r* Mlftre McINNES REPORT ON A PART OF THE NORTHWEST TERRITORIES X L< CANADA ■ DEPARTMENT OF MINES GEOLOGICAL SURVEY BRANCH Hon. W. Templbman, Minister; A. P. Low, Deputy Ministkb: R. W. Bbock. Director. REPORT ON A PART OF THE NORTH WEST TERRITORIES DRAINED BY THE WINISK AND ATTAWAPISKAT R1VEKS WILLIAM McINNES REPORT ON a TRAVERSE THROUGH THE SOUTHERN PART OF THE NORTH WEST TERRITORIES FROM LAC SEUL TO CAT LAKE IN 1902 BY ALFRED W. G. WILSON OTTAWA GOVERNMENT PRINTING BUREAU 1910 £^. /o2o CANADA DEPARTMENT OF MINES GEOLOGICAL SURVEY BRANCH Hon. W. Templeman, Minister; A. P. Low, Deputy Minister; R. W. Brock, Director. REPORT ON a part of THE NORTH WEST TERRITORIES OF CANADA DRAINED by the WINISK AND UPPER ATTAWAPISKAT RIVERS WILLIAM McINNES OTTAWA GOVERNMENT PRINTING BUREAU 1909 4074—1 3STo. 1080 Frontispiece. Plate i. Boulder ot green and reddish-purple Slate, Winisk river. Libran F woe To E. W. Brock, Director Geological Survey, Department of Mines. Sir,— I have the honour to submit a report on my surveys and explorations made during the seasons of 1903, 1904, and 1905, in the portion of the North West Territories of Canada drained by the Winisk river and by the upper branches of the Attawapiskat river. I heve the honour to be, sir, Your obedient servant, (Signed) WILLIAM McINNES. Ottawa, May, 1906. 4074 u N 801915 CONTENTS. Page. General remarks 7 Earlier explorations in the district 7 Surveys 8 Routes into the reeion 9 General description of the region 10 Geological summary 13 Direction of glaciation 21 Height of land region 21 Lower Winisk Paver region 21 Albany River and Attawapiskat River valleys 21 Winisk river 22 Attawapiskat river 35 Routes between the Attawapiskat and Winisk rivers 39 Routes between the Winisk river and Trout lake 41 Route between the Albany and Attawapiskat rivers 43 Cultivation of the land 44 Fish 45 Wild animals 45 Indians 46 Archaeology 49 Forests 49 Climate 51 Water temperature 52 Land shells 52 Fresh water shells 53 List of fresh water shells collected by W. Mclnnes 53 Index 55 ILLUSTRATIONS. Plate I.— Boulder of green and reddish-purple slate, Winisk river.. Frontispiece. " II.— Silurian limestone on the lower Winisk river 16 " III. — Treeless area fringing west coast of Hudson Bay 34 " IV.— Indians of the lower Winisk river 46 " V. — Lower Winisk river, showing banks of Silurian limestone, and characteristic forest 50 MAP. ]S T o. 1089. Portions of Northern Ontario, and of che North West Terri- tories, showing country drained by Albany, Severn, and upper Winisk rivers, etc. REPORT ON A PART OF THE NORTH WEST TERRITORIES OF CANADA DRAINED BY THE WINISK AND ATTAWA- PISKAT RIVERS. BY William McInnes. The present report deals with a tract of country lying within the unorganized North West Territories of Canada, between N. lat. 51° 10' and K lat. 55° 10', and between W. long. 86° and W. long 90°. This district forms part of what was known for a time, prior to the inauguration of the Provinces of Alberta and Saskatchewan, as the District of Keewatin, and lies between the northern boundary of Ontario and the southwestern shore of Hudson bay. It is drained by rivers running from the west into James bay and into Hudson bay respectively, and the report is, in the main, a description of one of the latter — the Winisk — throughout almost its entire length, and of the upper branches of one of the former, the Attawapiskat. Earlier Exploration in the District. As far as I have been able to learn, there are no references in the journals of the early explorers to the Winisk river. All concerned in the search for a northwest passage to the Orient, they were natur- ally led to give most of their attention to the passages between the Arctic islands lying at the extreme north end of the bay. The mouth of the Severn river was, however, visited by a number of them, and Henry Hudson and Thomas James explored the bay now known as James bay, then called Hudson's bays. Captain Thomas James, and Captain Luke Foxe (who styles himself in his journal, 'the northwest fdx'), seem to have been the only navigators who sailed along the coast between the Severn river and Cape Henrietta Maria, for the purpose of examining it. They describe a generally low shore, with shallow water, and make no allusion to having noticed the mouth of the Winisk river. It must 7 8 GEOLOGICAL SURVEY, CANADA have been, however, as Mr. Miller Christy points out, in the vicinity of the bay at the mouth of the Winisk river that the two vessels approached one another in August, 1631, when the two captains, both bearing letters from His Majesty King Charles I to the Emperor of Japan, were able to compare notes as to their discoveries, and when Captain Foxe, ridiculing James' action in keeping his flag continually flying at the masthead, said to him, to use the quaint language of his journal, ' Keepe it up then,' quoth I, ' but you are out of the way to Japan, for this is not it.' Mr. G. Taylor, of the Hudson's Bay Company's service, seems to have visited the river in 1808, and to have supplied the topographical details that appear on the Arrowsmith map. Dr. Kobert Bell, in 1886, descended the Attawapiskat river from the lake, which he named Lansdowne, to the sea, and published an account of the exploration in the Annual Report of the Geological Survey for that year. 1 The Fawn branch of the Severn river was explored by Dr. A. P. Low in 1886 2 and the Ekwan and Trout rivers by Mr. D. B. Dowling and Mr. W. H. Boyd in 1901. 3 No description of the Winisk has been published, though, without doubt, employes of the Hudson's Bay Company have traversed it, as, in the early part of the last century, posts of the Company were established at three points near the head of the river. The missionary priests from Albany, too, have descended the river, holding missions at the more important Indian centres. Surveys. In order to secure data for the compilation of a map of the region, the following surveys were carried out during the seasons of 1903-4-5 :— Surveys by micrometer telescope and compass, checked by astro- nomical observations for latitude, were made of the Winisk river, from the mouth to a point 190 miles from the coast following the course of the stream; from the foot of Wunnummin lake up to the outflow of the west branch at Misamikwash lake, a distance of 60 miles; down the west branch for 55 miles, and across by a portage route 24 miles in length to Trout lake at the head of the Fawn i Annual Report Geological Survey of Canada (New Series), Vol. II., 186G, Part G. 2 Ibid. Part E. 3 Summary Report Geological Survey of Canada, 1902. REGION OF WIN1SK AND ATTAWAPISKAT RIVERS 9 branch of the Severn river; of a route from the foot of Lake St. Joseph by way of the south branch of the Attawapiskat river to Fort Hope, a distance of 189 miles; and of 27 miles of the Albany river below Fort Hope. In addition to the above a number of track surveys, checked by latitudes, were made. These covered portions of the Winisk river; part of the Attawapiskat river; three routes connecting the Attawa- piskat and Winisk rivers ; a route from the Albany river at Eabemet lake to Lansdowne lake; and a route from Trout lake down the west branch of the Winisk river and across to the main river near Nibina- mik lake. Routes into the Region. While the number of possible routes to the Albany river from the Canadian Pacific railway is very great, there are but three that have been used to any great extent, one leaving the railway at Dinorwic station and reaching the Albany river by way of Lac Seul and its tributary the Root river, another one starting from Ignace and reaching the Albany by way of Sturgeon and Musibimega lakes, and another leading from Nipigon station by Nipigon river and lake and crossing to the Albany by way of the Ombabika and Opichuan rivers. The first of these is the best route in, particularly where a load is to be carried, as, though somewhat longer than either of the others, it is down stream or through large lakes for the greater part of the distance. For light canoes and a quick passage the route by way of Nipigon is preferable, on account of the shorter distance to be traversed. The greater part of the supplies used for the fur trade in the district are brought up the Albany river from James bay, a route including 300 miles of swift water where tracking is the only means of progression, and about 50 miles of alternating quiet water and rapids where portages are frequent. This is considered an easier route to Fort Hope, the headquarters of the trade, than any of the roads from the Canadian Pacific railway. The completion of the Grand Trunk Pacific railway will shorten very considerably the distance from this side, and render the whole region comparatively easy of access. From Fort Hope the heads of the Winisk and Attawapiskat rivers can be reached by several routes, none particularly difficult, but all made tedious by reason of the number of portages necessarv. 10 GEOLOGICAL SURVEY, CANADA General Description of the Region. The region may be roughly divided into three great areas, each with characteristic features: the Archaean area of the high interior plateau ; the boulder clay area ; and the limestone area of the Hudson Bay basin. The Archaean, of the three, comprises by far the largest extent of country. It consists of an elevated, undulating plain, with an average height of from 700 to 1,000 feet above sea-leveL The effects of long-continued subaerial decay and denudation, supplemented by the later cleaning up and smoothing action of a great glacier, are everywhere noticeable in the gently rounded out- lines of the very moderate elevations. On it all the larger rivers of the Hudson Bay watershed, and many of those flowing south and west, have their sources, the great muskeg areas acting as storage reservoirs, from which, even in the dryest season, the volume of drainage is large. It is along the parts of their courses lying within this area that the quickest descent occurs, falls and rapids that would afford water-powers being thus largely confined to the upper stretches of the streams. This condition is in contrast with that obtaining everywhere throughout eastern Canada, where the streams flow for the greater part of their length over the Archaean, and only come tumbling down from the elevations when low down in their courses after they have attained almost their maximum volume, thus making the eastern portion of Canada probably unequalled in the world in the matter of water-powers. It must not be thought, how- ever, that throughout the area now under consideration there is any scarcity of good water-powers. They occur in great number, but owing to the distribution of the Archaean highland before referred to, they are situated mainly far inland rather than near the coast. Though, considered as a whole, the central, elevated region cannot be spoken of as generally adapted for agriculture, there occur basins covered by heavy deposits of stratified sand and clay that seem to have been laid down in lakes held in between barriers formed by the walls of the retreating glacier and ridges of drift. An examination of some of these clays by Dr. Hoffmann shows them to be highly calcareous and somewhat siliceous, a composition that with the admixture of the surface vegetable mould should produce an excel- lent soil for general agriculture. The question of climate, which is, of course, of the utmost importance when considering the agricul- tural possibilities of a district, will be referred to more particularly REGION" OF WINISK AND ATTAWAPISKAT KIVEBS 11 iii another place. It may be said here, however, that the climatic conditions are, if somewhat adverse, not by any means prohibitory to the general cultivation of suitably situated tracts. Muskeg, alternating with low ridges of gravel and boulders, covers wide tracts, though, owing to the fact that the only practicable mode of travel through the country is by canoes, there is a tendency, perhaps, to overestimate the extent of such areas, as the natural canoe routes must follow the watercourses, and these in turn keep to the lowest elevations, and, therefore, show a proportion of swamp that is greater than the average of the district. It was noticed that the surface drainage became more perfect in that part of the region extending westerly towards Trout lake. Ascending the Winisk river from Weibikwei lake towards its headwaters this was very noticeable, the muskeg areas becoming infrequent and of smaller extent. The larger lakes throughout the district are confined to the Archaean area. They are all comparatively shallow, and so studded with islands, and broken by long, projecting points, that they seldom show any large expanses of open water. They occupy depressions in the superficial deposits, generally with a boulder clay bottom, and in no case was one found occupying a regular rock basin. The areas of the principal lakes are approximately as follows: — Wunnummin lake 60 square miles. "Weibikwei lake 40 " Lansdowne lake 38 " Ozhiski lake 25 " Wapikopa lake 24 " Eabemet lake 20 " Nibinamik lake 10 " The highest land lies about the headwaters of the south branch of thje Attawapiskat river, east of Cat lake, where an elevation of probably 1,500 feet above the sea-level is reached. The approximate heights of the principal lakes determined by barometric measurement is given below: — Eabemet lake, Albany river 900 feet above sea-level. Ozhiski lake, Attawapiskat river 910 Lansdowne lake, Attawapiskat river. . . . 815 Wimbobika lake, Attawapiskat river. . . . 1,300 " Weibikwei lake, Winisk river 670 " " 12 GEOLOGICAL, SURVEY, CANADA Wapikopa lake, Winisk river 750 feet above sea-level. Nibinamik lake, Winisk river 785 " Wunnummin lake, Winisk river 830 Misamikwash lake, Winisk river 865 The tract referred to as the boulder clay area consists of a broad belt of country, about 159 miles in width, lying between the Archaean highlands and the edge of the limestones of the basin of Hudson bay, overlapping the latter, however, so that the surface features of the two are generally quite similar. Gently undulating, and with a slight slope northerly and easterly, its general surface aspect is that of a great swamp, sparsely covered with stunted and deformed trees, that reach a growth approaching their normal only along the immediate banks of the rivers where drainage is afforded by frequent short gullies into the trenches that constitute the river valleys. The interior, to within a chain or two of the river-banks, owing to the impervious character of the till, is quite undrained, and consequently covered by a thick deposit of sphagnum moss from two feet to ten feet deep, the surface layer still growing, and even the bottom only bleached a little, but not at all oxidized. The short cool summer season, and consequent low tem- perature of the water that saturates the moss, is probably the prin- cipal reason for the absence of any of the visible effects of decay. The rivers flowing through this region have no real valleys, that is to say, they occupy trenches but little wider than the immediate channels in which they flow, cut down through the stiff, tough till, which stands up in nearly vertical walls that rise from the freshet mark on either side. At low stages of the water a slanting beach, often paved with boulders, slopes gradually from the foot of the bank to the edge of the diminished channel. A more or less continuous layer of marine clay, rich in fossil shells, overlies the boulder clay, ensuring, wherever it is present, a soil of good quality. The absence of other than swamp vegetation must be ascribed, then, to the almost total absence of drainage, and to the generally unfavourable climatic conditions. The third area, underlain by Silurian limestones and dolomites, presents essentially the same surface features as the till area. The folding of the limestones, however, though generally amounting to broad undulations only, gives to it somewhat more of relief, and the troughs in which the rivers lie have been excavated entirely through EEGION OF WINISK AND ATTAWAPISKAT EIVEKS 13 the mantle of till, and have cut down into the limestones to depths of from twenty to thirty feet. There is the same absence of any vegetation other than that having a muskeg habitat, excepting on the islands in the rivers and along their banks. The northern rim of this area consists oi a treeless plain, border- ing the shores of the bay, and varying in width from a mile and a half to three miles. It has an elevation of only a few feet above the level of high, spring tides, and is probably submerged on occasions when these tides happen to coincide with northeast storms on the bay. The sandy and gravelly surface is sparsely covered with bunchy grasses, and, early in August, was bright with the flowers of many sub-arctic plants, among which the Arctic daisy, Chrysanthemum arcticum, the yellow ragwort, Senescia pallistris, the painted cup, Castelegia pallida, a live-for-ever with small, bell-like blue flowers, purple vetches, and the large rose-coloured Epilobium were promi- nent. Geological Summary. The geological divisions recognized in the region under considera- tion consist of the following, in ascending order: — Laurentian. Keewatin, Lower Huronian ( ?), Silurian (Niagara), Pleistocene (Till, etc.), Post-pleistocene (Marine clays, etc.), LAURENTIAN. Biotite granite gneisses, varying in the proportion of their various constituents, in their attitude, and in the degree to which the gneissic structure has been developed in them, are widespread over the whole extent of country explored. Over great areas they have a stratiform appearance, the foliation showing an almost horizontal structure, with only very low, broad undulations. As at present constituted they, without doubt, include areas that differ widely in age, the com- paratively new granites, however, occurring in quantity quite insignificant in comparison with the volume of the older gneisses. Pegmatites, in veins and irregular masses, cut the gneisses practi- cally everywhere, and are, probably, though newer than the gneisses, almost contemporaneous with them in their present form. 14 GEOLOGICAL SURVEY, CANADA KEEWATIN. The Keewatin bands, made up of areas of basic rocks, in the main diorites, diabases, and chloritic and hornblende schists, but including a considerable volume of coarse conglomerates, though occurring as belts of considerable length and four to six miles in width, are of exceedingly small volume when compared with the whole extent of gneisses in which they are enfolded. Probably not more than a tenth of the whole Archaean area is occupied by them. In the region explored, between the Albany river and the overlap of the mantle of till, six apparently separate belts of these rocks were noted. They have all, in a general way, about the same trend, N. 70° E. The belt of these rocks crossing the Albany river at Petawanga lake and seen again on the route between the Albany at Fort Hope, just north of Eabemet lake, is the most southerly. It is made up for the most part of chloritic, feldspathic and hornblendic schists, and diorites in different stages of deformation, and has a width of about six miles. The gneisses bordering the belt on the south are finely foliated, hold a large proportion of black biotite, and are, in certain layers, thickly spotted with garnet crystals. Masses of coarse pegma- tite, cutting these gneisses, hold crystals of mica up to 2" in diameter. The next belt going northerly is situated about twenty miles north of the Albany river, and is well exposed along the banks of the Kawinogans river, which has cut its channel in these rocks for about seventeen miles. This band is from one to four miles in width, and is made up of feldspathic and chloritic schists, diorites and other basic rocks. It is flanked by biotite gneisses, with, at points close to the contact, occasional outcrops of hornblende granite-gneiss. Another belt, quite similar to the two above referred to, lies just north of Lansdowne lake. Further reference is made to it in the descriptions of the routes leading north from the Attawapiskat to the Winisk. The most interesting belts are the next two; the first, lying just south of Nibinamik lake, by reason of the occurrence in it of a large mass of hypersthene gabbro, similar to the nickel-bearing in- trusives of Sudbury; and the next, the Wunnummin' Lake band, on account of the extensive development in it of heavy beds of coarse conglomerate, holding pebbles, chiefly of various forms of granite. The most northerly band is apparently quite narrow, and was noted REGION OF WINISK AND ATTAWAP1SKAT RIVERS 15 only where a few isolated outcrops are seen near Kingfisher lake north of the Winisk river. SILURIAN. The Silurian section along the Winisk river seems to comprise, in ascending order, twenty feet of close-grained, hard, brittle, green and black ribboned slates, with bands and nodules of more highly calcareous material; six feet of a hard, dark-grey, rusty weathering, calcareous quartzite; ten feet of a much more calcareous form of the last named beds, so calcareous as to constitute an impure lime- stone rather than a quartzite. All of these lower beds, which are exposed at but one place on the river, where they are brought up by a compound anticlinal fold, are hard and baked-looking, with many small veins of quartz and calcite cutting them in all directions. Lithologically they are quite dissimilar to any of the strata compos- ing the rest of the section. Further effects of pressure are seen in the hardened condition of all the rocks, and in their cracked and fissured condition, the cracks filled with secondary quartz and cal- cite. The more massive beds described as calcareous quartzites are seamed in all directions by these white, reticulating veins, which are bought into strong prominence by their contrast in colour with the dark, rusty- weathering surfaces of the parent rock. There seems to be a gradual passage upwards from these beds, by the increase in their calcareous content, into impure limestones, and then into the next beds in the series, consist- ing of a series of slightly magnesian limestones, comprising eight feet of buff-coloured, slightly ferruginous, hard, close-grained, flaggy beds, with the texture of lithographic stone in certain layers; two and a half feet of more massive nodular limestone, the nodules of finer texture than the enclosing, slightly shaly matrix; and ten feet of rubbly, shaly limestone, with occasional sandy layers. Nodules of bluish opalescent quartz, with banded, agate-like structure, occur in the more compact beds throughout the series. Above these beds is a very persistent band, six feet in thickness, of a tufaceous-looking, vesicular limestone, the very distinctive char- acter of which makes it easily recognizable at many points along the river. Cavities in it are coated with crystals of calcite, and vesicles and cracks occurring in it are filled with a fibrous form of that mineral. The calcite occurs throughout the rock in irrregular masses that weather out to form cavities of irregular sizes and shapes. 16 GEOLOGICAL SURVEY, CANADA Immediately above this bed there occurs a 6" layer of a compact yellowish limestone, with but little magnesia; then six feet of very fine-grained, almost compact, very light buff coloured dolomite, containing a small quantity of argillaceous matter and occurring in heavy flag-like beds, tbe plates hard and clinking under the hammer. These are overlaid by twelve feet of buff-coloured impure magnesian limestone, shaly in certain layers; and ten feet of flaggy and shaly buff-coloured, somewbat nodular, magnesian limestone, the whole becoming disintegrated easily so as to show only nodular, crumbling surfaces. Broadly speaking, the strata may be said to lie almost hori- zontally, with a slight dip towards the shores of the bay, amounting to about the same as the descent accomplished by the river. Low undulations cause the same beds to recur again and again in the sections exposed along the river. The exposures are not continuous, long intervals where the overlying boulder clay only is seen inter- vening between the exposed sections, so tbat the generalized section given above, and tabulated on a succeeding page, is made up from an examination of separated exposures occurring along the river for a distance of eighty miles. Though the strata are uniformly buff- coloured and closely similar in general appearance, a few very dis- tinctive beds — notably the tufa-like limestone bed, which seems to be very persistent and to keep its distinctive characteristics — serve to connect the various exposures satisfactorily. The lowest beds, com- prising the thirty-five feet of strata brought up by the compound anticlinal fold, appear at only one place on the river. As no fossils were found in them their age can be inferred only from their appar- ently conformable position immediately underneath the fossiliferous Silurian strata. The corrugated surface of the dome of the anticlinal itself dips about ten degrees north of west, at a low angle varying from five to twenty degrees, and it is possible, though not probable, that the rocks noted by Mr. Dowling at Sutton Mill lake represent underlying beds brought up by a southeasterly extension of this fold. The calcareous nodules, which probably represent bands broken by the stress of the folding, weather out readily, where exposed to atmospheric action, leaving a rock full of holes. 4071— p. 16. REGION OF WmiSK AND ATTAWAPISKAT KIVEKS GENERALIZED SECTION ALONG THE WINISK RIVER. 17 in ft. 12 ft. 6 ft. 6" 6 ft. 10 ft. 2'-6" 8 ft. 10 ft. fi ft. 20 ft. T T » . I ^-r 1 rJ-rW 1 EJ. L T T~rr < l — T~-r- r -r ■ r It . -U-U I , ! ■ if fr^ ~T~T '. ' . i I " T^ "^Stt^Et ■ I *-!--.■ /5J 1^, ! 1 . _L 1 [ — r E^LXTTI f^ Flaggy and shaly, buff magnesian limestone. Shaly, impure, rusty-weathering, magnesian limestone. Flaggy, clinking dolomite. Compact, slightly magnesian limestone. Vesicular limestone. Slightly magnesian, shaly limestone with sandy layers. Slightly magnesian, nodular limestone. Slightly magnesian, clinking, flaggy limestone. Rusty-weathering, siliceous limestone. Dark grey, rusty-weathering, calcareous quartzite. Black and green, ribboned slates. i07l- 18 GEOLOGICAL SURVEY, CANADA A small collection of fossils was made from the beds overlying the vesicular band, in which Dr. Whiteaves has identified the following forms : — • Favosites gothlandica, Lamarck. Stropheodonta niagaraensis, W. and M. Leptcena rhomboidalis, Wilckens (sp.). *Camarotcechia (?) winishensis, Whiteaves. *C 'amar otoechia (?) coale^cens, Whiteaves. "Glassia variabilis, Whiteaves. *Actinocei'as Jceewatinense, Whiteaves. Trimerella, sp. indet. Orthis, " Spirifer, " Streptelasma, " Trochonema, " Euomphalus, " Loxonema, " Cyrtoceras, " Bronteus, " Encrinurus, " Though not a very satisfactory collection in itself for purposes of age-determination, the above-named species correlate the beds hold- ing them with those of the Severn river to the north, and the Ekwan river to the south, and collections from the three localities combined fix the age of the rocks very satisfactorily. The southern limit of the Silurian limestones cannot be fixed with any degree of exactness, owing to the heavy overmantle of till that conceals from view the underlying rock for a distance of 130 miles along the river. It seems probable, however, that it extends to the vicinity of N. lat. 54° 20'. Mr. Low found on the Fawn branch of the Severn, the nearest river to the west, the same wide area of country completely covered by till intervening between the most northerly exposure of gneiss and the first exposure of limestone. He thought it probable that the limestones extend under the till for a distance that would correspond very closely to that given above for the Winisk. East of the Winisk river the inland boundary of the Silurian bends suddenly to a direction nearly due south, crossing the * The two new species of Oamarotcechia, the Glassia and the Actino- ceras have been described by Dr. Whiteaves in Palaeozoic Fossils, Vol. III., Part IV., 1906, where further notes concerning the collection will be found. REGION OF WINISK AND ATTAWAPISKAT RIVERS 19 Attawapiskat river a little above N. lat. 52° 30', and the Albany one degree lower. PLEISTOCENE. The boulder clays of the Winisk river may be easily divided into an upper and a lower till, the one lying upon the gently undulating surface of the other. The upper bed is composed of a buff-coloured clay, drying slightly friable, with occasional large boulders, and many small pebbles and angular fragments of diorite, quartzite, gneiss, red and white sand- stone, jasper, etc. Its greatest observed thickness is about forty feet, measured from the surface of the lower till to the bottom of the fossiliferous marine beds. Xo stratification is apparent in it, and the large boulders are so rare, that, at a little distance, cut faces have the appearance of beds of pure clay. The lower till, the thickness of which was not ascertained, is composed of an extremely tough blue clay, with very many large boulders, semi-rounded and mostly well striated. Limestones and dolomites quite similar to the Silurian beds of the lower river make up a large proportion of the boulders, but others of gneiss, quartzite conglomerate, etc., are not uncommon. The sloping beaches extend- ing between low and high water marks are often a mosaic of the washed out material from the clay, forming very good examples of boulder pavements, the natural tendency of the rocks to arrange themselves with their flatter sides parallel to the surface resulting in an almost smooth floor, over which the spring floods seem to pass with little or no denuding power. The whole bed of the river is, in the same way, protected by a layer of heavy boulders that offers great resistance to the wear of the current, and that has practically stopped the further excavation of the channel at levels far from the bottom of the lower till. The accumulations of glacial drift are an important feature over this whole district. They form the highest elevations, and are the principal causes that define the shapes of the lakes and the direc- tions of the rivers. The influence of morainic ridges of boulders and gravel on the course of a river is strikingly seen in the case of the upper part of the Winisk river. The direction of the ice movement was about S. 23° W., and the course of the river is found to conform to this direction to a remarkable extent, that is, it makes its way 4074— 11 20 GEOLOGICAL SURVEY, CANADA eastward in a series of zig-zags, the lake-like expansions conforming in a remarkable way to the course of the morainic ridges of drift. The lakes occurring along the river are characterized by many- long narrow bays with the same trend, due to the drift ridges that bound them. The glaciation of the whole area shows most clearly that it is the result of the passage of a large glacier, continental almost in extent, moving in a general way a little south of west, but showing minor deflexions, that occurred probably at stages in the period of glacia- tion when the ice sheet was not at its greatest thickness and was more readily influenced by the surface contours. The general S.S.W. direction of movement is indicated not only by striae, chatter marks, and crag and tail sculpturing, but also by the character of the boulders enclosed in the till and scattered broad- cast over the Archaean area. The occurrence of the fossil-bearing limestones along the west coast of Hudson bay and James bay, and the entire absence of any rocks at all similar to them over the whole region farther south, makes the character of the travelled boulders derived from these rocks a sure index to the direction followed by the moving ice-sheet. Additional evidence is afforded by the occur- rence in the till of boulders and pebbles of jasper, hematite, quart- zite of a very distinctive character that Dr. Bell has recognized in place on the east coast of Hudson bay, and jasper breccia or con- glomerate. The wide tract of country lying between the Archaean gneiss and the first exposures of limestone, where the underlying rocks are completely concealed by the thick mantle of boulder clay, might be the source from which is derived many or all of these apparently foreign boulders, but their very close similarity to rocks that are known to occur on the east shore of Hudson bay makes it more probable that they have been derived from them. A few southwesterly striae that appeared to be possibly later than the prevailing ones might be interpreted to indicate a glacier travelling down a gathering ground such as has been assigned to the Keewatin glacier. The local variations of the striae from the general direction are so many, however, that it seems quite possible that they are only the records of deflexions caused by local surface relief, and made perhaps by a very much reduced glacier. No evidence of a glacier moving down towards the bay was noticed. The following list of glacial striae is arranged under three divisions — the height-of-land REGION OF WINISK AND ATTAWAPISKAT RIVERS 21 region where the striae may be considered to represent most truly the general course of the glacier, the Winisk Eiver channel where the direction of the striae seems to have been somewhat affected by the river course, and the valleys of the Albany and Upper Attawapiskat rivers, where the direction has been quite governed by the trend of the valleys. Direction of Glaciation. Height-of-Land Region — Kawinogans river S, 50 W. Hail lake g. 40° \\r Wapitotem river S. 38° W. Winisk river, eight miles above Weibikwei lake. S. 38° W. Winisk river, Wapikopa lake S. 32° W. Lower Winisk River Region — Winisk river below outflow of Winiskisis S. 6° W. at outflow of Tabasokwia S. 6° W. above Tashka rapid S. 10° W. " at Tashka rapid S. 30° W. at Boskineig fall S. 30° E. i mile below Boskineig fall .... S. 18° W. 1 " " .... S.10° E. " 2 miles " " . . . . S. 12° E. 8 " " . . . . S. 10° E. 13 " " .... S.24 E. 15 " " .... S. 26° E. Albany River and Attawapiskat River Valleys — Eabemet lake, northwest shore S. 83° W. " north shore S. 78° W. Albany river 10 miles below Eabemet S. 68° W. 12 " « S.67°W. 15 " « S.64°W. Ozhiski lake "W, Kabania lake N. 79 "vV. POST-PLEISTOCEXE. The marine clays, overlying the boulder clays along the Winisk river, were found to be generally fossiliferous, excepting near their most southerly extension where they are quite thin, and, as far as 22 GEOLOGICAL SURVEY, CANADA observed, do not hold fossils. From a collection made from these clays in 1903, Dr. J. F. Whiteavcs has identified the following species : — Pecten islandicus, Miiller. Mytilus edulis, L. Cardium ciliatum, Fabricius. Serripes Grcenlandicus, Ganelin. Macoma calcarea, Gmelin. Mya truncata, L. Mya arenaria, L. Saxicava rugosa, L. Buccinum tenue, Gray. Buccinum ? and, fresh water species: — Sphcerium striatinum, Lamarck. Limnoea palustris, L. The Winisk River. The Winisk river, though without falls in its lower course, and with a volume that would lead one to suppose it easily navigable by vessels of considerable size, is so rapid and so wide for a long dis- tance up from the bay that it would be difficult to find a channel for a steamer of even moderate draft. This is particularly true of the thirty miles of its course over the flat-lying limestone ledges that often form barriers quite across the river bed, on which there is a depth of only a few feet of water. The river has cut down into the limestones to a depth of more than forty feet, the strata rising in vertical walls to that height above mean low water level. There is evidence that the river followed its present channel in the limestones prior to the glacial period. It has since then not worn out for itself any valley beyond its immediate channel, which is a mere trench in the boulder clay in the upper stretches, and in the clay and underlying limestones farther down. The extreme tough- ness of the lower boulder clay, and the protection afforded by the great number of large boulders thnt wash out from it and coat the bottom and lower parts of the sides of the trench, have prevented any quick degradation of the banks, which stand up, raw and steep, like the sides of a newly excavated canal or railway cutting. The REGION OF "WINISK AND ATTAWAPISKAT RIVERS 23 more gently sloping parts of the bank, between high water mark and the foot of the boulder clay wall, are covered with a growth of grasses and small bushes, and, beyond latitude 54° 30', the nearly vertical boulder clay itself supports a growth of silver berry, Eleagnus argen- tea, and buffle berry, Shepardia, the almost snow-white foliage of the former standing out in strong contrast with the dark-green leaves and red berries of the latter. The Winisk river, along its upper course easterly to Weibikwei lake, has a distinguishable valley. The lower part of the river, how- ever, from the lake to the sea, has absolutely no valley outside of the steep-walled trough in which it runs. The upper Attawapiskat river, flowing in an easterly direction, has a fairly well-marked valley, comparable to that of the Albany, though of less extent. The upper parts of the river are roughly parallel to one another and to the Albany river, with which it is not at all improbable that the Attawapiskat was at one time connected, as the country now dividing them is characterized by high hills of glacial drift, filling up and concealing any former channels that may have existed. These are the very remarkable hills described elsewhere in this report in greater detail. In all the rivers on this slope is seen the tendency to split up into two or more channels, enclosing areas of land often many miles in extent. This feature is more marked in the case of the Winisk than in any of the others. Above Weibikwei lake one of these divisions of the channel occurs, enclosing an area of thirteen square miles; and below, the two branches known as the Winiskisis and the Taba- sokwia flow around islands with areas of about 480 and 180 square miles respectively. The former of these branches, flowing to the east at a point seven miles below the lake, joins the main river again sixty-five miles below. The Indians say that no important stream comes in to the branch, but a number of small streams makes it a river of considerable size at its confluence with the main channel, even at low water when no water is passing over the bar at its upper end. The volume of water in the river during the period of spring freshet must be quite ten times as great as at low water in mid- summer. The height reached by the water is, in many places, plainly indicated on the banks. Evidences of the destructive force of the ice, when running out in the spring, are common. Trees on some of the islands are found 24 (JKOLOGICAL SURVEY, CANADA broken and uprooted at heights of fifteen feet above the normal water level, and the boulder clay of the banks is ploughed and deeply scored at corresponding heights. The flat surface of the limestone bordering the gorge is evi- dently swept annually by the river when at its height, though the water surface in the gorge at ordinary summer level is thirty feet below the top of the limestone. No beds of lignite were observed, though a few highly carbon- aceous, sandy layers were seen to occur at water level, apparently beneath the boulder clay on the upper Winisk river. Owing to the frequent small landslides occurring along this part of the river it was impossible to fix the position of these beds with any degree of certainty. Though for so great a part of its course the river is bordered by high and steep banks of clay, landslides seem to be exceedingly rare, excepting where the country has been swept by forest fires. Where fires have recently taken place along the banks, denuding them of their protecting vegetation, small landslides are almost continuous. The Winisk is with little doubt the largest of the rivers discharg- ing into the west side of Hudson bay or James bay between the Severn and Albany rivers. Rising in the highlands lying to the south of Trout lake, it drains the large expanse of country lying to the east of the upper waters of the Severn river, and to the north of the spreading branches of the Attawapiskat. The watercourses of this section of country have been most inadequately represented on the existing maps, owing to the lack of knowledge of their positions, and a reference to the map accompanying this report will be neces- sary in order to understand the apportionment of the watersheds among the various rivers. From Misamikwash lake, above which the Winisk is divided into two main and many smaller branches, the river flows out by two channels, one quite insignificant in volume flowing to the north, and the other, a river of considerable size, flowing to the east. The former of these forms the head of the Asheweig or west branch of the Winisk, and the latter the main river. Diverging at a point situated in N. lat. 53° and W. long. 90°, these two streams unite 224 miles below, following the course of the main river, in N. lat. 54° and W. long. 87° 30'. From Misamikwash lake for twenty-five miles the river keeps a general easterly course. In this distance the descent is about thirty- REGION OF WINISK AND ATTAWAPISKAT K1VEUS 25 five feet, and occurs principally in a series of five rapids, at the lowest of which, just above Wunnummin lake, there is a very con- siderable fall. Between the rapids are stretches of swift water, varied by many lake-like expansions. The surrounding country is for the most part low, seldom rising to greater heights than fifty feet above the river. Few rock exposures are seen, what there are consisting of low, rounded knolls and ridges of well foliated biotite granite gneiss, generally with an almost horizontal foliation and often invaded by a coarser white granite or pegmatite. The banks are usually low, but in places the river is found impinging against a bank of im- stratified sand and gravel twenty to thirty feet in height. Below the rapid and fall just referred to a large stream comes in from the south, and the river widens out to form Wunnummin lake, a body of water of varying width, twenty-five miles in length. The trough in which the lake lies has been hollowed out mainly in a band of Keewatin rocks to whose trend it generally conforms. The most conspicuous rocks occurring in the belt are heavy beds of coarse conglomerate, very similar to that of Abram lake on the English river below Minnitaki lake. 1 With these are associated diorites and chloritic and hornblende schists, the whole striking about N. 70° E. and dipping at high angles. These rocks can, without doubt, be classed almost wholly with the Keewatin, though there are possibly small areas of lower Huronian, the basal beds of which would be represented by the conglomerate. About the lake almost the only eminences in view are low hills of unassorted drift, rising generally not more than fifty feet above the water level, but in one case forming a very striking cone-shaped eminence, rising perhaps 300 feet above the surrounding level. Owing to its inaccessibility this hill was not visited, but from its general aspect, and from the accounts of it given by the Indians, it evidently is one of those remarkable, isolated masses of drift seen on the south branch of the Attawapiskat, and noted also by Mr. Camsell as occurring in the country north of Cat lake. 2 From Wunnumin lake to Nibinamik lake, a distance of twenty- five miles, the descent is about forty-five feet, the fall occurring principally at three points, where series of heavy rapids break the course of the river. Between these are stretches of quiet flowing i Annual Report Geological Survey, 1901, Vol. XIV, p. 90 A. - Summary Eeport Geological Survey, 1904. 26 GEOLOGICAL SURVEY, CANADA water, where the current, though generally strong, flows along pla- cidly between banks of sand not generally high, but in places, where the current has worn into the side of a drift ridge, showing cut banks seventy-five feet in height. A stream known as Michikenis flows in from the south about six miles below Wunnummin lake, and a larger one, referred to again in describing the route from Trout lake, joins the river from the north five miles above Nibinamik lake. Nibinamik lake is an irregular body of water whose shape has been largely defined by ridgea of glacial drift. From inlet to outlet is but five miles, the lake, however, extending to the south for seven miles and to the north for four miles. A number of low ledges of fine, well-foliated biotite gneiss occur along its shores, cut by a coarse white gneiss that often is interbanded with the finer, giving the whole an appearance of stratification. The land rises gradually from the lake shores to heights of about sixty feet, a considerable thick- ness of sand and gravel concealing the underlying rocks, excepting at the immediate shores. A forest about one hundred years old, but never very large, covers the surrounding country. Spruce and tama- rack are the principal trees, with aspen, poplar, and canoe birch on the ridges. From the southern end of the lake, by a large brook entering the southeasterly bay, a route to be referred to again, leads to the Atta- wapiskat river. For the next twelve miles, between Nibinamik and Wapikopa Joke?, the river flows with a fairly stiff current, increasing to rapids at three places, and descends in all about thirty-five feet. No ledges arc- seen along the shores, the over-mantle of drift, rising in places to form ridges ninety feet in height, quite covering the underlying rocks. Wapikopa lake has a length northeasterly of thirteen miles, with a long irregular bay running to the north for fourteen miles, where it receives the waters of the river of the same name, a quiet flowing stream thirty yards wide, two to six feet deep, and with a sluggish current of about one mile an hour. Many exposures of biotite gneiss occur about the lake-shores, the foliation well marked, and dipping at angles of from forty degrees to horizontal. A coarser grey gneiss cuts these stratiform beds, and encloses in place? angular blocks of the finer black gneiss in such numbers as to constitute a breccia. REGION OF WIJSTSK AND ATTAWAPISKAT RIVERS 27 A newer reddish granite, with porphyritic crystals of red feldspar, occurs in heavy ledges near the west end. Green forest from thirty to one hundred years old clothes the shores of the lake on every side. From Wapikopa lake downwards to Weibikwei lake, a distance of thirty-eight miles, the river follows a most irregular course, and really constitutes a succession of lakes, with intervening rapids, the total descent being about eighty feet. The lake-like expansions are remarkable for the way in which the long narrow bays, running off from them, conform to the direction of glaciation. This is caused by the recurrence of parallel ridges of glacial drift, with a direction about 1ST. 30° E., the valleys between them forming the basins of the lakes. A number of small rapids occur where the river breaks through the drift ridges, and for ten miles immediately above the outflow of the channel coming in below Weibikwei lake the current is very swift, and heavy rapids occur, some of them over ledges of biotite gneiss. These rocks, the only exposures seen, are fine, banded black and grey biotite gneisses, dipping at various angles but preserving a general northeasterly trend. They are invaded by irregular masses of a coarser white gneiss, that sometimes occurs as bands conforming to their foliation, but often cuts them in the form of apophyses, and surrounds and encloses angular blocks and masses. Midway, at a point above Kanuchuan lake, where the river divides into a number of channels, a small brook flowing in from the south is the starting point for a route across to Lansdowne lake, and nine miles above Weibikwei lake a channel leads off to the north, rejoining the main river just below that lake. The southern channel of the river flows into the northwesterly bay of Weibikwei lake and discharges from its extreme northern end. Weibikwei lake has an extreme length of seventeen miles, and i3 seven miles wide. Two rivers of considerable volume flow into its southern end, the Michikenopik (stone fish-trap) — known on the old maps as the Fishbasket river — and the Wapitotem, up which the principal canoe route to the south leads. The lake, though of considerable area, nowhere shows any wide expanse of open wafer, consisting of a series of long, narrow 28 GEOLOGICAL SURVEY, CANADA channels, lying about north and south, between parallel low islands of sand, gravel and boulders, with a substratum of till reaching about the level of the top of the water. The passages are not generally more than half a mile in width, and only thirty feet in depth. The land about the lake is low, and has been almost entirely denuded of trees by recurring fires, excepting in a few localities where Banksian pine, tamarack, and spruce of fair size remain to show the character of the original forest. Sturgeon, whitefish, pike, and dore of good size are plentiful in the lake, and the Indians say that brook trout are not uncommon, but that lake trout do not occur. The only ledges about the shores are biotite gneisses that form low points near the southern end of the lake. The river discharges from the extreme northern bay of the lake by a short rapid, with a fall of three or four feet. Just below the rapid, at the head of a long bay that extends for several miles to the west, the channel which leaves the river ten miles above rejoins. This is probably really the main channel of the river. Below the junction the river flows for the first eight miles of its course over horizontally foliated ledges of banded, biotite gneiss, that cause an almost continuous succession of rapids with swift water between, down to the point of outflow of the Winiskisis, a channel that flows off to the northeast, to become reunited to the main river seventy miles below. At low water no water flows over the bar at the en- trance to this channel, though there is, at all stages of the water, a river of considerable size coming in at the junction, due, the Indians say, not to any single large stream, but to a great number of smaller tributaries draining the country between this stream and the heads of the Ekwan and Black-fence branch of the Attawapiskat rivers. Thirteen miles below the head of the island thus formed, another branch channel, called the Tabasokwia, splits off to the west and flows around an island about twenty-three miles long. For forty- five miles below the lake, or to the upper edge of the tfll-covered area, the river is an almost continuous rapid, the descent being probably as much as seven feet to the mile. At two points only do these rapids become cascades, both situated near the bottom of the very rapid section. At the Tashka rapid the vertical fall is not great, but at the Boskineig or Smoky fall there is a vertical pitch of about fifteen feet. The portage past the first of these rapids mounts over a low ridge of boulder clay, but cut banks, showing a section through the till, are first seen just above the Boskineig fall, where the river REGION OF WINISK AND ATTAWAPISKAT RIVERS 29 has cut down through twenty feet of an upper buff-eoloured clay, and six feet of an underlying, exceedingly tough blue clay holding many well striated boulders. Below the fall the cut banks of boulder clay become higher, and a few inches at the summit are seen to be stratified. Four miles below, in the thin layer of stratified beds at the top, the first fossil shells. Saxicava rugosa, were noted, proving these beds to be of post-glacial, marine origin. The height c.bove the sea is estimated to be about 350 feet. The banks, along this part of the river's course, are low. rising gradually from almost water level to heights of not more than fifty feet above it. Frequent exposures of biotite gneiss, generally nearly horizontal, but much disturbed by intrusions of a coarser white gneiss, and by veins and apophyses of pegmatite, occur all along the river. They are low, rounded, well-glaciated ledges, showing well marked stria- tion in a general direction varying from south to southwest, but showing occasional strise, that are probably later, having a direction about southeast. Down to this point, and for a few miles beyond,, the old forest has been destroyed by the same fire that swept the shores of "Weibikwei lake, and its place taken by a second growth about thirty years old. Occasional low bosses of biotite granite-gneiss are exposed along the shores for sixteen miles below Boskineig fall. A horizontal or gently undulating foliation is well developed, though the regular uniformity of their attitude is marred by frequent invading masses of coarse white gneiss and pegmatite. These exposures are the last that outcrop along the river until the outer rim of the limestones of the Hudson Bay basin is reached, 140 miles below. Though the bottom of the trough gradually becomes lower in reference to the surface of the till as the river is descended, at no place in this distance has degradation been carried far enough to expose the underlying rocks, the great number of boulders derived from the wearing away of the till probably becoming an increasingly impor- tant factor in retarding the wearing action of the current. Below the last exposure of gneiss the old forest still clothes the banks, the brule above referred to extending only to that distance. The banks of the river preserve, all along the part of its course lying within the till-covered area, a very uniform character. The shores between low and high water mark gradually slope \ip from the 30 GEOLOGICAL SURVEY, CANADA water's edge, and are often paved with boulders, and marked at the upper edge by a belt of low bushes and grasses. From high water mark tbe bank of boulder clay rises in an almost sheer wall, bare and raw looking, like the side of a recent railway cutting or canal; the lower till often rough with the great number of projecting boulders, but the upper smooth faced like a pure clay. Capping the upper clay is a very unequally distributed layer of marine clay, in places reach- ing a thickness of ten feet, but over long distances entirely wanting. The impervious character of the till, together with its nearly flat or gently undulating surface, gives to the country a muskeg-like character, even though it lies eighty feet or more above the bed of the river. Along the immediate banks, and for perhaps a chain or two back, there is a narrow belt of trees of fair size, and back of that stretches away a great level, plateau-like country, practically without drainage, and consequently moss-covered to a great depth, support- ing a stunted and deformed growth of black spruce and tamarack. There is no river valley, the trench cut in the boulder clay being but little wider than the actual bed of the stream. The comparatively stable character of the till walls is indicated by this belt of larger growth, as, were the disintegration proceeding at all rapidly, the ordinary condition of tree growth would prevail quite to the edge of the trough. At sixty-eight and seventy-seven miles, respectively, below Weibikwei lake, the Tabasokwia and Winiskisis channels rejoin the parent stream, the latter now of considerable volume. At a lake-like expansion studded with islands, situated seven miles below the inflow of the Little Winisk, the first tributaries of importance join the river, the Asheweig flowing from the southwest, and the Atikameg from the southeast. The former of these, which is slightly the larger, is the West Winisk of the old maps, and the stream referred to on a former page as flowing out from the main river at Misamik- wash lake 224 miles above. At its outlet it is a quiet flowing stream, with a good current, a chain or more in width, and having an average depth of about four feet. A short distance below this point white birches and balsam spruces are seen for the last time on the banks, and thence to the sea the forest growth, quite to the edge of the river trough, is composed entirely of black spruce and tamarack. The islands, and here and there a projecting point, however, continue to show groves of white spruce, balsam poplar, and aspen. REGION OF WINISK AND ATTAWAPISKAT RIVERS 31 After a course almost directly north, with slight curves to the east and west, for 126 miles, the river by a sharp turn suddenly changes its direction to a little south of east, and keeps that trend for seventy miles. Looking down the valley from a point a few miles above the elbow, the land to the north, beyond the turn, is seen to be elevated a little above the general level, the line of higher ground probably representing the northern edge of the Silurian basin. The abrupt turn made by the river, and its long detour to the east before resum- ing its normal northerly direction, may probably also be attributable to the presence of the barrier offered by the rim of the limestone area. Two tributaries, the Banipatau and the Pikwakwud, join the main river near the elbow. Both head near the Fawn branch of the Severn river, and by the last named there is a canoe route to the Severn. The Winino brook comes in from the north about half-way down the easterly stretch, and nine miles farther on an island six miles in length, known to the Indians as Atikminis, or Caribou island, divides the river into two channels of nearly equal volume. The almost sheer walls of boulde: clay, with their intermittent and irregular capping of marine clay, continue to rise in reference to the river bed, until at a point fifty miles above the mouth they attain a height of eighty-five feet above the water level, with a bed of but slightly beached and not at all decayed sphagnum moss on top. The marine clays with their contained fossils, a list of which is published elsewhere in this report, immediately underlie the moss. The lime- stones and dolomites of the Hudson Bay basin first outcrop at a distance of forty-two miles from the bay, measuring along the river. They are flat-lying, slightly magnesian, flaggy limestones, forming the bed of the river, but not appearing above the water. Within a very few miles, however, the slope of the river carries it below the surface of the limestones so that they form low walls, gradually in- creasing in height in reference to the surface of the water until, four miles below, the river flows through a gorge cut to a depth of thirty feet in the limestones and dolomites. This is probably a part of an old pre-glacial channel, as from here on down towards the sea the limestone walls, capped by boulder clay, alternate with banks that show till only down to high water mark. The surface of the country, extending back from the sides of the river-trough, has the 32 GEOLOGICAL SURVEY, CANADA same plateau-like character stretching away as far as the eye can sec. as in almost level, moss-covered plain, with only a sparse growth of stunted trees. The limestones show gentle undulations, but are, broadly speak- ing, nearly fiat, with a slope northerly corresponding closely with the descent of the river. A small collection of fossils, determined by Dr. Whiteaves, is referred to more at length on another page. They serve to satisfactorily fix the position of these beds as Silurian, and of about the age of the Niagara. At a projecting point on the southeast bank, twenty-six miles from the mouth, an entirely different set of rocks is brought to the surface, in the form of a double anticlinal fold, whose axis strikes south 70° east. These consist of banded green and black slates and calcareous quartzites, the whole very hard and baked looking. No actual contact with the overlying dolomites or limestones is seen, so that it is not possible to say with certainty whether or not the two sets are conformable. It seems very probable, however, that the upper beds, that gradually merge upwards from a calcareous quart- zite into a highly siliceous limestone, underlie conformably the lowest stratum of limestone. No fossils were found in these beds. The ribboned character of the slates, their bright coloration, and the occurrence in them of streams of more highly calcareous pebble- like pieces that are very suggestive of broken limestone bands, give to them a most striking appearance, and would make their recogni- tion, if exposed at any other place on the river, almost a certainty. It was considered at the time that these might represent a part of the Nastapoka series noted by Mr. Dowling about thirty miles to the east, on Sutton Mill lake. There does not seem, however, to be a sufficient similarity between these beds and those described by Mr. Dowling to warrant this correlation. Below this point, and down nearly to the mouth of the river, the limestones and dolomites, for the most part a repetition of the same beds lying in low undulations, are almost continuously exposed, forming low cliffs, overlain by a thick mantle of boulder clay. The river, along this part of its course, is about thirty chains wide, with many expansions three-qnartrrs of a mile or more in width, and dotted with islands. The Mattawa, a river of considerable volume, by which there is an Indian canoe route to the Ekwan river, comes in from the east twenty-four miles from the mouth, and ten miles farther down the Mishamattawa, or Big Mattawa, flows in from the west. This stream REGION OF WI.MSK AND ATTA WAPISKAT EIVESS 33 is used by the Indians as an inland canoe route to the mouth of the Severn, which is reached by ascending the stream almost to its head and crossing thence to the Shagamu, which flows into the west shore of Hudson bay about a day and a half's journey below the Severn. For the last twenty-five miles of its course before reaching the shores of the bay, the river has an average width of about three- quarters of a mile, but expands to over a mile at many places. An almost continuous line of islands divides it into a number of channels all along this part of its course. For the last twelve miles above the sea these islands are generally low, and clothed only with grasses and low bushes, but varied by occasional, more elevated ones that support groves of balsam poplar of good size. Above this the islands are mostly masses of till that have resisted the wear of the current; they are higher and generally well wooded with large white spruce, that attain diameters as great as two feet, and are tall and straight. The current is swift for the whole distance from Weibikwei lake to the mouth, a distance of 240 miles, though across the boulder clay area, and through the limestones, the descent is comparatively uni- form. Though there is water enough all along for tracking canoes, a channel suitable for larger boats could only be found by following a very tortuous course, and by frequently crossing from side to side, where the flat limestone ledges, approaching the surface, form almost continuous barriers across the current, with perhaps only one break where the water has any considerable depth. This even slope is characteristic of all the rivers flowing from the great central Archaean plateau downwards to the west coasts of Hudson and James bays, after they have passed the more elevated Archaean country and reached the gently sloping till-covered area. The Albany, the Attawapiskat, and the Severn rivers are other examples of this. The absence of any valley might be interpreted to mean that the river, in its present form, is very recent. It must be borne in mind, however, that evidence of a considerable age is afforded by the gorge in the limestones where the river flows in a channel cut down at least forty feet into the flat-lying strata, and all along in its passage through the sedimentary belt its pre-glacial age is indicated by the cliffs of limestone that appear alternately on the one side and on the other, with boulder clay forming the banks in the intervening spaces, constituting what is practically a broad, shallow, partly till-filled gorge all the way. 4074—3 34 GEOLOGICAL SUBVEY, CANADA It seems evident then that through the boukler clay area, until the limestones are reached, the present channel does not necessarily represent an older valley, but that below, through the limestones, the river has resumed possession of an older, pre-glacial channel. Approaching the mouth the banks become lower, and for the last few miles are not generally more than about fifteen feet high, and are composed of stratified clays and sands. Bordering each side of the river at the estuary, and extending back from the shore of the bay to form a belt from two to five miles in width, a treeless tract four or five feet above ordinary high tides extends away to the east and north, and is probably continuous, almost without interruption, up and down the west shore of the bay. It is a comparatively level plain, intersected, however, by many channels that are filled at high tide, with a gravelly and sandy surface sparsely covered by clumps of grass and brightened by many species of sub-arctic flowering plants. The river has an easterly direction just at its mouth, and the south shore consequently becomes, without change of direction, the coast of the bay; and it is only by the turning away to the north of the opposite shore that the actual mouth of the river can be fixed. At this point the estuary has a width of about three miles. It is generally shallow, large boulders showing above the surface even at high tide, while at low tide bars of sand, gravel and boulders are exposed. The ordinary rise and fall of the tide is only about six feet, but this is sufficient, so flat is the bottom of the bay in this neighbourhood, to expose at low tide wide sand flats extending far out from the actual shore i; ne and dotted with large blocks and boulders, mainly of limestone, that in places are heaped together to form points and low ridges that remain uncovered even at high tide. The shallow character of the bay was further evidenced, when the mouth was visited in August, 1903, by the barrier of pack ice that formed a continuous line across the estuary, about five miles off shore. The small sailing vessel used by the Hudson's Bay Company for the transport of supplies from the post at the mouth of the Severn river to the Winisk river is forced, by the shallow water off the mouth, to make a long circuit, following the channel of the river from far out in the bay. The length of the Winisk actually traversed, from Misamikwash lake to the mouth, is 365 miles. As it is a riv~r of considerable volume at the upper point reached, it may be confidently stated that its total length is well over 400 miles. 4074— p. 34. REGION OF WmiSK AND ATTAWAPISKAT RIVERS 35 Its volume was estimated to be about 25,000 cubic feet per second in midsummer, at a point twenty-five miles above the bay. To avoid the difficult navigation of the west coast of Hudson bay, the Indians have well-known routes both east and west from the Winisk, the western leading to the Severn river by a stream called the Mishamattawa, which enters the Winisk six miles from the mouth. From near the heal waters of this stream the Shagamu river is reached by a portage route, and that stream is descended to the coast, which is reached at a point about a day and a half's journey from the mouth of the Severn river. The eastern route leaves the Winisk eleven miles from the mouth by its tributary the Shamattawa. This stream is ascended to a large lake on its course, and one of the tributaries entering the lake is utilized to reach a stream flowing into the Ekwan river by which the western side of James bay is reached. By this route the hazardous journey for canoes along the exposed west coast and around the point of Cape Henrietta Maria is avoided. The Attawapiskat River. The Attawapiskat river was examined to the main forks twenty miles above Lansdowne lake, and its southern branch, the Kanu- chuan, for 135 miles farther, where it overlaps the foot of Lake St. Joseph at a distance of about fifteen miles to the north. A micrometer survey was made of the greater part of this dis- tance, connecting at one end with Lake St. Joseph and at the other with Tort Hope post on Eabemet lake. The Attawapiskat watershed was first reached at Wimbobika and Kapichegima lakes, lying about twelve miles to the northwest of the northeasterly end of Lake St. Joseph. The upward continuation of the river is represented by two large brooks flowing in from the west, and one, known as the Kice-stalk river, from the north. The latter affords a canoe route to Cat lake. This has been traversed by Mr. Jabez Williams, of the Hudson's Bay Company, who reports that biotite gneisses only are exposed along the route. These lakes, both long, narrow and trending about east, parallel to the prevailing strike of the gneisses in that vicinity, are separ- ated by a low ridge of chloritic, feldspathic hornblende-schists, that occur in a belt, at this point not more than three-quarters of a mile wide. The westerly extension of this belt was not traced, but it prob- 4074—3* 36 GEOLOGICAL STJBVEY, CANADA ably does not reach the shore9 of Lake St. Joseph, as it appears to be tapering in this direction. Easterly it was traced pretty con- tinuously, as the stream valley has been excavated in these rocks practically all down its course. The outlets of these two lakes unite a few miles below to form the small river known to the Indians as the Kawinogans, or No- Pikerel river. For twenty-five miles below the junction the river ha- a width of only from one to two chains, and is swift flowing and broken by numerous rapids. At frequent intervals exposures of chloritic and feldspathic schists outcrop, striking both to the north and south of east, or parallel to the general course of the river valley. Associated with the schists are more or less schistose diorites, and massive pyritous quartz diorites. At the edge of the belt is a strip of hornblende granite gneiss similar to the biotite gneiss, excepting that in it the biotite has been replaced by hornblende. The trend of the belt of basic rocks would carry it to the south of the long narrow lake called by the Indians Kagabades-dawaga. Excursions inland from the south shore of this lake revealed no out- crops, and as no further exposures of these rocks were seen on the river, the belt probably terminates in this direction not far east of the head of the lake. Along the lake shores ledges of rock were seen at only one point, where obscurely foliated biotite gneisses are cut by a later red granite of medium grain. Stratified fine white quartz sand, underlain by blue clay and over- lain by gravel, forms banks from ten to thirty feet in height all along both sides of the lake. Among the peach pebbles, which occur in great variety, are in- cluded dolomites and fossiliferous limestones, as well as many large semi-angular blocks, indicating that the underlying clay is probably a till. Where the banks are low, and fresh sections are afforded by the work of the waves, a layer of peat from two to three feet thick over- lies the clay. From the south shore of the lake a rolling, sandy- covered slope, the surface coated with white moss, and supporting an open growth of jackpine, white birch, and spruce, gradually rises to the summit of a ridge two hundred feet or more above the river. Along the side of the ridge, which is entirely of drift material, are numerous cirque-like depressions sixty to ninety feet deep, with REGIOX OF WmiSK AND ATTAWAPISKAT RIVERS 37 steeply-sloping sides, and in a few eases holding up small ponds of water. The opposite or southeast side of the ridge falls away abruptly, at as steep an angle as the sand will assume, to another rolling sandy plateau that extends for miles to the southeast. The Otosk or Elbow river, probably the longest of the various branches of the Attawapiskat, as it heads near the northeast end of Cat lake, flows into the lake from the northwest, about half-way down its northern side. Eleven miles below, after flowing in an easterly direction past a number of rapids, with occasional outcrops of biotite granite-gneiss, the river expands to form Kakawizida lake, a shallow body of water ten miles in length and a mile wide. The same rolling, sandy plain, with extensive tracts of muskeg where it approaches the south shore, surrounds the lake. Beyond the muskeg area, which extends for two miles or more back from the lake, the land gradually rises to about a hundred feet, where glacially planed surfaces of gneiss, coarse and obscurely foliated, outcrop through the drift covering. Beyond, the sandy flat gradually gains in elevation southwards for five or six miles, and then rises sharply to form a ridge of gravel and boulders 300 feet above the lake, only a few feet wide at the summit, and falling away abruptly to the south and east to a well-wooded Valley. An open forest of banksian pine covers the whole of the sand plateau. From the summit of the ridge described others are seen, appar- ently of similar character and with the same general east and west trend. Twenty-nine miles farther down the river, which still keeps an easterly direction, Ozhiski or Mud lake occupies a shallow trough, twenty-one miles long and a little over two miles wide at the broadest part. Shelving ledges of biotite granite-gneiss, lying nearly hori- zontal, or gently undulating, occur at many points along the shores. The country traversed by the river for the last fifty miles above the lake is characterized by very heavy deposits of drift, mostly stratified and often from fifty to sixty feet in thickness. Where sections are exposed along the river or lake shores, by the wear of the water, the greatest thickness is seen to be occupied by very fine, white, quartz sand and siliceous clay, underlain by a tough blue clay, in fine laminations, and overlain by irregularly distributed deposits of coarse sand and gravel. Underneath the whole, tand resting imme- diately upon the bed-rock, are deposits of till of unequal thickness, thatt at no place are exposed in section. 38 GEOLOGICAL SUBVEY, CANADA Occasional lenticular layers of indurated calcareous material, one to two inches in thickness, holding approximately 59 per cent of calcium carbonate, occur in the siliceous clays. Two specimens of the clay were examined by Dr. Hoffmann, one from the neighbour- hood of Ozhiski lake and one from higher up to Kanuchuan river. Differing only in the proportion of their lime content, they are described as slightly ferruginous, feebly plastic, readily fusible clays, holding a large quantity of siliceous grit and containing from 2*" to 30 per cent of calcium carbonate. In combination with the vegetable mould of the surface these clays should form a soil very suitable for general agriculture, though they are evidently not of value for in- dustrial use as clays. Flowing out from the north side of Ozhiski lake the river con- tinues northerly for fifteen miles, with many heavy rapids and a high average rate of flow, to an elbow, where it changes the direction of its course sharply to 'the east. Ledges of well foliated, banded, biotite granite-gneiss protrude through the drift mantle at frequent intervals along the river valley, generally lying at low angles, but in places very much contorted and crumpled. The prevailing strike is about northeast. The Pinei- muta, or north branch of the Attawapiskat, comes in from the west just at the elbow. Though somewhat smaller than the south branch, this is a river of considerable volume. For the first few miles above the forks it is broad and smooth-flowing, with banks of clay and sand, and is then broken by a high fall, above which it receives a large tributary that drains Totogan lake, lying a short distance to the north of the south branch, above Ozhiski lake. Above this the Indians say that the river takes a very long bend to the north and then southwest, and heads near the sources of the Pipestone branch of the Winisk. From the elbow the river, now nearly doubled in volume, flows easterly for twenty miles into the long southwesterly bay of Lansdowne lake. It is a succession of lake expansions, with connecting rapids, which, though they are rough, can all be run by loaded canoes. Kabania, eleven miles long and generally quite narrow, is the largest of these lakes. The land about the lake is low and drift covered, nearly horizontal, but contorted ledges of banded, biotite gneiss, with glaciated surfaces, showing at intervals. Lansdowne lake, and the lower Attawapiskat ri7er to James bay, have been described by Dr. Bell in his report published in 1887. REGIOH OF WINISK AND ATTAWAPISKAT RIVERS 39 Routes between the Attawapiskat and Winisk Rivers. The tract of country lying between the Attawapiskat and Winisk rivers was crossed by three canoe routes, two starting from Lans- downe lake and one from the Attawapiskat river, ten miles above the lake, and striking the Winisk at Weibikwei lake, between Wapikopa lake and Ivanuchuan lake, and at Nibinamik lake respectively. The first-named route leaves the extreme northeasterly bay of the lake, and reaches the height-of-land by way of a small boulder-strewn brook, flowing through low land with occasional gravel and boulder ridges of moderate height. After crossing the divide the route follows the course of the Wapitotem river, through numerous lakes down to the south bay of Weibikwei lake. For the whole distance .the country is characterized by drift ridges, rising from seventy to one hundred feet above the general level, with areas of muiskeg and low, sand- covered flats occupying the intervening valleys. For the first thirteen miles north of Lansdowne lake no exposures of rock in situ are seen, the drift cover hiding completely the underlying rock. A low ridge of slightly schistose, hard, chloritic diorite, specked with iron-pyrites and striking east and west, is the first outcrop observed. The width of the band of which it forms a part cannot be determined even approximately, as to the north the first rock outcropping through the drift occurs on Mistassin lake six miles farther on, and to the south the nearest is on Lansdowne lake nineteen miles away. These, in both cases, are biotite gneisses, the last being the first of a series of exposures that occur at intervals all the way down the stream to Weibikwei lake. The trend is in a general way about east and west, though satisfactory strikes are seldom seen owing to the contorted character of the strata, due principally to pegmatite in- vasions where the foliation is plain, or to obscure foliation. The prevailing type of rock is a hard, reddish, banded, biotite gneiss, lying nearly horizontal, stratiform in appearance, and cut by irregular masses and veins of coarse white pegmatite. The distance across by this route is sixty-five miles, and for the whole distance the country, excepting a few low, muskeg areas, has been repeatedly swept by forest fires, so that many of the ridges show surfaces of bare boulders and gravel, and others a second growth of banksian pine, white birch, aspen poplar, spruce, and tamarack. In the muskeg tracts only spruce and tamarack grow, and the trunks do not attain a size to be of industrial value. 40 GEOLOGICAL SURVEY, CANADA Low, rounded bosses of biotite gneiss, varying from very coarse to quite fine and containing a large proportion of biotite, are exposed at intervals to beyond Sagaminnis lake. The prevailing strike is a little west of south. At the northeast end of a long portage between two small lakes, lying about midway in the series, one of these low bosses is composed of interbanded fine quartzose gneiss and horn- blende schist, the fine gneiss resembling a finely micaceous, schistose quartzite, and the whole striking in conformity to the foliation of the gneisses that are exposed at no great distance on either side. The strata are much shattered and seamed with quartz veins con- taining iron sulphide. This is probably an offshoot from, or con- tinuation of the belt to be next referred to. Crossing another divide the route continues to Nibinamik lake, through numerous small lakes occurring along the course of a small tributary flowing northwesterly into the most southerly bay of the lake. The stream valley follows the trend of a belt of basic rocks from one to two miles wide, and traced in a compound curve northerly, northwesterly, and northeasterly for twelve miles. Chloritic and hornblende schists, associated with highly altered and sheared quartz diorites, are the prevailing rocks at the lower end of the belt. Farther north on the band more massive, hard diorites, and coarse diabases altered in places to obscurely schistose chloritic rocks, occur with the schists, all striking parallel to the longitudinal axis of the belt. At intervals for a distance of more than two miles massive ledges of hypersthene gabbro, similar to the Sudbury nickel- bearing irruptive, whose relations to the other rock masses were not clearly seen, but which occur at or near the western edge of the belt, are associated with a massive hard, dark-green diabase. The belt, striking northeasterly, passes just to the east of Nibina- mik lake and should cross the Winisk river a few miles below the foot of the lake. Owing to the continuous drift covering no ex- posures of rock in situ were seen along this section of the river. The most westerly route traversed ascends the Pusabiwan river, a tributary entering the Attawapiskat from the north at the foot of Kabania lake. For the first few miles to the north of the river no exposures of hard rock are seen, 'the surface consisting- of rolling hills of sand and clay. Beyond, though the country is for the most part drift covered, numerous outcrops of biotite gneiss, flat-lying or gently undulating, are seen along the river and lake shores to the REGION <>F WIXISK AXD ATTAWAPISKAT RIVERS 41 height-of-land separating these waters from those of the Alichi- kenopik brook flowing into the south end of Weibikwei lake. North- erly from here the route follows a series of small lake6 lying near the heads of streams flowing northeasterly into the Winisk, for a distance of twenty miles. Large areas of muskeg, and low sandy flats, occupy the greater part of the area traversed, diversified only by sand, gravel, and boulder ridges that nowhere rise to elevations of more than eighty or ninety feet above the general level. The second route, leaving the northwesterly bay of Lansdowne lake by a portage over a low ridge of unassorted sand, gravel and boulders, ascends a small brook through a series of lakes situated along its course, for a distance of eight miles, to a divide between the Attawapiskat and Winisk watersheds. Occasional outcrops of biotite granite-gneiss lying at low angles are seen to within about three miles of the height-of-land, beyond which, after a short interval completely drift covered, exposures of massive diorite, and hornblendic and chloritic schist are seen, for a distance of about four miles. These, without doubt, are extensions westerly of the belt of these rocks, described in connexion with the first route as crossing a short distance to the north of Attawapiskat lake. Continuing north the route follows a small stream downwards to Mameigwess lake, a body of water covering a considerable area, but of very irregular outline and broken by many islands and long; points. Biotite gneisses are the only outcrops that show through the drift deposits covering the greater part of the surface. From the foot of Mameigwess the route follows a number of small lakes to a small stream, which it descends to a southerly channel of the Winisk river fifteen miles below Wapikopa lake. Biotite gneisses only are exposed all the way through to the main river. Routes between the Winisk River and Trout Lake. Two canoe routes between the upper waters of the Winisk and Severn rivers were explored. The most westerly of these leaves the Winisk at Misamikwash lake, and the other at the first northerly expansion above Nibinainik lake. Descending a small outlet that flows through a boulder-choked channel from the northeasterly bay of the lake, the first-mentioned 42 GEOLOGICAL SIKVKV, CANADA route follows this stream — that by the addition of tributary brooks gradually becomes a river of considerable volume — northwards for fifty miles to a small lake known on the old maps as Sturgeon lake. For this distance the channel has a steep gradient, and the route is impeded by frequent rapids. Several lakes occur along its course, the largest, ten miles long and a mile and a half wide, lying not more than two miles to the north of Misamikwash lake. The country is generally low and drift covered, with only occasional exposures, all, excepting a few isolated outcrops of hornblende schist near King- fisher lake, of biotite granite-gneiss. From Sturgeon lake, a small tributary from the west, draining a chain of small lakes with connecting rapids, is ascended for thirteen miles to the divide. The rapidfs are many of them rough, and all are shallow, so that the stream is navigable with difficulty even by light canoes. The obstructions are caused by erratics that have been washed out from boulder and gravel ridges that cross the stream at frequent intervals. From the divide, Nemeigusabins lake and its outlet, a small stream with many rapids, lead to the southeast corner of Trout lake. The shores of Trout lake in the vicinity of the mouth of Nemeigusabins brook and for eight miles or more westerly are generally low and boulder strewn, the land back from the hike rising: gradually over morainic ridges of gravel and sand. Occasional out- crops of banded biotite gneiss, well foliated and lying horizontally, or gently undulating, occur here and there in low, rounded ex- posures near the lake shores. Most of the country seen near the lake has been burnt over, and 'the present forest, over all but very wet muskeg areas, is a second growth of small size. Avoiding the shallow streams between Sturgeon and Trout lakes an alternative route follows an almost direct line through nine small lakes or ponds, connected by ten portages aggregating a little over rive miles and a half in length. The section traversed is a nearly flat, sand-covered plain, with occasional low, drift ridges and extensive areas of muskeg. The second route referred to follows the west branch of the Winisk down stream from Sturgeon lake for thirty-three miles in an easterly and then southerly direction, to a small lake where the river changes its course to a northerly direction. The country traversed by the river is similar to that crossed by the main Winisk in one of its most striking features, namely, the REGION OF WINTSK AM) ATTAWAPTSKAT RIVEES 43 occurrence of parallel glacial ridges that deflect the course of the channel and of the lakes to a series of zig-zags conforming to the trend of the glaciation. The country is, however, more level and not so well drained as that bordering the main river; the proportion of swampy land is larger and the forest growth consists largely of black spruce and tamarack. Leaving the west branch a short divide is crossed, and a stream, flowing southwesterly, probably into one of the northern bays of Wunnummin lake, is ascended in a southeasterly direction through an almost continuous chain of lakes, with short rapid intervals of river joining them, for twenty-one miles, to a minor divide separat- ing the headwaters of this stream from another small river flowing southeasterly to the Winisk above Nibinamik lake, a distance of thirty-six miles. The country is of the same general character, and the lakes, and to some extent the river channel, show the same parallelism to the glaciation, due as before to the ridges of trans- ported boulders and gravel. The covering of drift material is so universal, and the relief so small, that the underlying rocks can seldom be determined. Wher- ever outcrops occur they are biotite granite gneisses, so that if the Wunnunimin Lake belt of conglomerates and schists extends to this distance easterly, as would seem probable, they are entirely concealed by surface deposits, and cross the route at one of the long intervals without exposures. Route between the Albany and Attawapiskat Eivers. The route principally used between the Albany and Attawapiskat rivers leaves the former river at Eabemet lake and reaches the latter at Lansdowne lake, traversing a distance of seventy-five miles. The first thirty miles from the Albany through Eabemet, Rib, and Keno- zhe lakes to Machawaian lake was traversed by Dr. Bell in 1886, and has been described by him in his report on '' An Exploration of Portions of the Attawapiskat and Albany Rivers,' published by the Geological Survey in 1887. The belt of diorites and felsitic, chloritic, and hornblende schists that crosses the Albany river at Petawanga lake crosses this route just north of Eabemet lake, in a band about nine miles wide, running N. 70° E. Eor the balance of the distance to Lansdowne lake, wherever outcrops are seen they are of biotite 44 GEOLOGICAL SURVEY, CANADA granite-gneiss of medium grain, striking about east and west, and banded fine black biotite gneiss cut by a coarse gneiss that encloses blocks of the finer. From the northwesterly bay of Machawaian lake the divide between the Attawapiskat and Albany waters is crossed, at a distance of two miles to the north of the lake, by a portage seventy-four chains in length, traversing a muskeg with occasional ridges of transported gravel and boulders. Manitush lake, two miles long, lying at the north end of the portage, discharges northerly by a small stream, barely navigable by canoes, into Marten Drinking river, which the route follows through Hail lake to Wintawanan lake, from which there is a route westerly through an intervening small lake, to the south branch of the Attawapiskat river at Ozhiski lake. The Marten Drinking river, rather shallow and with a number of rapids along its course, is nevertheless navigable by canoes down to its mouth at one of the southerly bays of Lansdowne lake. The country between the two rivers in the neighbourhood of the route is a high, rolling plateau, rising, midway, about a thousand feet above the sea, or a hundred feet above the Albany at the point of departure. Large areas of muskeg abound, from which rise low, rounded bosses of gneiss, and ridges of sand, gravel, and boulders. To the west of Machawaian lake the country is much more broken and rises to higher elevations. This more elevated region extends in a belt westerly past Trout and Cedar lakes, and without doubt continues still farther west, forming the height-of-land between the Albany and the south branch of the Attawapiskat. This country is referred to in the description of the route down the Kanuchuan river on a previous page, where the hills are stated to be composed of transported material to their summits. Cultivation of the Land. In the matter of the actual cultivation of these northern areas we have little to go upon. At the Hudson's Bay Company's posts at Fort Hope and Osnaburgh potatoes have been grown, and small gardens maintained from the time of the establishment of the posts. and little difficulty has been experienced in maturing the common garden vegetables of Ontario, though occasionally the frosts of late summer have cut off all but the hardier kinds. As the posts were REGION OF WINISK AND ATTA WAJ'ISKAT KIVERS 45 located with a view to their favourable situation for the purposes of the fur trade with the Indians, neither one is situated on ground well suited for cultivation, and much better results might reasonably be expected were trials made on more favourably situated tracts. An Indian cultivating a small garden plot at the head of the Pineimuta branch of the Attawapiskat river succeeds in raising good crops of potatoes and turnips. Fish. Whitefish and sturgeon are the best food fishes, and occur in most of the lakes. Both are taken in nets, and the latter also by spearing from scaffolds built out over rapids in the rivers. Dore and pike are also generally distributed over the whole area, and form an important source of food supply, though the sucker among the fishes, like the rabbit among the mammals, holds the most important place, as it can be caught everywhere, not only in the larger lakes but also in the smaller ponds and streams. Brook trout were actually caught only in the Winisk river near its mouth, and in the streams running into the Albany river, but were seen in the rapids below Weibikwei ; the Indians assert that they occur also in the lake itself. Lake trout were caught in large numbers in Trout lake at the head of the Severn river, but are not found in either the Winisk or Attawapiskat waters. Wild Animals. The moose (Alces americanns) has been found as far north as the southern shore of Weibikwei lake, in N~. lat. 52° 50', though tracks were actually seen during our exploration only as far north as the Attawapiskat river. Even here it is not nearly so plentiful as farther south in the belt of country lying ntar the Canadian Pacific railway and extending for about 150 miles north of it. Caribou (Rangifer caribou) range all over the district. No red deer are found anywhere throughout the region. The fur-bearing animals, though not so plentiful as they once were, are still fairly abundant throughout the district; the otter and the beaver from long-continued trapping are less numerous, perhaps than any other species. 46 GEOLOGICAL SURVEY, CANADA Bears (Ursus americanus) seem to be able to hold their own pretty well, and are still taken in good numbers. There is probably but one species of the common black bear, though the Indians and traders differentiate from this the brown bear, which they claim differs from the black, not only in colour and size, but also in disposi- tion and habits. Wolves (Canis lupus), though scarce, are not unknown. Foxes (Vulpes vulgaris), including the red, silver, black and cross varieties, are numerous, though they vary in numbers with the periodic increase and decrease in the numbers of the hares. Lynxes (Lynx canadensis) are fairly plentiful. Otters (Lutra canadensis), and Pine martens (Mustela ameri- cana), are taken in good numbers, and beavers (Castor fiber) occur more sparingly. Minks (Putorius vison), and muskrats (Fiber zibethicus), are plentiful. These, with skunks (Mephitis mephitica), weasels (Putorius vulgaris), and wolverines (Gulo luscus), make up the number of the merchantable furs. The rabbit (Lepus americanus) occurs abundantly all over the district, and is, perhaps, the most useful of all to the Indians, as it affords, during the winter particularly, both food and clothing. That the raccoon occasionally strays as far north as N. lat. 52° is shown by the fact of one being taken by an Indian woman on the upper Attawapiskat river in 1903. Indians. The Indians of the district, numbering about 700, are nomadic trappers, living principally upon fish, and obtaining from the Hudson's Bay Company, and to a smaller degree from other fur traders, the limited amount of necessaries that'are not supplied by the country. A few have small huts built of logs, with fireplaces and chimneys of wattles and mud, in which they live for a part of the year, but the greater number content themselves with winter teepees constructed of poles covered with sheets of birch bark, and summer tents of cotton; indeed, house building is such an arduous task for the Indian that the traders in the district have a saying to the effect that as soon as an Indian completes a house he dies, this result being due, not to the unwonted labour involved, but to the arrival of ex- treme old age before the work is finished. Plate iv Indians of the Lower Winisk river. 4074— p. 4 REGION OF WINISK AND ATTAWAPISKAT RIVERS 47 They are of the Ojibway tribe, though mixed to a certain extent with the Crees of the Hudson Bay basin, the purest Ojibway stock being found among the bands about the heads of the rivers. They seem to be men of larger frame than the Crees of the coast. A greater proportion of nominal Christians are found among these Indians of the far interior than among those nearer the front, in the hinterland of Ontario. This result is due in about equal measure to the efforts of the Koman Catholic church, which maintains a permanent mission establishment at Albany, with an educational home for children, and sends visiting missionary priests periodically among the Indians of the interior; and to the Anglican church, which maintains the missionary diocese of Afoosonee, by which resi- dent missionaries are supported at various points in the interior region. The Indians seem to accept readily the forms of Christian worship, and take great pride in their proficiency in memorizing the religious formulas presented to them. The mode of life followed by these Indians offers great obstacles to the work of the missionaries, who are able to reach them for purposes of instruction for only short periods during each year. .For the same reason, that is on account of their nomadic life, the teaching of the children can be carried on only in the same desultory way. Notwithstanding these disadvantages, practically all the Indians can read and write the syllabic characters designed and introduced by James Evans, an early Wesleyan missionary among the Crees. The introduction of this system of writing has proved a great boon to the Indians in their intercourse with one another. Written entirely phonetically it is unhampered by irregularities, and can be readily acquired by one Indian from another. So general is their knowledge of this sign language that every Indian camping-place, and every point where canoe routes diverge, become local post offices, where letters written on birch bark, often, of course, containing only nn account of trivial occurrences, but giving the opportunity to convey news of importance, are left for the information of follo\%ing parties. It is very doubtful whether the Indian has advanced much in general prosperity from the days when he lived in primitive savagery. His teepee was the same then as now; his weapons are now mere 4S GEOLOGICAL SURVEY, CANADA effective, but game is less plentiful; he wears better clothes, or clothes that one associates with civilization, but not probably so well adapted to his needs and way of living as his old raiment of skins. Even now he has to fall back upon rabbit skins, the only furs that he can afford to sacrifice to his personal use, for protection in winter. The skins are cut into strips, each skin, by being cut spirally, pro- ducing; a continuous strip. These strips are sewn together at the ends, and twisted into ropes which are woven loosely into blankets and rough coats that very effectually keep out the most extreme cold. Fish are taken with net and spear, and in trap-weirs. These arc constructed of spruce poles driven in a line into the bottom of streams, and interwoven with twigs so as to fence off the greater part of the water, and force it to run in volume only through a gate arranged so that the water flowing through the opening quickly drops away through the interstices of a platform of poles, leaving stranded all fish coming down with the current. One or two families will often camp by the side of one of these ' mechiken ' for weeks at a time, supplying their wants entirely from the stranded fish, and smoke-drying any surplus collected. This is accomplished by simply stringing the split fish on poles and hanging them in the smoke-laden atmosphere of the teepee. The fat dropping from the fish in drying is carefully collected and preserved for future use in bags made of the skins of embryo rabbits, the bladders of pike, or in similar receptacles ingeniously improvised from the materials at hand Wild rice, a staple among the Indians farther south, is too rarely met with throughout these northern regions to form any part of the Indians food supply, and to supplement his diet of fish and flesh he has only the various berries in their seasons and the small amount of flour that he is able to buy from the trader in exchange for his surplus furs. For tea, when the imported article is not available, the small twigs of the trailing red cedar are used. Taken as a whole, they appear to be a fairly healthy lot, though many suffer from diseases of the skin brought on probably by a too constant diet of fish. The greatest mortality is caused by pulmonary diseases, to which they are very prone, and to the occasional outbreak of epidemics of measles, etc., that sometimes prove widely fatal. They are far from cleanly in their personal habits, a few weeks' residence at a place in the summer time generally rendering it no longer habitable by reason of the accumulated filth. REGION OF WINISK AND ATTAWAPISKAT RIVERS 49 With the exception of occasional small log huts, the Indians of the region dwell in teepees covered with birch bark, though the cotton tent, made from materials bought from the traders, is now widely used during the summer months. Near the mouth of the Winisk river, many miles north of the ranges of white birches, a winter teepee, made after the plan generally used for birch bark wigwams, was covered with blocks of moss cut from the muskeg. Archaeology. Chipped flints were found in numbers scattered along the beach of an island in Attawapiskat lake. Two fairly perfect arrowheads were found at the same place, one chipped from white quartz and the other from flint, derived apparently directly from the drift, where it occurs as small boulders which have been carried primarily from the nodular beds in the limestones of the Hudson Bay basin. At camping-places of the Indians broken specimens of Pecten islandicus were noticed among the debris of the camps. These shells occur in a very perfect state of preservation in the marine clay, and are still used by the Indians along the river as very convenient substitutes for spoons. Forests. The average size of the trees growing within the country ex- plored is not great. On exceptionally favourable tracts the spruces attain sizes quite large enough for commercial use as sawn lumber, and large areas would afford good pulpwood. Evidences of the constant recurrence of forest fires over the area are everywhere plainly seen. The brule areas, varying from quite small patches to large tracts, are of every age; some are so old the forest has attained the full height of the old growth and the newer age of the trees can only be ascertained by a reference to their rings of growth, and others so recent that no vegetation covers the blackened surface. These fires are generally the result of the carelessness of Indian travellers, but may sometimes be traced to the igniting of a dry, standing tree-trunk by lightning. The oldest trees found in the whole area were growing on a till-covered island, about fifty miles from the mouth of the Winisk river. The complete isolation from the mainland by broad channels ensured its protection from fires having their origin outside its own borders. The spruces growing 4074—4 50 GEOLOGICAL SURVEY, CANADA here were found by their rings of growth to be between 270 and 280 years old. The diameters and ages of trees, growing in a number of different localities throughout the region, were noted, and are given in the list below : — Tamarack, Winisk Black spruce ■ Tamarack Black spruce Aspen poplar river. 32 miles from mouth 32 32 32 50 05 «5 Diameter j in Age, inches three by rings, of feet from I growth. ground. near bank . . fi5 G5 below Wapikopa lake. Wapikopa lake 10 chains back Nibinamik lake above Nibinamik lake. 9 12 12 8 10 8 6 3 3 10 9 6 9 5 15 15 100 125 153 75 275 130 115 105 80 130 145 135 75 75 130 130 The rings show that the growth is generally rapid for the period between five and thirty years, and afterwards exceedingly slow. The northern limit of a number of the common trees of northern Canada falls within the district, and of one species both the northern and southern limits. There is a black birch that the Indians call the squirrel-bark birch. Specimens of the wood and foliage of this tree were submitted to Professor John Macoun, by whom they were forwarded to Dr. Sargent, of the Arnold Arboretum, for determination. Dr. Sargent has named this birch Betula fontinalis. It was not seen growing in abundance anywhere in the district, though occasional trees were noted at various points between the Albany and Winisk rivers, tbe most southerly occurrence being in N. lat. 51° 28' on Dog-hole brook flowing into Lake St. Joseph, and the most northerly in N. lat. 52° 40' on the Wapitotem river flowing into Weibikwei lake on the Winisk river. The largest tree noticed had a diameter of six inches at three feet from the ground, and a height of about thirty feet. Where seen it was growing near the banks of rivers or lakes, in moist 4074— p. 50. REGION OF WINISK AND ATTAWAPISKAT RIVERS 51 localities. A table is subjoined of the observed northern limits of a number of species. Northern Limits of Trees. White elm, Ulnus americana, Albany river N. lat. 51 J 30' Black ash, Fraxinus sambucifolia, Eabemet lake... " 51° 50' Mountain maple, Acer spicatum, between Attawa- piskat and Winisk rivers " 52° 25' Mountain ash, Pyrus americana, between Attawa- piskat and Winisk rivers " 52° 38' Banksian pine, Pinus hanksiana, Weibikwei lake. . " 53° White cedar, Thuya occidentalis, Weibikwei lake. . " 53° 05' Balsam spruce, Abies balsamea, Winisk river " 54° 15' Canoe birch, Betula papyracea, Winisk river " 54° 25' Aspen poplar, Populus tremuloides, Winisk river. . " 54° 45' The northern limits of balsam poplar, tamarack, and black and white spruce lie beyond the mouth of the Winisk river, the most northerly point examined. Climate. The climate, as would be expected in these latitudes, and in a wilderness country approximately a thousand feet above sea-level, is somewhat severe. The summer temperature, though on occasional days rising as high as 85° Fahr., averages very much lower, and the nights are, practically, always cool. Frosty nights often continue into the early summer, and recur again in the autumn before most grain-crops would be ready for harvesting. Temperatures were taken with the thermometer during two seasons, and these, averaged, gave the following results for the months of July and August on the lower Winisk river, and for July, August and part of September on the upper Winisk and upper Attawapiskat rivers : — 6 a.m. noon. 6 p.m. Lower Winisk river 57° 69° 57° Upper Winisk and Attawapiskat rivers.. 47° -5 61° -6 58° The only points in the region where any attempts at cultivation of the land are made are the two Hudson's Bay Company's posts at Osnaburgh. near the foot of Lake St. Joseph, and at Fort Hope, on Eabemet lake. 1074-^1 52 GEOLOGICAL SURVEY, CANADA At these posts small kitchen gardens and potato-fields are main- tained with some success, though neither place is favourably situated for the purpose, the soil in both cases consisting of an almost pure sand. Timothy and clover grow luxuriantly, and all the common garden vegetables thrive at both places. Indian corn, however, is not sufficiently filled out for table use when caught by the frost. Barley has been successfully grown at Osnaburgh, and the potato crop, wherever a suitable tract of land has been utilized, has been generally fairly good at both places. The first killing frost in 1903 occurred on the night of September 3, and in 1904 on the night of August 30. The temperature of the water in a number of the larger lakes and rivers was taken by thermometer at six inches below the surface, and is given in the following, very uniform list : — Water Temperature. Lake St. Joseph, Albany river, June 28 59i° Annimwash lake, Albany river, July 5 58° Kagabades-dawaga lake, Attawapiskat river, July 16 62° Attawapiskat river, August 8 . 60° Weibikwei lake, Winisk river, August 9 62° Nibinamik lake, Winisk river, August 23 58° Winisk river, August 24 57° Land Shells. A small collection of land shells, made during the summer of 1904, has been examined by Dr. Whiteaves, who enumerated the following species. It was noted that in actual number of individuals there was a decided and progressive decrease as the latitude in- creased : — • Vertigo ovata, Say. Conulus fulvus (Miiller). Zonitoides arhoreus (.Say). Vitrea hammonisl (Strom). Pyramidula striatella (Anthony). Succinea vermeta, Say. Succinea retusa%, Lea. Succinea ovalis, Gould, non Say. REGION OF WINISK AND ATTAWAPISKAT RIVERS 53 Freshwater Shells. Collections of the freshwater shells of the region were made each year, and submitted to Dr. Whiteaves for determination, who has furnished the subjoined list, which for convenience has been tabu- lated according to watershed areas : — List of Freshwater Shells collected by Mr. W. Mclnnes in 1903- 4-5 on the Winisk, Attawapiskat, and Albany Rivers, on the Root and English Rivers, near Lac Seul, and on the Severn River at Trout Lake. BY J. F. Whiteaves. Lampsilis luteola, (Lamarck) Anodonta marginata ? Say Anodonta fragilis, Lamarck Anodonta Kennicottii ? Lea Sphcerium simile, Say Sphcerium Walkeri, Sterki Sphcerium cr.iarginatum, Prime Sphcerium stamineum, Conrad Sphcerium (Musculium) secure. Prime. Sphcerium (Musculium) partumcium, Sphcerium jlavum, (Prime) Sphcerium rhomboideuin, (Say) Sphcerium striatinum, Lamarck Sphcerium — Pisidium cimipressum, Prime Pisidium altile, Anthony Pisidium fallax, var. errans, Sterki Pisidium variabile, Prime Pisidium affine, Sterki Pisidium Sargenti, Sterki Pisidium Mainense, Sterki Pisidium abditum, Haldeman Pisidium Roperi, Sterki Pisidium politum, Sterki Pisidium rotundatum, Prime Pisidium pauperculum var. crystallense. Sterki Pisidium vesiculare, Sterki Pisidium splendidulum, Sterki, var Pisidium scutellatum, Sterki Pisidium. medianum, Sterki Pisidium milium, Held, Small var Pisidium sp. nov ? Pisidium — ? (near P. abditum) Winisk river. Atta- wapiskat river. Albany river. English and Root Trout lake, Severn 54 GEOLOGICAL SURVEY, CANADA Valvata tricarinata, Say Valvata sincera, Say Amnicola liniosa, Say Limncea mcgasoma, Say Limncea stagnalis, L Limncea stagnalis, appressa Limncea ])alustris, Muller Limncea citascopium, Say Limncea galbana (Haldeman) Dall. Planorbis trivolvis, Say Planorbis corpulcntus, Say. Planorbis bicarinatus, Say Planorbis companulatus, Say Planorbis albus, Mtiller Planorbis hirsutus, Gould Segmentina armigera, Say Physa hcterostropha , Say Ancylus parallelus, Haldeman Wini.sk river. Atta wapiskat river, Albany river. English and Root Trout lake, Severn river. INDEX. A Page. Agricultural land 10, 38 Albany river 23, 33 Animals, wild, of the district 45 Archaean area 10, 20 Archaeology 49 Arrowsmith map, details for, supplied by G. Taylor 8 Asheweig, west branch of Winisk 24,30 Atikameg river 30 Atikminis island 31 Attawapiskat river 35 " character of 33 " descended by Dr. R. Bell 8 " micrometer survey of 35 " probable connexion with Albany river 23 ' surveys of, in 1903-4-5 9 B Banipatau river 31 Bell, Dr. Robert, descended Attawapiskat river 8 east coast Hudson bay 20 report on Lansdowne lake and Attawapiskat river. 38 route between Albany and Attawapiskat river, tra- versed by 43 Big Mattawa river 32 Boskineig rapid 28 Eoulder clay area 10, 12 Winisk river 19, 21, 28, 29, 30, 31, 32 Bcyd, W. H., Ekwan and Trout rivers explored by 8 Camsell, Chas., drift masses noted by 25 Caribou island (See Atikminis) 31 Cat lake 35,37 Clay, boulder (See Boulder clay) " marine 21,30,31 " specimens examined by Dr. Hoffmann 38 Climate, character of 51 Cowling, D. B., Ekwan and Trout rivers explored by 8 rocks at Sutton Mill lake noted by 16, 32 56 GEOLOGICAL SURVEY OF CANADA E Face. Eabemet lake 11 Ekwan river 8 Elbow river 37 E^ans, James, syllabic characters designed by 47 Explorations, early, in the district 7 F Fawn branch of Severn river 8 F]sh in rivers of district 45 Wubikwii lake 28 Fishbasket river (See Michikenopik) 27 Fishing:, method of 48 Food of the Indians 48 Forests (See also timber) 49 Fort Hope, headquarters of FT. B. trade 9 Fossils 12,18,21,22,29,31,32,36 Foxe, Capt. Luke 7 G Geological divisions of district 13 Geology along Attawapiskat river 36, 38 ' between Albany and Attawapiskat rivers 43 " Attawapiskat and Winisk rivers 40, 41 Winisk river and Trout lake 43 Glaciation 20,21,27,37,43 H Hoffmann, Dr., examination of clays 38 Hudson, Henry 7 Hudson's bays 7 Huronian, Lower 13, 25 Irdians 43 J James, Capt. Thomas 7 K Kabania lake 3S Kagabades-dawaga lake 36 Kakawizida lake 37 Kanuchuan branch of Attawapiskat river 35 Kepichegima lake 35 Kawinogans river 36 Keewatin 13, 14, 25 REGION OF WINISK AND ATTAWAPISKAT RIVERS 57 L Page. Lake St. Joseph 35 Lakes of district, character of 11 Lansdowne lake 11, 38 " named by Dr. R. Bell 8 Laurentian 13 Low, A. P., character cf country on Fawn branch of Severn 18 Fawn branch of Severn river explored by 8 M Mattawa river 32 Mica 14 Michikenis river 26 Michikenopik river 27 Misamikwash lake 12,24 Mishamattawa river 32, 35 Mud lake (See Ozhiski lake) 37 Muskeg, extent of 11 N Nastapoka series 32 Nibinamik lake 11, 12, 26 Nickel-bearing intrusives 14 No-Pickerel river (See Kawinogans) 36 Ivorthwest passage 7, 8 O Otcsk river 37 Ofchiski lake 11, 37 P Peat 36 Pikwakwud river 31 Pineimuta river 38 Pleistocene 13, 19 Post-pleistocene 13, 21 Potatoes grown at H. B. posts 44, 45, 52 R Rapids on Attawapiskat river 37, 38 canoe route YVinisk river to Trout lake 42, 43 Kawinogans river 36 Marten Drinking river 44 Winisk river 25, 27, 28 Rice, wild, rarely met with 48 Pice-stalk river 35 Routes between Albany and Attawapiskat rivers 43 " " Attawapiskat and Winisk rivers 39 AVinisk river and Trout lake 41 " into the region 9 s Page. Sargent, Dr., birch named by 50 Skagamu river 33. 35 Shamattawa river ■ 35 Shells, freshwater 53 land 52 Silurian 12, 13, 15, 32 Smoky fall 28 Tsbasokwia branch of Winisk river 23, 28, 30 Tashka rapid 28 Taylor, G., visited Winisk river 8 Temperature of water 52 Timber 49 " between Attawapiskat and Winisk rivers 39 " Winisk river and Trout lake 42, 43 " on Attawapiskat river 36, 37 " Nibinimak lake.. 26 " Wapikopa lake 27 " Weibikwei lake 2S " Winisk river 29,30,33 Totogan lake 38 Trees, northern limit of 51 Trcut river S Vegetables, garden, possibility of raising 44, 51 Vegetation along Winisk river 23 of area 13 W Wapikopa lake 11, 12, 26 river 26 Wapitotem river 27 Water-powers 10 Weibikwei lake 11, 27 V'hiteaves, Dr., fossils identified 18, 22, 32 shells determined by 52, 53 Williams, Jabez, H. B. Co 35 Vvimbobika lake 11, 35 YViiiino brook 31 Winisk river, character of 22 estuary of 34 length and volume 34 section along 17 surveys of in 1903-4-5 S visited by G. Taylor „_ 8 missionaries. Winiskisis branch of Winisk jiverr. .... • - -rivY \ 23, 30 Wimnumminlaie-T-r-Tr . . . . . . ~ . p£ttl) Wvrh \ 11, 12. 25 The jfcM* B le West 91° ii'oui Greeirwdeli 90*30 r- MAP 9A on parts of the m\ H(v'MR^ ( m rr u ALBANY, SBVEB NORTHERN <>.\TA1 MAP 9A Kx|lluiN'll lioiltoN on parts oi' tkp B amjWIMSK RIVERS, ETC. UO and NORTH WEST TERRITORIES LITHOMOUNT PAMPHLET BINDER Manufactured 6y SAYLORD BROS. Uc. , Syrecm*, N. V. Stockton, Calif. UCLA-GeCogy/Geophysics Library F H06M18re L 006 568 709 7 ■■ D Q01068 427 2