occurs in irreg-ular growth-forms and aggregates of grains. Small amounts of biotite are found as- sociated with the mag-netite and olivine, and also some green horn- blende associated with the aug-ite. In the Barley Hill area, the rock is of a slightly different char- acter. Megascopically it presents phases not seen in the other area, and in the thin section, the differences are not less marked. In short, the Barley Hill rock shows a transition, both in texture and in mineralogical composition from a diabase to a diorite. The type nearest the olivine-diabase just described, much resembles it in the hand-specimen, although somewhat less fresh in appear- ance and slightly finer-grained. Examined microscopically, it shows the diabasic structure, although less perfectly than the oliv- ine-diabase. Olivine does not occur, while brown hornblende and biotite are equally prominent with the augite. The hornblende and augite occur together in complex intergrowths, and also in separate plates. Biotite is usually associated with the magnetite, which is more abundant than in the olivine-diabase. The other rocks of this area which have been examined are quite different megascopically, being more feldspathic and thus lighter in color, while the texture is more granular than diabasic. Biotite becomes a prominent constituent. In the thin section, the feldspar is seen to be more acid than in the diabasic types, and the pyroxene has become of less importance as a constituent. It occurs in intergrowths with the biotite and hornblende and is DIABASIC AND GRANITIC INTRUSIVES. 63 apparently older. The biotite is the most important of the ferro-mag-nesian constituents and often occurs in quite idiomorphic plates. Its association with the mag-netite is again noticeable. The green hornblende is also often idiomorphic in the prismatic zone, althoug-h irregular shreds also occur. Both hornblende and biotite show a tendency to alter to chlorite. Quartz occurs in relatively small amounts, although in larg-e irregular areas, being* the last constituent to crystallize. Apatite is very plentiful in very long- slender needles, which often extend throug-h several different constituents. Titanite is found associated with the biotite. The intergrowths of aug-ite and hornblende in these rocks seem to be the result of simultaneous crystallization of the two minerals and not of a secondary paramorphic chang-e of the pyroxene. The fact that the biotite is intergrown with the aug-ite in a similar way shows that the complexity of the intergrowth does not nec- essarily imply the derivation of the hornblende from the aug-ite. The occurrence of idiomorphic hornblende is sufficient argument for part, at least, of the hornblende being- original. The view that seems sustained by the microscopic study is that these rocks represent a series of rock-types, the result of a differentia- tion of one mag-ma. One end-member, which seems to characterize the more southern area, is an olivine-diabase, while the Barley Hill rock is principally a quartz-mica-diorite, rich in hornblende and containing- varying- amounts of aug-ite. While the two types were easily distinguished in the field, no relations were observed other than those of transition from the one type to the other. Both the diabase and the diorite have been quarried to some ex- tent. The hardness of the olivine-diabase interferes with its g-en- eral use, yet this stone is well fitted for monumental purposes, since it is of uniform color and takes a hig-h polish. Were the de- mand sufficient, it could be economically worked at many points along- the southern shore of Vinal Haven. RELATIVE AGE. Along- the shore north of the Reach, there is a peculiar ming-ling- of the two types of intrusives. The hills which form the long- peninsula are composed of olivine-diabase, but anyone standing- on the deck of a steamer passing- throug-h the 64 GEOLOGY OF THE FOX ISLANDS. Reach cannot fail to observe that this dark rock is striped and checked with lighter bands, in a manner almost fantastic. The diabase is cut by sheets and dikes of granite, in such numbers that the darker rock almost loses its continuity in this labyrinth. Both along this shore and at places on the western hill, the old contact surface of the two rocks seems to have been just reached by erosion. Some of the larger dikes of granite can be traced for a considerable distance into the diabase area, and at the contact, the granite is of a porphyritic character, but becomes more granular away from the contact. These relations afford most conclusive evidence as to the granite being a later intrusive than the diabase. At other localities, in this area, as well as in the Barley Hill area, the relations are similar, and there can be no doubt but that the granite is younger than the diabase and diorite. RELATIONS TO DIKES. The general mineralogical similarity of these intrusives to the two types of dikes is suggestive. In both acid and basic types of dikes there are phases which approach structurally as well as mineralogically the granites and diabases to the south. The field relations also fail to justify a sharp distinc- tion between the two occurrences. In the case of the acid dikes at the head of Seal Cove and of Long Cove, the connection with the granite mass can be traced, and a similar relation with the quartz- porphyries north of Mill River is most probable. Wherever the two types of dikes are observed in contact, the acid is the later; and the occurrence west of the mouth of Long Cove, where the granite cuts the large diabase dike, also confirms the hypothesis that the dikes and other intrusives are connected genetically, the more basic type preceding in time the granitic intrusion, with its accompanying acid dikes. TIME AND CONDITIONS OF INTRUSION. The evidence afforded by the apophyses from the granite mass cutting the quartzitic schists and the acid volcanics has already been mentioned. The younger age of granite has also been shown by an angular frag- ment of the characteristic acid tuff being included in the granite near the contact. The determination of the age of the diabase and diorite forming the large areas to the south depends upon the correlation of these with the basic dikes. According to this view, DIABASIC AND GRANITIC INTRUSIVES. 65 the diabasic and dioritic intrusives are younger than the rocks which lie to the north of the granite contact. Although there is no doubt of the intrusive character of the granite, yet a definite determination of the age of the intrusion is somewhat conjectural. The holocrystalline character of the granite is usually believed to express slow crystallization under conditions essentially plutonic. This hypothesis of a covering- at the time of the intrusion of the granite magma necessitates the presence of thousands of feet of sediments or volcanic deposits of later age than those now represented on these islands. The probability of the tilting- of the older series to their present posi- tion before the granite intrusion is also to be considered. The Calderwood's Neck schists and their supposed equivalents to the west show no constant relation of strike and dip to the granite contact, and the g-eolog-ic structure of the mass of acid volcanics, so far as it is evident, does not seem to have been determined by the granitic intrusion but rather to antedate it. Furthermore, it is probable that time must be allowed for the diabasic intrusions between the eruptions of the acid lavas and the intrusion of the granite, and the former intrusives doubtless also consolidated beneath a considerable thickness of rock. Therefore, if the sedi- mentary and volcanic rocks described on the preceeding pages are of Niagara and pre-Niagara age, a considerable interval of post- Niagara time must have elapsed before the intrusion of the granite. ALTERATION. CHARACTER OF ALTERATION. The Fox Island volcanics have suffered changes since their consolidation as lavas and deposition as tuffs, and to explain this subseqent history is one of the phases of the geologic problem. In the long- periods of time since their eruption, these volcanic rocks have had a varied history. They have been buried by later sediments, cut by igneous rocks at least two distinct times, uplifted by mountain-making- forces, and exposed to atmospheric agencies. All the changes which express the effect of these varying- conditions may be considered under the term al- teration, which thus comprises the work of processes acting- under surface conditions and those of greater depth and less superficial character; that is, alteration includes weathering- and metamor- phism. 1 Both processes have contributed to the final result, and a sharp line of separation between the two cannot be drawn, and indeed would be unnatural, since, in nature, there is perfect transi- tion from the one set of conditions to the other. STRUCTURAL AND MINERALOGICAL. The chang-es which the rocks of the Fox Islands have suffered are of two kinds, structural and mineralog-ical. The former is of a metamorphic character and referable to dynamic action which caused a crushing- of the rock. Dynamic metamorphism has been important only in the case of the old diabase flows which have been converted into the green- stone schists of North Haven. Here the effects are the production of a schistose matrix around columns of compact rock, and the fracture of the mineral constituents. The absence of any similar dynamic metamorphism in the case of the younger rocks is known through a vary delicate test. The quartz phenocrysts in the quartz -porphyries, which occur both in the volcanic series and as later intrusives, show not the least trace of peripheral granulation i. This use of terms is as given by Dr. Williams in his lectures upon metamorphism, in 1894 at this University. See also Bull. 62, U. S. Geol. Survey, p. 36, 1890. 66 ALTERATION. 67 or even of undulatory extinction. So brittle is this mineral that it is not possible to believe that these quartz-porphyries have been subjected to the same strains as the greenstones to the north. The mineralogical changes in both the older diabase flows and the later volcanic series are much more important than the structur- al changes just mentioned. In many of the rocks examined, these changes in mineral composition have been so complete that none of the constituents have escaped destruction. In general, these changes have been of the character of pseudomorphism, so that al- though the mineralogical composition of the rock has been much modified, its structure or texture remains. These molecular re- arrangements, by which certain minerals are replaced by others, are, in great part, such as are believed to result from processes that are of the nature of weathering rather than of metamorphism. Even in the greenstone it is seen that the degree of mineralogical change is not dependent upon dynamic action, while no cases are known oi purely contact metamorphism, except in the Calder- wood's Neck series, as mentioned on page 29. For convenience in treatment, the mineralogical changes may be discussed as they are exhibited in the alteration of the feld- spathic and of the ferro-magnesian constituents. The mineralogical alteration of the feldspars of the green- stone has been of the character of saussuritization. In this process, aggregates of clear zoisite and secondary feldspar grains have replaced the original feldspar laths. In a few cases, epidote is present, but as shown by Cathrein 1 , the epidotization of feld- spar differs from the more common type of saussuritization only in the greater amount of iron required. In the andesitic lavas, epidotization has been quite as im- portant as the production of zoisite. Large areas of the latter mineral were observed, however, and the relations seem to in- dicate its derivation from the feldspar. Calcite also occurs as one of the alteration-products of the feldspar, being usually found within the altered crystal. In the green schists, on the other hand, although the calcite undoubtedly was derived from the feld- spar in great measure, it usually filled amygdules or occurred in i Zeitschr. f. Kryst., vol. 7, p. 234. 1883. 68 GEOLOGY OF THE FOX ISLANDS. irregular areas. The most striking- case of the alteration of feld- spar is in the andesite described on page 32. In this rock, the labradorite phenocrysts often show altered cores, the alteration being- to an aggregate of muscovite, chlorite and calcite, and the boundary between this core and the unaltered border is a sharp line. (Plate I, Fig-. 3.) Such a zonal relation can be explained as resulting- from a difference, either physical or chemical, between the core and the rim. The alteration would be favored in the former case by the presence of inclusions in the core, and in the latter, by the more basic character of the core. Comparison with other feldspar phenocrysts in this rock shows that they are more basic in the interior, and while inclusions are present, they occur only near the outer rim. Therefore, this zonal alteration is believed to have originated from differences in chemical composi- tion, the more basic core being- more easily attacked by solutions 1 . The ferro-mag-nesian constituents of these rocks show one principal kind of alteration, that of chloritization. Some small amounts of what appeared to be actinolitic fibers were observed in the tuffaceous greenstone; but the absence of any considerable amount of secondary hornblende in these pyroxene-bearing- rocks is a remarkable feature. Such uralitization has been considered a truly metamorphic process, while the hydrous chlorite minerals are rather the products of weathering- 2 . The chlorite has the optical characteristics of pennine and shows different modes of de- rivation from the augite. In the older green schists, the former presence of the pyroxene is only implied from the shape of the triangular areas of chlorite between the feldspar laths; but in the andesites and diabase-porphyries of the later series, the chlorite forms pseudomorphs after the pyroxene phenocrysts or occurs in association with the partially altered augite. In the latter case, the progress of the mineralogical change is seen to be along de- finite planes in the augite, especially the orthopinacoid (100), which is the normal solution plane 3 . It may be added that a 1. Similar cases have been described by various writers; vid. Grooser : Tsch. Min. u. Petrog. Mitth., vol. 13, p. 13; and Matthew: Trans. N. Y. Acad. Sci., vol. 14, p. 212. 2. Williams: Bull. 62, U. S. Geol. Survey, pp. 52-56. 3. Judd: Min. Mag., vol. 9, p. 192. 1890. ALTERATION. 69 similar relation to definite crystallographic planes was noted in the arrangement of some peg's of epidote in an altered feldspar crystal 1 . The alteration of the ores is most apparent in the large amounts of leucoxene present in many of these rocks. To a great extent, this leucoxene appears to be titanite and its origin from ilmenite is shown most beautifully in some of the diabase-porphyries. The titanite and rutile found in the greenstones are probably derived from ilmenite. Certain apparent irreg-ularities in the processes of alteration de- serve mention. In the greenstones, the mineralogical chang-es were less marked in the rocks which 1 had suffered most structural metamorphism; the feldspars have been crushed, but yet are clear and show no development of zoisite. In the North Haven vol- canics, specimens which had their feldspars completly altered, often contained augite crystals barely affected, and vice versa. These facts prove that the mineralogical changes are not necessarily of the nature of reactions between the components of the different constituents. DEVITRIFICATION. In the glassy lavas, the process of devit- rification is the pre-eminent mode of alteration. The presence of granular ag-greg-ates of crystalline material in portions of the rock, that also show structures believed to be characteristic of glass, points to some chang-e having- taken place. Devitrification is the process that will account for this chang-e of amorphous g-lass to de- finite crystallized minerals, and it is recog-nized, therefore, as of chief importance in the explanation of the differences between the old rhyolites and their recent equivalents. The fact that these g-lassy lavas contain spherulites and other crystalline structures that are undoubtedly original, makes it a difficult problem to de- cide to just what extent devitrification has contributed to the pre- sent rock-structure. The absence of any subsequent dynamic action in these lavas, however, favors the separation of primary and secondary crystallization. This feature also shows that dynamic action is not needed to initiate the processes of devitri- fication. i. Gentil believes that the alteration of nepheline proceeds in a similar way along di- rections of feeble cohesion. Bull. Soc. Franc, d. Min., vol. 17, p. 108. 70 GEOLOGY OF THE FOX ISLANDS. NOMENCLATURE OF ALTERED IGNEOUS ROCKS. In view of the changes which these volcanic rocks have undergone since their consolidation, questions of classification and terminology naturally arise. The usage in the nomenclature of altered igneous rocks has been varied and in the case of the volcanic rocks, it has been such as to cause the true nature of the older rocks to be misin- terpreted. Present and acquired differences have been so magni- fied as to be regarded as genetic and inherent. Thus "older" and "younger" volcanic rocks have been separated as fundamentally distinct. In the present study, the usage adopted has been that of re- taining the name of the equivalent recent volcanic rock for the Paleozoic lava. Genetic identity has been recognized in this manner, and the geologic relations interpreted accordingly. How- ever, it is realized fully that these old andesites and rhyolites are not at present exactly similar to the Tertiary andesites and rhyolites. It is doubtful if these, differences are of classificatory value, yet, an ideal nomenclature should give recognition to secondary changes which have contributed so much to the present character of the rock. In the study of similar old volcanics in the South Mountain area, Miss Bascom 1 suggested a terminology for the acid volcanic rocks, in which devitrification had been the process of alteration, and her term, aporhyolite, has been used in the preceeding pages to signify a devitrified rhyolite. In the case of the North Haven andesites and porphyries, there seems to be a need of some terms, which will express the altera- tion of these rocks, without affecting their position in a genetic classification. In the present instance, the alteration has been not structural, but mineralogical or pseudomorphic. Thus, the origi- nal nature of the rock can be determined and deserves full recogni- tion. Refinement in such a terminology is to be avoided, since it has been observed in these volcanics that the degree of alteration may vary greatly within a small area, and the recognition of these different degrees of alteration would be impracticable. Moreover, secondary differences are not to be emphasized, since they are not always essential to the interpretation of geologic history. How- i. Op. cit., p. 828. ALTERATION. 71 ever, it seems certain that the ideal nomenclature of altered ig-- neous rocks has not been secured; and it is with the hope of calling- attention to this feature in the study of ancient volcanics that these pag-es have been devoted to the subject of alteration. GEOLOGIC HISTORY. ACADIAN TROUGH. The correlation and interpretation of any rock series involves a consideration of the broader topographical features during- the period in which these rocks may have been formed. Such a view must comprise the whole geologic province of which the special area is a part; and thus it becomes necessary to consider briefly the Paleozoic features of what is now the coast- region of Maine. In his papers on the Archaean axes of North America, Prof. Dana outlined the features of this area, at the be- ginning of the Paleozoic 1 . The Mount Desert Range with a south- western trend marked the old coast-line, which, thus, approximate- ly coincides with that of to-day. To the southeast lay a great basin of sedimentation, the Acadian trough, which extended from Northern Newfoundland to the present Bay of Fundy and thence to Massachusetts and Rhode Island. This was probably not an open sea, but one protected by barriers, now represented by Nova Scotia and the more shallow parts of the present sea-floor, which mark the southern boundary of the Gulf of Maine. In this trough, sediments were deposited throughout Paleozoic time and in its various portions great thicknesses of sedimentary rocks are ex- posed to-day. PKE-NIAGARA VOLCANIC ACTIVITY. Along the border of such an area of great sedimentation, volcanic outbreaks might naturally be expected. In Southern New Brunswick, the pre-Cambrian rocks have been found to be made up, in large part, of surface volcanics, lavas, breccias, and tuffs, petrographical description of which has been published by Dr. W. D. Matthew 2 . Throughout 1. Bull. Geol. Soc. Am., vol. 7, p. 36. Am. Jour. Sci. (3rd series), vol. 39, p. 378. 2. Trans. N. Y. Acad. Sci., vol. 14, p. 187. 1895. Bull. Nat. Hist. Soc. New Brunswick, No. 13. 1895. 72 GEOLOGIC HISTORY. 73 all of the maritime districts, the presence of such volcanics has been recognized by the Canadian geologists 1 . On the Maine coast, the volcanic deposits described by Professors Shaler and Davis are considered to be younger, of Silurian and Devonian age. Pre-Niagara volcanic activity was a factor also, in the geologic history of the Fox Islands. The greenstones forming the north- ern portion of North Haven originated as amygdaloidal flows of diabase and accompanying deposits of coarse and fine ejectamenta. The lava-flows were characterized by contraction phenomena similar to those observed in the lavas of Vesuvius and other mod- ern volcanoes. These old lavas and tuffs are metamorphosed to such an extent that their original characters are greatly obscured. This relatively greater degree of metamorphism seems to indicate dynamic action during the time-interval between the eruption of these lavas and the post-Niagara volcanics, which do not show any structural metamorphism. It seems probable that further geologi- cal study will reveal the presence of other areas of these older volcanic rocks on the Maine coast. NIAGARA SEDIMENTATION. Throughout these periods of vulcanism, sedimentation doubtless continued in adjoining por- tions of the Acadian trough. Indeed, on account of the loose nature of many of their products, the volcanoes would furnish abundant material to be easily eroded and transported to the Acadian Sea. In the area under discussion, the Calderwood's Neck schists represent a thick series of arenaceous and argillaceous sediments, which were deposited, doubtless, in early Paleozoic time, but their exact age has not been determined through lack of fossils. Similar old sediments have been described by Shaler as occurring on Mount Desert 2 . Sediments of Niagara age occur on North Haven and their varied character and rich fauna doubtless point to their being younger than the metamorphosed schists on Calderwood's Neck. The basal members of this Niagara series were deposited upon the diabase flows, which were evidently considerably disintegrated on their surface, before this transgression of the sea in Niagara time. 1. For review of literature : Williams: op. cit., pp. 14-21. 2. Op. cit., p. 1037. 74 GEOLOGY OF THE FOX ISLANDS. POST-NIAGARA VOLCANIC ACTIVITYC While the six hundred feet of Niagara sediments represent an interlude in the volcanic activity at this point on the border of the Acadian trough, the break was not one of long duration. Even within the Niagara series, there are indications that abundant fine volcanic dust fell into the quiet waters and were deposited forming the red shale which occurs at several horizons in the series. Such deposits simply mark the commencement of a period of vulcanism which has left a full record of its nature. Flows and streams of andesites, andesite-porphyries, and diabase-porphyries; beds of volcanic blocks, more or less rounded and cemented with fine dust or volcanic mud or even molten lava of later eruption; deposits of the fine volcanic sand and dust that fell on the slopes of the volcano or in the shallow waters of the ad- joining sea; such are the elements that constitute the record of the earlier part of the volcanic outbreak. Later, the volcanic vent or vents, from which these lavas, breccias, and tuffs of the North Haven series had been erupted, became less important. Already in the upper portions of the series, there were indications of an intermittent character in the eruptions, since several beds of sediments, not purely pyroclastic, occur between the purely volcanic members. Other volcanic pro- ducts of a decidedly more acid composition were next erupted, pro- bably from different centers. These acid volcanics had their fore- runners, the quartz-porphyries occurring within the North Haven series. Now, however, the acid rock-types became more import- ant, and rhyolitic lavas, taxitic and spherulitic, were erupted and spread over the surface. Fragmental material was also ejected and deposited in well-defined beds. In both its phases, this volcanic activity, that directly followed the deposition of Niagara sediments, was essentiall} 7 similar to the vulcanism of today. In view of the vicissitudes of their sub- sequent history, it is indeed remarkable that these old volcanics so well express the conditions of the volcanic activity during Paleo- zoic time. What was the extent of the volcanic area, it is difficult to state. Study of the Fox Islands rocks points to the fact that only a small part of the volcano or group of volcanoes is represented in these GEOLOGIC HISTORY. 75 islands. Correlation with the volcanic series of Mount Desert would be highly conjectural, but careful g-eolog-ic study of the whole Penobscot Bay area may afford data for a more complete determin- ation of the limits of this volcanic group. It has been the purpose of the present study to contribute to such a history of the Paleo- zoic vulcanism of this portion of Maine coast. DEVONIAN INTRUSIONS. At some time after the close of the volcanic activity represented in the rocks exposed on the Fox Islands, there was a tilting* of the series, with some folding" and faulting-. There is no evidence that the Niag-ara rocks were tilted before the eruption of the overlying- volcanics, but there are indi- cations in the relations of the acid volcanics to the granite, that this movement antedated the granite intrusion. There is no further evidence bearing- directly upon the time of the intrusions of the granite, diabase, and diorite. In other parts of the Acadian province, however, the great granite intrusions of Nova Scotia and New Brunswick belong- to the close of the Devonian 1 . Such an ag-e for the Vinal Haven granite-mass would afford an interval for the deposition of a cover of sediments, sufficiently thick to assure granular crystallization of the granitic and the more basic mag-mas; and therefore, these intrusions are considered, provisionally, as of Devonian ag-e. The character of these intrusions seems to have been of the nature of a displacement of the older rocks on one side of a definite line of fracture. Absorption or fusion of the country-rock by the granite is not at all probable, since at the contact, the granite shows no endomorphic effects other than a slig-ht chang-e in tex- ture. Study of the granite contact with the quartzitic schists and acid volcanics sug-g-ests that the intrusion was accomplished by the movement of the rock that lay to the south of the fracture line, a displacement which may have been horizontal, or in part vertical, and thus of the nature of a laccolitic intrusion, the unsymmetrical side being- to the north. Althoug-h the exact nature of this in- trusion is a matter of conjecture, the true intrusive character of the granite and other holocry stall ine rocks seems well established. J. W. Dawson : Quart. Journ. Geol. Soc., vol. 44,. p. 814. 76 GEOLOGY OF THE FOX ISLANDS. Thus, following* the eruption of the volcanics of basic and acid composition, there was a similar sequence of basic and acid plu- tonics. The g-eologic history of the Fox Islands, so far as rock-making- is concerned, closed with the intrusion of the granite. The later history has been that of destruction and degradation of the differ- ent rocks, and the carving- of the rock-masses to their present re- lief, by atmospheric processes. LIFE. George Otis Smith was born in Hodg-don, Maine, February 22, 1871. His early education was obtained in the public schools of Augusta and Skowheg-an, Maine, and in 1889, he entered Colby University, from which institution he was graduated in 1893, with the degree of A. B. In the three years following, he has pursued graduate courses of study at Johns Hopkins University in Geology under Professors Williams and Clark, Dr. Mathews and Messrs. G. K. Gilbert and Bailey Willis, and in Chemistry under Professors Remsen and Renouf. He has held the University Scholarship in Geology, 1894-5, and the Fellowship, 1895-6.. During- the field- seasons of 1893, 1894, and 1895, he has been eng-aged in g-eological work in the States of Michigan and Washington in the employ of the United States Geological Survey. PLATE I. FIG. 1. OH vine-bearing- andesite. North Haven. Olivine phenocrysts, older than the feldspar, and now altered to serpentine. Magnetite and apatite inclusions in the olivine. FIG. 2. Andesite. North Haven. Flow structure developed by feldspar laths of ground- mass. FIG. 3. Augite-andesite. North Haven. Feldspar phenocrysts with altered core, and other crystals of labradorite zonal. FIG. 4. Rhyolitic tuff, North Haven. Ash-structure developed by glass sherds, now devitri- iied. Fragment of lava with quartz phenocryst. FIG. 5. Aporhyolitic flow-breccia. Vinal Haven. Glass fragment with obsidian structure. FIG. 6. Spherulitic aporhyolite. Vinal Haven. Spherulitic intergrowth partially altered to a granular mosaic. Nicols crossed. PLATE I X 22 FIG. I. X I 34 FIG. 2. X 22 FIG. 3. X 44 FIG. 4. X 22O FIG. 5. X 44 FIG. 6. i i 11 ii m a YD ,'0053