PRINCETON UNIVERSITY CONTRIBUTION TO THE GEOLOGY OF NEWFOUNDLAND NO. 5 Pre-Cambrian Rocks of South east Newfoundland BY ARTHUR FRANCIS BUDDINGTON A DISSERTATION PRESENTED TO THE FACULTY OF PRINCETON UNIVERSITY IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY, APRIL, 1916 Reprinted from the JOURNAL OF GEOLOGY, Vol. XXVII, No. 6 September-October, 1919, pp. 449-79 Pre-Cambrian Rocks of South east Newfoundland A DISSERTATION PRESENTED TO THE FACULTY OF PRINCETON UNIVERSITY IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY, APRIL, 1916 BY ARTHUR FRANCIS BUDDINGTON EARTH SCIENCE* LIBRARY Accepted by the Department of Geology, April, 1916 PRE-CAMBRIAN ROCKS OF SOUTHEAST NEWFOUNDLAND 1 A. F. BUDDINGTON Brown University The following paper gives the more important results of an investigation based upon field work carried on during the summers of 1913 and 1914. Southeastern Newfoundland is important geologically in that it affords, as far as is known, the most complete section of the later pre-Cambrian rocks along the eastern coast of North America, and an understanding of this section should aid in the correlation of the later pre-Cambrian at other localities. The term later pre-Cambrian, as here used, is possibly the equivalent of the terms Proterozoic or Algonkian. LOCATION AND GEOGRAPHY The Avalon Peninsula (Fig. i) forms the southeastern portion of Newfoundland and is attached to the main island by a narrow, rugged isthmus, in places but three miles wide, which separates Trinity Bay on the north from Placentia Bay on the south. The peninsula itself is in turn almost split in twain again by St. Mary's and Conception bays. The major portion of the rocks here described lie along the eastern side of Conception Bay, at the head of this bay, and in the vicinity of St. John's; but they are believed to be typical of the entire peninsula and of the later pre-Cambrian rocks of eastern Newfoundland in general. Except for the mining industry conducted on a large scale at Great Bell Island in Conception Bay, fishing is almost the sole occupation of the inhabitants. As a consequence of this and of the unfavorable character of the interior of the peninsulas, habitations 1 Thesis presented to the faculty of Princeton University for the degree of Doctor of Philosophy. 449 444308 45 A. F. BUDDINGTON are confined almost exclusively to the shore lines, and often to small valleys at the mouths of brooks which may be from a fraction of a mile to several miles apart. Geological and Submarine ToDoaraphical Map of Newfoundland^^ ^ FIG. i LITERATURE The literature on Newfoundland pre- Cambrian geology is well summarized by Van Hise and Leith ( 1 909) . Jukes ( 1 843 , p. 5 1 ) first separated the Cambrian from the pre-Cambr-ian and noted the PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 451 unconformity existing between them. He was followed by Murray (1881), who gives accurate descriptions of the pre-Cambrian sedi- ments and classes them as the intermediate series, possibly equiva- lent to the Huronian of Canada. Walcott (1899, p. 219) proposed the name Avalon for the terrane lying between the basal beds of the Cambrian and the Archean gneisses of Newfoundland . Walcott (1900) described a new pre-Cambrian terrane conformable above the Signal Hill series, which he called the Random and included in the Avalon terrane. Howley (1907) published a map in which he outlined the distribution of the Avalonian series and also distin- guished a formation composed of interbedded volcanics and aqueous deposits, which he separated from the base of the Avalonian terrane and called Lower Huronian. The succession worked out by Murray and Howley, the addition by Walcott, and the modification pro- posed by Howley are given below. Feet Random Sandstones, quartzitic sandstones, and sandy shales i ,000 Red conglomerate 500 Signal Hill Dark-red sandstones 1,320 Greenish or gray fine-grained sandstones 1,300 Momable slates Dark-brown or blackish slates 2,000 Torbay slates Green, purple, pinkish, or red slates 3,300 Conception slates Greenish slates Lower Huronian Mixed igneous and aqueous deposits in-a highly meta- morphosed condition PHYSIOGRAPHY In general, the striking features of the Avalon Peninsula are the parallel lineaments of the topography; the ever-present, bold, rocky cliffs of the coast ; the flat-topped uplands with innumerable lakes and ponds (Jukes counted 153 from the top of Powder Horn Mountain); and the fiord bays with their deep interiors and shallower thresholds. The topography is that of a much-dissected plateau at an altitude of from 500 to 700 feet, with monadnocks rising to heights of 1,000 to 1,100 feet, and submerged river valleys deeply gouged by glaciers and invaded by the sea, constituting fiord bays and their arms. The remnants of two definable peneplains may be repre- sented in the present topography: one on the highland peneplain 452 A. F. BUDDINGTON of the plateau surface (Fig 2), the other represented in the broad river valleys and lowlands at a height of 250 to 300 feet. The trend lines of the more impressive and marked physio- graphic features have been controlled for the greater part by the presence of ancient fault and fracture lines, which have to a large extent determined the distribution of the underlying rocks, with their varying degrees of resistance to erosion and weathering and the consequent parallel lineaments of the present topography. Joints have certainly played a prominent part in localizing the erosive agents. FIG. 2. Highland peneplain on granite; interior of the St. John's Peninsula; about eight miles east of Holyrood. An uplift of the land following the period of glaciation is indi- cated by several facts. The amount of this uplift has been stated by Daly (1891, pp. 257-58), judging from the lower limit of undis- turbed glacial erratics at St. John's, to be about 575 feet; but the presence of perched bowlders in precarious positions on the tops of hills at much lesser elevations (especially at about 300 feet) around Conception Bay renders this estimate of doubtful value. GLACIATION The results of glaciation expressed in the present topography point to the presence of local ice caps flowing into the individual bays in a direction perpendicular to the major outlines of the bays PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 453 at each point, and hence controlled in a great degree by the gross features of the pre-glacial topography. The directions of ice movement are easily decipherable from the abundant striations, the stoss and lee sides of hills (Fig. 3), and numerous bowlder trains which may be traced to their source. Every bay indenting the coast of the Avalon Peninsula presents all the essential characteristics of a fiord, such as high, steep, straight, parallel walls, threshold across its mouth, and small, insignificant streams entering its head. In Trinity Bay (Fig. i) the inner, deepest portion is over 1,100 feet deeper than the sill across the mouth. Where the relations are known there is a remark- able parallelism between the strike of the major fault planes and FIG. 3. Glaciated hills with lee and stoss sides southeast of Harbour Main, Conception Bay. Lee sides are on the north-northeast ends of hills. the trend of the coast line, but the writer could find no evidence that faulting was directly responsible for the formation of the fiords. The faults and fractures controlling the lineaments of the bays are of ancient date older than the peneplains and their influence on the present topography has been indirect. The formation of the sills across the mouths of the bays by deposits from currents is improbable, since this theory will not hold for the inland fiords of Gander Lake, Red Indian Lake, or Grand Lake, the latter with its deepest part not less than 988 feet below sea-level and with its outlet 116 feet above sea-level. In view of these conditions glacial overdeepening of river valleys probably associated with some submergence followed by the invasion of the sea seems the best- adapted explanation for these fiord bays 454 A. F. BUDDINGTON GENERAL GEOLOGY The later pre-Cambrian beds have undergone at least two periods of folding, one at the close of the later pre-Cambrian, possibly the Penokean revolution, as denned by Blackwelder (1914), the other the Taconic. As a consequence of this, and of probable later movements, the beds are much disturbed, more or less closely folded, and very considerably faulted. The core of a major anticline composed of complexly faulted and folded beds of the volcanic series, cut across at a small angle by a stock of grano- diorite, is exposed at the head of Conception Bay. At Chapel Cove its eroded surface is overlain by beds constituting the south end of a northward-pitching synclinal fault block of Cambrian sediments, in which Conception Bay is excavated. We have here apparently the phenomenon of the location of a younger syncline in sediments deposited in a basin on the eroded crest of an anticline. Cutting across the east flank of the anticline is a batholith of granite forming the backbone of the St. John's Peninsula and called by the writer the Holyrood granite batholith. This is bordered on the. east by a narrow band of the volcanics overlain by successively higher beds involved in minor folds until the trough of a major syncline in the Signal Hill series is reached east of St. John's. According to Murray and Howley's map of 1881 the Carbonear Peninsula is formed of later pre-Cambrian beds which overlie the volcanics and are involved in two major synclines in the troughs of which the Signal Hill series appears, with an intervening anticline on the core of which the Torbay series appears. The Placentia Peninsula, according to the same map, is composed of a syncline of beds in the trough of which the Signal Hill series lies, while the volcanic series forms its western border and appears on the flank of an anticline. Metamorphism, although not extreme, has yet affected the rocks to a considerable degree. The beds lie in folds which, although, probably not closed, yet approach that state, and a slaty cleavage at an angle to the bedding is prevalent throughout the entire bedded series. Along localized zones basaltic flows and breccias have been changed to chloritic schists, and rhyolite flows have been analogously altered to pinite, quartz-pyrophyllite, or pyrophyllite PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 455 schists (Buddington, 1916). Yet the granite, granodiorite, and many of the volcanics show little or no affects of metamorphism and are quite unaltered. Whereas metamorphism has been a moderate factor, however, faulting on a tremendous scale and often of remarkable intensity has been the predominating feature of the diastrophic processes. These faults trend in the main a little east of north, and the throw along many of them is measurable in thousands of feet. A few of the more important ones are the fault between the Cambrian and pre-Cambrian along the east side of Conception Bay from Topsail to Cape St. Francis, with a throw of 8,000 feet, as determined by van Ingen; the fault a short distance inland from the west side of Colliers Bay in Conception Bay, which may be traced for almost ten miles and separates the Cambrian from the pre-Cambrian and the Conception slates from the volcanic series; and lastly, the fault on the west side of Random Sound, which may be traced for fifteen miles and brings Cambrian beds against pre-Cambrian granite and the Conception slate series. UNCONFORMITY BETWEEN CAMBRIAN AND LATER PRE-CAMBRIAN The uncomformable relations of the Lower Cambrian beds to those of the later pre-Cambrian, with the possible exception of the Random formation, can be definitely established. At eleven widely separated localities in Conception Bay the actual unconformable plane of contact between the Lower Cambrian and the later pre- Cambrian may be observed. At Brigus the Lower Cambrian rests with an angular uncon- formity on the upturned beveled edges of members of the Avondale volcanic series. Along Colliers Bay the Lower Cambrian with a gentle dip rests unconformably on more or less vertical basalt flows and breccias of the Avondale volcanic series. At Chapel Cove the Lower Cambrian beds rest on the eroded surface of a quartz syenite and contain pebbles of the underlying rock. At Duff's Station the Lower Cambrian overlies granite, occupying hollows and filling deep, narrow wedgelike spaces between joint planes so as to resemble a network of veins on the surface. At Upper Gullies it rests with unconformable relations on a gabbro mass. On Kelligrews Brook 456 A. F. BUDDINGTON a chloritized granite gneiss is overlain by Cryptozoan limestone beds of Lower Cambrian age. Just west of Manuels three exposures of the unconformable relations are presented. At the first locality limestone rests unconformably on granite and on a rhyolite dike in the granite. A shale bed of Cambrian age overlaps the limestone and also rests directly on the granite and contains pebbles of it. A near-by conglomeratic limestone consists almost entirely of cobbles of rhyolite similar to the rock of the rhyolite dike. At the second locality a deep-fissure deposit of conglomeratic limestone with pebbles of the country rock occurs in diabase. At the third locality a Lower Cambrian conglomeratic limestone carrying Hyolithes and Coleoloides rests on dark-red rhyolite, the pebbles consisting almost entirely of the underlying rock. Again, in the bed of Manuels Brook a coarse conglomerate rests unconformably on volcanics and on a granite dike intrusive in the volcanics. The conglomerate is of Lower Cambrian age and consists of cobbles and bowlders of the underlying granite and volcanics. On Trinity Bay, along Smith and Random sounds, the Lower Cambrian, of older age than that on Conception Bay, rests discon- formably on beds of the Random formation in the trough of a syncline. The Random may be either conformable or disconform- able with the Signal Hill series. In either case the Lower Cambrian must be unconformable with the Signal Hill series and all lower formations, since it must necessarily have transgressed all the later pre-Cambrian formations in passing from its disconformable posi- tion on the Random, the youngest formation of the pre-Cambrian, in the trough of a syncline, across the flank of an anticline on to the lowest later pre-Cambrian formation, the Avondale volcanics, on the core of an anticline in Conception Bay. AVONDALE VOLCANICS Howley (1907) mapped a series of rocks which he called Huronian and described "as composed of mixed igneous and aqueous deposits in a highly metamorphosed condition." Parallel to their trend they are mapped by Howley as outcropping for a distance of about two hundred miles in a north-northeast-south- southwest direction, being submerged beneath the sea at each end. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 457 In view of the unknown age of this formation, and because of its characteristic development and typical outcrops in the vicinity of Avondale, Conception Bay, the writer proposes the name Avondale volcanics to designate this series, which comprises rocks of direct volcanic origin, together with a small amount of interbedded more or less water worn volcanic materials. They constitute a formation of wide geographical extent 200 miles along the strike and 100 miles or more across the strike and of great but unknown thick- ness, at least several thousand feet, forming so far as is known the basal member of the later pre-Cambrian or Avalonian system of Newfoundland. Interbedded rhyolite and plagioclase basalt flows, with corresponding breccias, crystal, lithic, and vitric tuffs, volcanic dust beds, and average tuffs, together with volcanic conglomerates, sandstones, and slates are comprised within the series. The beds are in a highly disturbed condition, with a prevailing steep dip, and are affected by intense and often profound faulting. They are exposed on the cores of major anticlines or are brought to the surface on the upthrow sides of great faults. The degree of metamorphism which the rocks have suffered is in general moderate, though locally intense, and is manifest for the larger part merely in the development of a slaty cleavage. Secondary minerals, except those of surface origin, are confined almost exclusively to sericite and quartz in the acid rocks and chloride with sericite, epidote, and calcite in the basic rocks. Along certain shear zones in the acid rocks pyrophyllite, quartz-pyrbphyllite, and pinite schists have formed through replacement, while in certain basic rocks copper ores have been deposited along close-set narrow fractures and in vesicles of the flows. Ore in quantities of commercial importance, however, has not yet been found. In view of the great age of the rocks the lack of intense meta- morphism and the frequent preservation of delicate primary structures are striking. Indeed there is little doubt that, with search, practically every structure and texture found in the Tertiary and recent volcanics might also be found in these ancient pre- Cambrian volcanics. One of the most interesting discoveries in this series was the presence of volcanic necks intruded through the breccias and tuffs. 458 A. F. BUDDINGTON Near Manuels a plug of plagioclase basalt about 400 by 450 feet, with columnar jointing perpendicular to the surface and with dikes and apophyses radiating from it into the adjoining rocks, was found intruded through basaltic and rhyolitic breccias. A rhyolite neck about fifty feet in diameter and circular in shape, nearly every foot of the contact of which could be seen, and a neck of basic tuff with fragments up to three feet in diameter of the underlying rocks, through which it was drilled, and with radiating dikes of tuff, were also found. The largest neck forms a conspicuous hill about three miles south of Holyrood and seems to have been one of the main centers of volcanic activity in this region. Here we find a great elliptical stock of rhyolite porphyry about a mile long and half a mile wide, while in the surrounding country occur exceptionally coarse rhyolite breccias, with many blocks up to two feet in diam- eter, interbedded with rhyolite flows and tuffs. A chemical analysis of this rock, given below, shows it to be very similar to the rhyolite flows at Manuels. TABLE I CHEMICAL ANALYSIS or RHYOLITE PORPHYRY FROM VOLCANIC NECK SOUTH OF HOLYROOD SiO 2 75.85 K 2 O 4.47 A1A 13 .03 H 2 O+ 35 Fe 2 O 3 i .82 H 2 O- 05 FeO *.. .32 MnO 05 MgO 05 GaO 38 Total 99.71 Na a O... :.../ 3-34 The rhyolite flows are quite variable in color, ranging through reddish, purplish, and greenish grays. They are usually aphanitic in texture, and dense, chertlike in character, sometimes felsitic, with a finely rough feel. They may be homogeneous and massive, or may be marked by beautiful, well-developed, and well-preserved banded textures, and are very frequently spherulitic. The spheru- lites vary from microspherulites up to those as large as a man's head, or even occasionally attain a diameter of two feet (Fig. 4), but in general they are of medium size (Fig. 5). The flows may also be characterized by a platy jointing more or less parallel to the flowage planes, or by flow breccia, eutaxitic (Fig. 6), or perlitic structures. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 459 As seen in thin section the texture is usually microgranitic or an irregular polarizing aggregate of quartz and feldspar. Spherulitic, axiolitic, and microspherulitic textures are common, and rarely a felt of microlitic feldspars in a groundmass of quartz is observed. A micrographic texture characterizes portions of an eutaxite from FIG. 4. Very coarsely spherulitic rhyolite; the largest spherulite is twenty-two inches in diameter; about five and one- half miles south of Manuels, Conception Bay. near Avondale. A chemical analysis of a specimen from a 50-foot rhyolite flow follows. The rock is a dense, red, banded felsite with a microfelsitic to finely macrocrystalline and micropoikilitic texture. TABLE II CHEMICAL ANALYSIS or RHYOLITE FROM FLOW AT MANUELS SiO 2 76 . 24 A1 2 O 3 .13.94 Fe 2 O 3 89 FeO.. 13 MgO 27 CaO 1.07 Na 2 2.55 K 2 O 4-95 H 2 0+ 15 H 2 0- 03 Total.. ..100.22 460 A. F. BUDDINGTON FIG. 5. Medium spherulitic banded rhyolite; about five and one-half miles south of Manuels, Conception Bay. Photograph by G. v. I. I 'W^fv -JH* x - * * i Oi v'-:l : ;i^.::: ; : ' f V FIG. 6. Microphotograph of eutaxitic structure in rhyolite. Hill 467, south of Avondale, Conception Bay. Natural light, X$2 diameters. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 461 Basalts attain a tremendous development along Colliers Bay, extending south toward Brigus Junction. They are associated with a great development of basaltic breccias and often exhibit amygdaloidal or Howage structures. The flows are very much altered throughout and may be chloritized with resultant green- gray rocks, or hydrated and thoroughly impregnated with iron oxide giving purplish- or reddish-brown hues. These alterations are probably due to thermal waters, as they occur independent of what FIG. 7. Coarse rhyolite breccia; about two and one-half miles south of Man- uels, Conception Bay. Photograph by G. v. I. might be expected as the result of weathering. In texture they vary from felsitic to finely porphyritic, and in thin section show a felt of plagioclase laths or microlites with a distinct fluxion structure in a chloritic or indistinct altered groundmass. Chlorite pseudo- morphs, replacements of some ferromagnesian mineral, possibly augite, occur as abundant minute phenocrysts in some flows. The breccia beds exhibit a varied assortment of colors and a wide range in the diameter of their component fragments; but by far the predominant portion of the rhyolitic breccias are various hues and tones of reddish- and purplish-gray, whereas the basaltic breccias 462 A. F. BUDDINGTON consist of dark greenish-gray fragments in dark reddish-, purplish-, or greenish-gray matrices. Both basaltic and rhyolitic breccias exhibit approximately the same range in the diameter of their fragments, from a fraction of an inch to four feet (Figs. 7 and 8). The general run of the fragments FIG. 8. Coarse basalt breccia; west side of Blue Hills, south of Conception Bay would probably average a few inches in diameter. In general, the rhyolitic breccias and the basaltic breccias contain few foreign fragments, especially the coarser beds; yet beds in which angular fragments of rhyolite and basalt and crystals of orthoclase and plagioclase are mingled are frequent; and rhyolitic breccias and basaltic breccias are often found interbedded. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 463 Crystal tuffs are of frequent occurrence interbedded with the breccias and tuffs. They are remarkable, in view of their great age and the vicissitudes which they have undergone, for the frequent presence of a perfectly preserved vitroclastic structure in their groundmass, whether rhyolitic or basaltic in nature. These devitrified glass shards are such a prominent feature of some of the tuffs from the vicinity of Red Rock Lake near Brigus that they JT IG g Microphotograph of vitroclastic groundmass of rhyolitic crystal tuff; near Red Rock Lake, south of Brigus, Conception Bay. Natural light, Xs6 diam- eters. might well be denominated vitric tuffs (Fig. 9). In other cases the groundmass is an indistinct, microcrystalline, granular dust. Many of these tuffs are soaked with hematite and hence are reddish in color. Associated with the phenocryst-like crystals are frag- ments of either porphyritic or felsitic basalt or rhyolite, or of devitrified obsidian exhibiting trichites, microspherulites, spheru- lites, axiolites, perlitic cracks, pumiceous structure, or any combina- tion of these. The crystals consist principally of plagioclase and 464 A. F. BUDDINGTON orthoclase, with quartz and magnetite in the acid tuffs, and of plagioclase, augite, magnetite, apatite, and foreign crystals in the basic tuffs. Some of the basic tuffs resemble typical palagonite tuffs (Fig. 10). The lithic tuffs consist predominantly of fragments of rhyolite or of pilotaxitic basalt, according to their nature, in a groundmass FIG. 10. Groundmass of basaltic crystal tuff, with devitrified shards of glass bordered by rims of fibrous, brownish material. Hill 937 near Holyrood. Micro- photograph. Natural light. consisting of finely comminuted materials similar to the lithic fragments, with crystals of feldspar and quartz in minor amount. The volcanic-dust beds are frequently well bedded and chertlike in character, often translucent in thin edges, with a conchoidal to subconchoidal fracture. The waterworn volcanics comprise volcanic conglomerates, sandstone, and a rare red shale bed. The prevalence of volcanic PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 465 conglomerates, both stratigraphically and geographically, inter- bedded with the volcanic breccias, flows, and tuffs, is one of the most cogent arguments for the subaerial origin of the volcanic series as a whole; for both basalt and rhyolite conglomerates are associated with and interbedded between completely angular basalt breccias and well-rounded basalt conglomerates, and between similar angular rhyolite breccias and well-rounded rhyolite conglomerates. In some of the basalt conglomerate beds bowlders up to three feet or FIG. ii. Basalt volcanic conglomerate; Turks Gut, Colliers Bay, Conception Bay. Photograph by G. v. I. more in diameter are not rare (Fig. n). Such beds are in places 200 feet thick and grade upward or downward into sandstones or tuffs. Some of the sandstone beds are cross-bedded and contain abundant fragments of red shale constituting typical thon-gallen beds. Others are ripple marked, but these features are rare. In the course of the work no reasons were found for assuming the volcanics to have had other than an essentially subaerial origin. All the structures and textures of the flows and the characters of the associated volcanic products are compatible with such a hypothesis, while the particular criteria pointing to such a conclusion are as follows : 466 A. F. BUDDINGTON The predominance and widespread distribution of the red and brown volcanic tuffs and breccias, which owe their color to satura- tion and cementation with hematite, point toward subaerial exposure rather than toward local effects of volcanic gases. The constant association of volcanic conglomerates or sandstone with the breccias, tuffs, and flows indicates the work either of river erosion, or wave work, or both. The first seems 'more probable in view of the constant recurrence of conglomerate beds in the strati- graphic succession, as well as the fact that beds of conglomerate 150 to 200 feet, thick with well-rounded bowlders up to three feet in diameter, are exceptional within marine deposits, even within marine volcanic deposits. On the other hand such conglomerates are commonly found associated with subaerial volcanics of all ages. Those of Newfoundland are such as might be expected to form along river valleys draining a region of great active volcanic cones such as this probably was. Although some of the conglomerates may have been deposited in standing water, the indications are that for the most part they were deposited on a land surface during quiescent periods between successive volcanic outbursts which repeatedly buried them with the products resulting from extravasation and explosion. The volcanics are, so far as known, the oldest formation of the later pre-Cambrian in this district. They are intruded by granite, and together with the granite are overlain unconformably by Lower Cambrian sediments. The pre-Cambrian volcanics of the Blue Ridge of Virginia and Maryland present similar relationships (Keith, 1892), and in the present state of knowledge it seems prob- able that in early later pre-Cambrian times a chain of volcanic cones extended from Newfoundland to North Carolina and farther, in a zone more or less parallel to the present coast line. CONCEPTION SLATE SERIES The Conception slate series was examined by the writer along the east side of Conception Bay from Portugal Cove to Topsail; along the west side, Colliers Bay north to Brigus; near La Manche on the Isthmus of Avalon ; and along Smith and Random sounds on Trinity Bay. Owing to folding, faulting, and covered exposures PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 467 the relations of this series to the beds above and below were not seen. The series consists predominantly of thin-bedded, green-gray, dense, halleflinta-like slates, at least several thousand feet thick, which may be compact without cleavage, or dynamically meta- morphosed, with a resultant well-developed cleavage; and to a lesser extent of green-gray feldspathic sandstones and conglom- erates. The slates often resemble at a superficial glance banded flow rhyolites, but occasional sandy layers betray their true origin. The slates may vary from very fine-grained or dense siliceous quartzites to argillaceous slates, or from dense feldspathic quartzites to chertlike feldspathic slates. Thin, intercalated beds and layers of feldspathic sandstone are common, but the predominant char- acter of the rocks is that of a slate. The feldspathic character of the typical slate and its normally fresh, undecomposed character are shown very well by the two chemical analyses which follow. It will be noted that the soda is in excess of the potash, contrary to the usual rule for sediments derived from well-weathered materials, and that the analyses resemble those of certain rhyolites. The high soda content may be due to plagioclases washed in from basic tuffs. TABLE III CHEMIQAL ANALYSES or MEMBERS or CONCEPTION SLATE SERIES i~-_. 2 SiO 2 . . 68.92 77 J 3 A1 2 O 3 17.67 14.01 Fe 2 O 3 i 43 2 25 FeO ^ 48 Undet. CaO 8s 64 MgO I . IO 36 Na 2 O K 2 O 3-oi i .46 3-59 i .01 Ign. Loss i-75 .76 Total QO 76 QO . 7s 1. Feldspathic slate, near Brigus, Conception Bay. 2. Feldspathic quartzite, Robinson's Bight, Random Sound, Trinity Bay. The sandstones are formed of angular to subangular grains of quartz, fresh plagioclase, orthoclase, rhyolite, plagioclase basalt, and tuffs. The rhyolite fragments exhibit axiolitic, spherulitic, and banded textures. 468 A. F. BUDDINGTON The conglomerate outcrop near La Manche, and the pebbles, are all of volcanic rocks such as basalt, basalt porphyry, rhyolite porphyry, felsite, etc., imbedded in a matrix showing innumerable flashing cleavage faces of white plagioclase. Since these beds in their well-bedded and unoxidized characters give evidence of their being deposited beneath a permanent water- level, and since they consist almost exclusively of materials similar to the rocks comprised within the underlying volcanic series of subaerial origin, it is probable that an unconformity exists between the Conception slate series and the Avondale volcanics, and that the former are derived from the latter, at that time a more or less loose accumulation of volcanic ash and breccias which were swept into the sea in a comparatively fresh and undecomposed condition. TORBAY SERIES The Torbay series was not studied by the writer. Murray's description of it is inserted here for completeness: Green, purple, pinkish, or red slates in frequent alternations; the texture of these slates is generally extremely fine, and in some cases they approach in hardness to jasper or chert. The fracture is often conchoidal, and the imperfect cleavage parallel with the bedding; but in many instances the rock has a good cleavage at right angles to the stratification and is well adapted for roofing purposes. The exposed surfaces weather for the most part a yellowish white. Some beds seen by the writer on the Isthmus of Avalon strongly resemble banded argillites; cobbles of green and red argillite were also found in the drift of the St. John's Peninsula, presumably derived from- the Torbay series. MOMABLE SERIES These beds are described by Murray (1881, p. 145) as follows: Dark brown or blackish slates of St. John's, with ripple marks very dis- tinctly displayed on some surfaces, and in which some obscure organic remains have been found resembling those found in c, and another supposed to be the shelly casing of some description of Annelid. The cleavage of this slate is sometimes very regular, oblique, or at right angles to the bedding, but in parts it also cleaves parallel with the stratification. Towards the top are frequent layers of hard, fine-grained, greenish sandstone interstratified, not usually over 6 or 7 inches in thickness. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 469 The beds seen by the writer are strongly ripple-marked, blackish- gray, thin-bedded, very fine-grained, shaly sandstones, with thin films of black carbonaceous or graphitic shale. In thin section the sandstone is seen to be made up of sharply angular grains of quartz, with an occasional fresh feldspar fragment and bits of carbonaceous material. The grains average about 0.05 'mm. in diameter. From a lithologic point of view the most striking feature of these rocks is their content of carbonaceous material, which, although slight and often absent, yet serves to contrast them with the over- lying and underlying formations. In contrast to the succeeding reddish-brown feldspathic sandstone formation, the gray, fine- grained quartz sandstones and shales of the Momable suggest an origin under a permanent water-level at shallow depths in which flourished organisms whose presence is indicated by the carbonaceous content of the shales and by the possible fossil Aspidella. Attention has been called by many observers to the resemblance between the gold-bearing series of Nova Scotia and Murray's Intermediate series of Newfoundland. It is to the beds comprising the Conception, Torbay, and Momable series lying below the Signal Hill series and above the Avondale volcanics that this lithologic resemblance applies. Judging from Fairbault's descriptions (Mal- com, 1912, pp. 46-47) the Golden ville quartzites present certain resemblances to the Conception slate series, the banded argillite division of the western part of the field to the Torbay series, and the Halifax formation to the Momable series. Specimens of the Halifax slate collected by the writer from an outcrop in the yards of the Canadian Pacific terminal at Halifax are indeed indistinguish- able, except for a more prevalent cleavage, from many characteristic specimens of the Momable series. Before any correlation may be suggested, however, it is necessary that a more intimate knowledge of the conformable or unconformable relations within the New- foundland series shall be known, as well as more definite information concerning the relations of the gold-bearing series of Nova Scotia to the Cambrian. The Momable beds are also similar to the Animikie sediments as described by Coleman in respect to their content of carbonaceous material. 470 A. F. BUDDINGTON SIGNAL HILL SERIES As near as can be judged from the descriptions of Murray and Howley (1881) and from the experience of the writer the Signal Hill series outcrops mainly in the troughs of synclines or synclinoria over an area extending for a length of about 200 miles in a north- northeast-south-southwest direction and about 75 miles at right angles to this. The series consists of a very thick succession of reddish-brown, occasionally green, feldspathic sandstones and conglomerates, with intercalated shale beds and about 1,300 feet of greenish-gray, thick- bedded feldspathic sandstones at the base, giving a. thickness estimated at about 10,000 feet. Murray's estimate of 3,120 feet for this series referred only to the beds on the west side of St. John's harbor, which constitute but a part of the series. Thon-gallen or intraformational conglomerate beds are abun- dant. The majority of the conglomerate beds may be appropriately described as gravel or pebble beds, with the pebbles varying in size from one-fourth of an inch to an inch in diameter, and prevailingly subangular to rounded. Because of the, predominating sandy character of such a thick series of sediments; the repetition and often considerable thicknesses of conglomerate beds; the presence of so much fresh plagioclase and orthoclase throughout the rocks; the prevailing subangular character of the component grains; the predominating red color due to interstitial hematitic mud, to films of hematite, and to oxidized grains of basalt, rhyolite, and magnetite; the constant recurrence of thon-gallen beds; frequent cross-bedding; lithologic alternations of sandstone, conglomerate, and shale; and the absence of any limestone, fossils, or carbonaceous materials, the writer has been led to conclude that these beds originated as domi- nantly flu via tile deposits of subaerial origin in a subarid climate. It is quite possible that the basal 1,300 feet of green-gray sand- stones accumulated under a permanent water-level; but if so the water was apparently drawn off or excluded at the time of accumu- lation of the succeeding reddish-brown series. In thin sections from specimens of deep, livid-brown sandstone from Signal Hill the rock is found to consist of angular to subangular PRE-CAM BRIAN ROCKS OF NEWFOUNDLAND 471 grains of quartz, fresh plagioclase, orthoclase, and rhyolite and granophyre, with rarely a fragment of basalt. Accessory minerals comprise grains of primary epidote and occasionally abundant grains of magnetite. A film of hematite coats each grain, and in some specimens the interstices are filled with secondary quartz, sericite, or occasionally with epidote. The sericite is a secondary product, the result of recrystallization during shearing. In one section a rounded grain of primary epidote is coated with a film of hematite and is surrounded on the outside of this by a growth of secondary epidote. The cement of the green sandstone seems to be of an epidotic nature, apparently arising through a recrystallization of an impure argillaceous material. A chemical analysis of the reddish-brown sandstone serves to confirm the highly feldspathic nature of this rock. TABLE IV CHEMICAL ANALYSIS OF FELDSPATHIC BROWN SANDSTONE FROM NEAR SIGNAL HILL, ST. JOHN'S SiO 2 71.38 Na 2 O 2.28 A1 2 3 14-25 K 2 i . 99 Fe 2 O 3 4. 75 H 2 O+ i . 20 FeO 46 H 2 O- 21 MgO 46 CaO 3.01 Total . r 99-99 That the red color of the brown sandstones is due to hematite and to the oxidation of its iron content is evident from a comparison of the ferrous and ferric contents of the red and green sandstones. Although both have similar total contents of iron, 4 . 74 and 5.11 per cent respectively, expressed in terms of ferrous iron, the brown sandstone shows 4.75 per cent of ferric oxide and only o .46 per cent of ferrous oxide, whereas the green sandstone shows 2 . 54 per cent of ferric oxide and 2 .82 per cent of ferrous oxide. The red color of the sandstones depends in varying degrees upon the primary deposition of a hematitic mud in the interstices of the sand grains; upon the deposition of hematite as a cement around the sand grains; upon hematite existing in the grains themselves as in the cleavage cracks of feldspar, or in oxidized basalt and rhyo- lite; and upon the hematite resulting from the oxidation of magnetite grains in situ. 472 A. F. BUDDINGTON The pebbles of the conglomerates in the vicinity of Signal Hill consist roughly of about three-quarters rhyolite-porphyry and the remainder granophyre, intraformational red shale flakes, and quartz, with occasional pebbles of basalt. The rocks of which these pebbles are formed are very similar to the rocks comprised within the Avondale volcanic series, and it is quite possible that they have been derived from that formation. 1 In their stratigraphic relations, lithology, and mode of origin the Signal Hill series resemble the Jotnian of Sweden, the Torridonian of Scotland, and the Keweenawan of the "Canadian Shield." In the absence of vulcanism, so far as yet observed, they stand in contrast to the Keweenawan and Jotnian, but resemble the Torri- donian. This is of special interest, since it has been shown by Hayes and van Ingen that this district during the succeeding Cambro-Ordovician period had a history which coincided even in minor detail with that of Wales. INTRUSIVE ROCKS The later pre- Cambrian rocks, more especially the early later pre-Cambrian beds, are intruded by several batholiths and stocks of igneous rock and are very intensively cut by dikes. In general, it may be said that the salic rocks have been intruded as molten masses of batholithic or stock type, with few apophyses or dikes, whereas the older basaltic or gabbroic magmas, to a preponderating extent, have been intruded as dikes. The rocks, with the exception of the quartz syenite and granodi- orite, whose position is unknown, are given in the order of their intrusion, beginning with the oldest. Those around Conception Bay comprise hornblende granite gneiss, hornblende gabbro and plagioclase gabbro without olivine but locally quartz-bearing, basic granodiorite, biotite granite and granophyre, quartz syenite, aplite and granophyre dikes intruding the granite, and younger dikes of rhyolite porphyry and diabase intruding all the older rocks. Near 1 Murray describes the pebbles of the conglomerate at Signal Hill as composed "chiefly of white quartz, but with occasional pebbles of brown or red jasper, syenite or gneiss and slate." The writer examined 150 pebbles with the entirely different result given above. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 473 Clarenville on Trinity Bay a mass of gabbro is intruded by granite, and both in turn are cut by younger dikes of aplite, rhyolite, and porphyrite dikes. The hornblende granite gneiss shows a cataclastic texture and is believed to be the oldest rock in the district. Its outcrop was noted in only three areas of very small extent along the border of the Holyrood granite batholith. The gabbros occur as small, irregular patches on the borders of and within the area of the Holyrood granite batholith, and also as huge dikes in the Conception slate series. They are predominantly of a blackish or dark speckled gray color weathering to an ashen gray. In grain they vary from fine to medium, and pegmatitic facies are entirely lacking. Mineralogically there seem to be two types, which can be distinguished by microscopic methods alone: one characterized by hornblende without augite; the other an augite gabbro in which the pyroxene may be unaltered or partially altered to hornblende or uralite. Microscopically the augite gabbros are found to be pre- dominantly hypautomorphic-granular in texture, occasionally becoming ophitic. The augite and plagioclase may be fresh, or partially or completely altered; and the alteration of these minerals seems to have proceeded independently, for fresh augite may be associated with altered plagioclase, and vice versa. The plagio- clase may be partially altered to sericite, or flecked with sericite and partially replaced by quartz or epidote. The augite may be partially or completely altered to compact green hornblende or to an aggregate of uralite fibers or of uralite, quartz, and chlorite. The ilmenite is usually altered to leucoxene and is very small in amount. Pyrite is rare and magnetite relatively abundant. The peak of Holyrood Butterpot is formed of an elliptical mass of gabbro about 1,200 feet by 500 feet, completely surrounded by granite. It is peculiar in having quartz in small amount filling the interstices between the feldspars in such a manner as to indicate its origin as an original constituent, the last product of crystalliza- tion. Some of the augites are twinned, and others show idio- morphic basal sections. A chemical analysis of this rock is given on page 476, No. 2. 474 A. F. BUDDINGTON About six miles south of Manuels, lying between a series of green slates and the Holyrood granite, a more or less elliptical-shaped mass of gabbro forms a conspicuous ridge in the topography. This gabbro has a tendency toward a gneissic structure but without any particular evidences of dynamic metamorphism. It is quite inti- mately penetrated by ramifying apophyses of granite and by epidote veins, which are especially common along the joint planes. In thin section the rock is seen to be hypautomorphic granular to ophitic in texture and to consist essentially of light-green hornblende and labradorite. The feldspars are extensively altered to sericite, and some of the hornblende to intergrown fibers and blades of epidote, zoisite, and quartz. Many of the hornblendes are twinned parallel to the orthopinacoid, and by an occasional resemblance of crystal form give indications of an apparently primary origin. Whether this rock is the result of recrystallization during the intrusion of the granite or whether it is primary is difficult to say. A chemical analysis is given on page 476, No. i. In this connection it may be well to note that where the road crosses Seal Cove Brook there is an outcrop of hornblende porphyrite with phenocrysts of hornblende up to 10 mm. in length and plagioclases averaging 2-3 mm. in length. In thin section this rock is found to consist of idiomorphic phenocrysts of labradorite and hornblende in a groundmass in some places consisting of a microcrystalline, elsewhere of an almost crypto-crystalline, aggregate of plagioclase microlites, quartz, and an altered ferromagnesian mineral in sparse amount. The plagioclases show a zonary banding. To the west and south of Woodfords is a stock of granodiorite about two miles long and one-half mile wide, intrusive into the beds of the Avondale volcanic series. Dikes from this mass are excep- tionally rare, but enough evidence was found to prove its intrusive nature. The rock is light colored and medium grained, consisting of pink orthoclase, black hornblende, and quartz. The latter is abundant in small grains which show upon a fresh surface only when closely examined. On the weathered surface, however, they stand out in relief and give to the rock the appearance of a typical granite. With the microscope the rock is found to consist of andesine, ortho- clase, augite, and quartz, with abundant rods and grains of apatite. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 475 The colorless augite is almost completely altered to pleochroic green hornblende and chlorite. A chemical analysis of this rock is given on page 476, No. 3. Outcrops of a quartz syenite are found along the shore to the north of Chapel Cove. This rock has a very limited distribution, occurring in fault blocks brought into their present position by thrusting, and in places the rock is much crushed by the stresses to which it has been subjected. In the hand specimen the rock appears to be a white, medium-grained syenite, consisting of an aggregate of feldspars averaging 5 mm. in length. In thin section the rock is seen to consist of an aggregate of plagioclase feldspars, predominantly albite and to a minor extent oligoclase, with a groundmass of quartz and orthoclase in micrographic intergrowth filling the interstices. The rock is crushed and exhibits cataclastic texture in a high degree. A chemical analysis of this rock will be found on page 476, No. 4. The backbone of the St. John's Peninsula is mapped by Howley (1907) as Laurentian. This rock in its northern half was found by the writer to bear intrusive relations to the Avondale volcanics and to constitute a granite batholith five to six miles in width and about forty miles in length, if its boundary to the south continues to coincide with that of Howley's Laurentian area. The eastern boundary, where it was followed by the writer for several miles, is delimited by a fault plane. Along the west side the beds are very much disturbed and folded. The rock is a pink, medium-grained, equigranular, biotite granite, consisting essentially of quartz, orthoclase, albite, oligoclase, and chlorite, the latter believed to be derived from biotite. Fresh flakes of original biotite, however, are found only in occasional sections. A chemical analysis of the rock appears on page 476, No. 5. One of the most striking and interesting features of the geology along Smith and Random sounds at the head of Trinity Bay is the way in which the early green slates of that region have been rent, riven, and intruded by a great series of approximately parallel dikes of porphyrite and basalt. At .one locality there were counted 32 dikes in a distance of about 350 yards. As a rule, however, the dikes are not as frequent as this ; yet they are never absent for any 476 A. F. BUDDINGTON great distance. Their general strike is approximately north-south. About half the dikes are between two and five feet in width, and the rest are about equally divided between dikes ten to fifteen feet and those thirty or more feet in width. In addition to the dikes in the slate series several dikes and sills were found intruding the Cambro-Ordovician rocks, and dikes in the Signal Hill and Random formations are present but rare. The porphyrite dikes are variable in their mineralogical composition and comprise hornblende por- phyrites, labradorite porphyrites, and augite porphyrites or basalts. TABLE V CHEMICAL ANALYSES OF INSTRUCTIVE IGNEOUS ROCKS AROUND CONCEPTION BAY i 2 3 4 5 SiO 2 4.8 =C4 r i -IA irn ig 60 ^2 71 82 A1 2 O 3 2O 62 o x -OH- l8 Q4 18 21 16 85 16 07 Fe 2 O 3 r -2Q -2 20 2 73 Q7 I 22 FeO c c-i 6 4.0 1 62 i 6=: I 60 MgO 4 77 A cc 2 c;i 18 4O CaO 8 16 8 84 4 23 I 37 i 66 Na 2 O* 3 14 2 . 10 3 84 ^ OQ 3 27 K 2 O i . 23 I .OQ 3 26 2 7? 3 07 H 2 O+ 1.85 2.12 1.85 .89 71 H 2 O- *7 . II . 20 CX J 9 TiO 2 . 20 34 Undet. . 21 Undet. MnO 1 9 .29 .16 .09 13 Total 00 77 00 ^0 00 70 oo 66 100 14 1. Hornblende gabbro from about 6 miles south of Manuels. 2. Quartz bearing gabbro from Holyrood Butterpot. 3. Granodiorite from near Woodfords. 4. Quartz syenite from Chapel Cove. 5. Biotite granite from interior of St. John's Peninsula. The former are found in the pre-Cambrian slates, but only the last named in the Cambro-Ordovician beds. In the region about Conception Bay the granite of the Holyrood batholith, the gabbro, quartz syenite, and dikes of rhyolite porphyry and diabase are all overlain 'unconformably by Lower Cambrian sediments. Salic dikes were not found intruding beds younger than the Conception slate series, and femic dikes are far less frequent in formations above this series than in those below it and constituting it. Whether this is due to the different character of the formations, to their different stratigraphic position, or to their different age, PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 477 is an open question. What little evidence there is seems to point toward a period of batholithic and dike intrusion at a time succeed- ing the period of deposition of the Conception series and preceding that of the Signal Hill series. No dikes whatever were found intrud- ing the Cambro-Ordovician beds in the region of Conception Bay, but in the vicinity of Trinity Bay infrequent dikes of a distinctive character do intrude these beds and undoubtedly belong to a younger period of intrusion, possibly that of the Taconic revolution, or related to the period of vulcanicity represented by vast flows of lava in the Notre Dame Bay region about fifty miles to the north- east. This period of intrusion during the pre-Cambrian corresponds to a similar period of intrusion of granite into the volcanic series of the Blue Ridge of Maryland and Virginia described by eith (1892), or the Proterozoic batholiths of the Lake Superior region described by Allen (1915, p. 717) which he suggests are post-Middle Huronian or pre-Upper Huronian. Geographically, chemically, and physically the volcanics, plutonics, and dikes of the Conception Bay region appear to be genetically allied. Thus the gabbros, quartz-bearing gabbro, hornblende porphyrite with quartzose groundmass, granodiorite, quartz syenite, biotite granite, granophyre, and aplite form a series of rocks with an ascending silica ratio, belonging to the same geographical area and very probably to the same period of intrusion. Again, the analyses of the Holyrood granite, rhyolite porphyry near Holy rood, and a rhyolite flow from the Avondale volcanics near Manuels, constituting a batholith, a volcanic neck, and a volcanic flow respectively, show but slight variation in composition, although the texture varies widely. The chemical differences between the granite on the one hand, and the rhyolite porphyry and rhyolite on the other, are the usual relative increase in silica and loss of alumina, lime, magnesia, and iron in the volcanic facies. Again, the plu tonic gabbros, a volcanic plug of plagioclase basalt near Manuels, and the Blue Hills basalt flows may be cited as textural facies of a similar basic magma. The repeated recurrence and abundance of intru- sions of magma of basaltic composition in dikelike form throughout the entire peninsula, in contrast to the restricted local character of the salic intrusions with their associated infrequent dikes, rather 478 A. F. BUDDINGTON impress one with the idea that a basaltic magma was the one from which the salic differentiates originated. CONCLUSION The points brought out in this paper which it is desired to emphasize are as follows: 1 . The confirmation of the local glaciation of the Avalon Penin- sula, of the control of faults and fractures over the lineaments of the topography, and of the fiord nature of the coast. 2. The description and confirmation of the presence of a thick series of volcanics at the base of the pre-Cambrian section. 3. The presentation of evidence pointing toward the origin of the members of the later pre-Cambrian series, as follows : a) The Avondale volcanics as subaerial accumulations of vol- canic materials derived from volcanoes of the central type. b) The Conception slate series as materials derived from rocks resembling the Avondale volcanics swept into the sea in a com- paratively fresh, unaltered condition. c) The Momable formation as well-decomposed marine sedi- ments with traces of organic life. d) The Signal Hill reddish-brown sandstone series as dominantly subaerial fluviatile deposits in a subarid climate, derived from a re-working of volcanic rocks resembling members of the Avondale volcanics. 4. The evidence of a comagmatic series of igneous rocks intruded probably at some time during early or middle Proterozoic time. 5. The entire series from bottom to top is derived directly or indirectly from rocks of volcanic origin. The author wishes to acknowledge his indebtedness to Professors van Ingen and Smyth for their cordial assistance and advice rendered to him throughout the course of the work, and to Mr. B. F. Howell for his assistance during the summer of 1913. PRE-CAMBRIAN ROCKS OF NEWFOUNDLAND 479 BIBLIOGRAPHY Allen, R. C. "Revision of the Correlation of the Huronian Group of Michigan and the Lake Superior Region," Jour, of GeoL, XXIII, No. 8 (1915), pp. 705-24. Blackwelder, Eliot. "Summary of the Orogenic Epochs in the Geologic History of North America," ibid., XXII, No. 7 (1914), pp. 633-54. Buddington, A. F. " Pyrophyllitization, Pinitization, and Silicification of Rocks around Conception Bay, Newfoundland," ibid., XXIV (1916), 130-52. Daly, R. A. " Geology of the Northeast Coast of Labrador, " Harv. Coll. Mus. Comp. Zool. Bull., XXXVIII (1902), 205-70. Howley, J. P. Geological Map of Newfoundland, 1907. Jukes, J. B. General Report of the Geological Survey of Newfoundland during the Years 1839 and 184.0. London: John Murray. Keith, Arthur. "Geologic Structure of the Blue Ridge in Maryland and Virginia," Am. GeoL, X (1892), 362-68. Malcom, Wyatt. Gold Fields of Nova Scotia. Canada Geol. Surv., Mem. No. 20, 1912. Murray, A. and Howley, J. P. Geological Survey of Newfoundland. London: Edward Stanford. Walcott, C. D. " Pre-Cambrian Fossiliferous Formations," GeoL Soc. Am. Bull., X (1899). . "Random, a pre-Cambrian Upper Algonkian Terrane," ibid., XI (1900), 3-5. RETURN EARTH SCIENCES LIBRARY TO* 230 Earth Sciences Bldg.. 642-2997 LOAN PERIOD 1 14 DAYS 2 3 4 5 6 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS Books needed for class reserve are subject to immediate recall DUE AS STAMPED BELOW UNIVERSITY OF CALIFORNIA, BERKELEY FORM NO. DD8, 6m, 477 BERKELEY, CA 94720