CHAPTER VIII.

GEOLOGY.

The following summary of the geology of the northeastern coast of America and of the Arctic islands is based, so far as the southern and eastern portions of that region are concerned, upon the observations made on the cruise of the _Neptune_, and is supplemented by the reports of previous explorers in the areas beyond the limit of the voyage of that vessel. The geological work of the Arctic explorers until recent years was necessarily poor and disconnected owing to the absence of trained men, and to such work being of secondary importance among the objects of the expeditions.

The observations of these earlier explorers were carefully gathered from the different narratives, and ably summarized by Dr. G. M. Dawson in a report on the Geology of the Northern Portion of the Dominion of Canada, published in 1886. This work has been largely followed in the present report, but such corrections have been made as are justified by the knowledge gained in explorations since the date of its publication.

The notes on the geology of the southern part of Baffin island are from the observations of Dr. R. Bell, while those on Ellesmere island and the Sverdrup islands are based on the work of P. Schei, the geologist of the _Fram_.

The rocks of the Arctic islands and of the northern coasts of the continent present an almost continuous ascending series from the Archæan to the Tertiary, while the upper loose material represents various phases of the Glacial age and of the subsequent Post-Glacial deposits.

Only isolated attempts have been made to subdivide the Archæan complex into Laurentian, Huronian and other members of the system. This lack of subdivision is due to want of detailed knowledge; not to the absence of the various members of the complex in these northern regions, where the greater number of the members are known to occur.

The Palæozoic rocks are well represented on the islands by thick deposits extending upwards in a continuous series from the Cambro-Silurian to the upper beds of the Carboniferous. Rocks older than the Galena-Trenton are only found in the northern part of Ellesmere island, where a series of beds appears to connect the Upper Huronian formations with the lower members of the Cambro-Silurian.

Mesozoic rocks are found on the northern Parry islands, on the Sverdrup group and on the western and northern sides of Ellesmere island.

Tertiary formations occur on the northwestern islands, on the northern part of Ellesmere, as well as on the northern and eastern parts of Baffin island.

The former presence of a continental ice-cap is attested along the northwestern shores of Hudson bay and in the southern part of Baffin island, by the rounded and polished rock surfaces, which are everywhere well marked by the ice striae, often in several sets showing changes in the direction of the ice movement. On the east side of Baffin the rock surfaces show signs of rounding and smoothing by ice, but the striae are not well marked, and the glaciation does not appear to have been nearly so intense as to the south and westward. Passing northward up the western side of Davis strait and Baffin bay the evidence of intense glaciation becomes less and less, that on Ellesmere the present condition of the local ice-covering would appear to represent nearly as great an amount of glaciation as ever occurred there.

The sequence of earth movements and physical conditions, read from the geological formations of the northeast, are as follows: An ancient floor of crystalline rocks, largely of igneous origin, represents the most ancient crust of the earth. These, associated with ancient bedded deposits and cut by dark basic intrusions of trap and allied rocks, were at a very early period so crushed and foliated that it is now impossible to separate them. Upon this ancient complex was laid down a series of bedded deposits, chiefly sandstones and dolomites, associated with contemporaneous traps, as may be seen along both the shores of Smith sound. Following this came a great outburst of granite and other acidic igneous rocks which, over large areas, inclosed, penetrated, compressed and otherwise altered the sedimentary deposits to such an extent that it is now impossible to separate them from the older complex upon which they were originally deposited. Only in a few comparatively small areas were the conditions of the granite intrusion such as to allow the sedimentary deposits to preserve their original unaltered conditions. All the above rocks are grouped in the Archæan, and further and closer examination will probably show that it contains all the members of the Laurentian and Huronian found in the more southern Archæan regions of Canada.

Except in the northern part of Ellesmere, there is a considerable time-break in the geological sequence in the northeast between the Archæan rocks and the Cambro-Silurian strata which rest unconformably upon them. Schei found at Bache peninsula, on the eastern side of Ellesmere, a series of stratified sedimentary rocks resting upon the northern flank of the Archæan and containing fossils of Cambrian age. These deposits have a thickness of nearly 1,500 feet, and are overlaid by limestones containing Cambro-Silurian fossils.

The Archæan rocks at the time of deposition of the lower beds of the Cambro-Silurian limestones appear to have extended southward from the vicinity of Bache peninsula in a gradually widening ridge along the western side of Baffin bay and Davis strait. In this manner they attained a width of seventy miles on the southern side of North Devon, and occupied the entire southern shore of Baffin island, being separated from the great area of Labrador by the depression of Hudson strait, which probably existed at that early period. Islands of Archæan rocks may also have risen above the surface of the Cambro-Silurian sea in the present island of North Somerset and on Melville and Boothia peninsulas, as well as on other portions of the northern coasts of the mainland, to the west of Hudson bay.

The western Cambro-Silurian sea filled the present depression of Hudson bay, and extended far to the south and westward of its present limits, outliers of limestone containing fossils of this age, and very similar in mineral character, being found in the valleys of the great lakes of Manitoba. From Manitoba these rocks have been traced southward into the United States, so that at the time of their deposition the Cambro-Silurian sea occupied a great basin open to the Arctic ocean and extending southward into the middle of the continent.

This was the time of maximum encroachment of the northern ocean, after which the land gradually rose, and the sea slowly receded. Owing to the great lapse of time and the eroding of the thick ice-cap in the more southern regions, it is exceedingly difficult to now trace the boundaries of the narrowing sea during Silurian and Devonian times. Cambro-Silurian limestones containing fossils which refer them to the Galena-Trenton, are widespread over the northern islands and in a wide margin along the western and southern shores of Hudson bay. Outliers of these rocks occur at the head of Frobisher bay in the southwest part of Baffin island and on Akpatok island in Ungava bay. As before stated, similar limestones are found in the lake valleys of Manitoba, and it is quite possible that these limestones were once continuous with those of Hudson bay, the present break having been caused by the erosion of the glacier.

The upward continuation of these limestones containing Silurian fossils occupies a corresponding but slightly circumscribed area. These Silurian limestones form the characteristic abrupt cliffs of the islands on both sides of Lancaster sound, and continuing southward occupy the larger parts of Southampton, Coats and Mansfield islands in Hudson bay, as well as the low lands of the western part of Baffin island. They are not well marked, and are probably considerably thinner on the western side of Hudson bay, but are found in the Winnipeg basin.

The Devonian gradually emerges from the Silurian in the cliffs of the islands to the north of Lancaster sound, and forms the lower parts of the cliffs of the southern side of Ellesmere. Devonian fossils are not found in the limestones of the islands of Hudson bay, and only occur in a narrow belt on the low lands to the west and southwest of James bay. Similar rocks form the upper beds of the Winnipeg basin.

There is no break in the passage from Devonian to Carboniferous in the rocks forming the Parry islands and the southern part of Ellesmere, where Carboniferous rocks occupy wide areas on these northern islands, but are not found to the southward of Lancaster sound, showing that the Palæozoic sea had retreated that far north before the close of the Devonian.

The land rose above the ocean at the close of the Carboniferous, and with the exception of the northern parts of the Parry islands, the Sverdrup group and the western part of Ellesmere has not been deeply submerged since. Rocks of Mesozoic age, belonging to the Alpine Triassic, have been found in the last-named places, but in no other localities to the southward within the limits of this report.

Considerable earth movements occurred at the close of the Mesozoic period, causing those and older rocks to be highly tilted and folded.

Another slight submergence took place in the Miocene Tertiary, when shallow water deposits of sand, gravel and clay, associated with beds of lignite, were laid down in the wide valleys along the margins of several of the Arctic islands. Such deposits are known to exist in Banks island, on the western side of Ellesmere and along the northern and eastern sides of Baffin island. There is little doubt that other deposits of this age will be discovered when more systematic search has been made for them in these northern regions. From the character of the fossil plants found in these deposits there can be little doubt that during the Miocene the climate of these northern islands was much warmer than at present, and approached a tropical condition.

The conditions of the land and water surfaces during the Glacial period differed little from those at present, except that there has been a considerable uplift of the land, as proved by the marine terraces found along the coasts. The maximum uplift probably amounted to 700 feet along the eastern side of Baffin island, and was perhaps slightly less on the islands farther north, where Schei reports beaches 600 feet and upwards above the present sea level. This being the case, a new explanation must be found for the depression and subsequent uplift of the land covered by ice, if the uplift be practically the same in northern Ellesmere, where the accumulation of ice is nearly as great to-day as at any previous time, while in the southern part of Baffin island a great thickness of ice was present during the Glacial period and has now completely disappeared. Perhaps we have been taking cause for effect, and the uplift due to some unknown cause may have been the cause of a lessening of the ice; certainly the almost equal rise of the land throughout the Arctic islands is an argument against the subsidence of the northern lands being due to the burden of the ice-cap, and the subsequent uplift due to the disappearance of that burden.

DETAILS OF THE GEOLOGY BY ISLAND GROUPS.

It is exceedingly difficult to write a readable, concise and comprehensive account of the geology of the territory included in this report without subdividing it in some manner. This has been attempted by considering the different formations under their separate headings, and dividing the territory, as has been done in the geographical description, into groups of islands, and considering each of the great geological divisions separately.

ARCHÆAN.

_Hudson Bay and Hudson Strait._

The territory comprised in this group includes the islands of Southampton, Coats, Nottingham, Salisbury, Charles and Resolution, along with the shores of the northwest part of Hudson bay, and the south shore of Hudson strait.

The geological information concerning this group obtained prior to the present voyage is contained in the reports of Dr. Bell, Tyrrell and the writer.

Crystalline gneisses, schists and granites occupy the eastern and northern parts of Southampton, extending northward from Seahorse point to Frozen strait at the northern end of the island. The rocks near the junction of the Archæan with the Silurian at Seahorse point are largely a very quartzose, light-gray mica-gneiss, associated with bands of rusty-weathering, fine-grained mica-gneiss holding graphite in small flakes, the rusty colour being due to the decomposition of small grains of pyrite disseminated through the rock. This rusty gneiss closely resembles that found in the vicinity of Cape Wolstenholme at the entrance to Hudson strait, and both appear to be similar to the sillimanite gneiss of the Grenville series of southern Canada. Both of the above rocks are cut and twisted by masses of a coarser granite-gneiss pink to red in colour, with pearly feldspar and smoky quartz. All are cut by dikes of feldspathic pegmatite containing much of the pearly feldspar.

The only notes relating to the Archæan area to the north of Seahorse point are those of Parry and Back, both of whom mention the occurrence of granites and crystalline rocks in several places to the northward.

The band of Archæan rocks which crosses the eastern part of Coats island has never been examined closely, and our knowledge of it is confined to observations made from the ship in passing.

The high rocky shores of the eastern and northern sides of Salisbury island were closely followed by the _Neptune_, so that the red, crystalline rocks forming its cliffs could easily be seen. The prevailing rock was red, or pink, and only occasionally were darker masses seen.

Nottingham, which lies south of Salisbury in the western part of Hudson strait, was visited by Dr. Bell in 1884 and 1885, when he examined the rocks in the neighbourhood of Port De Boucherville, in its southeast part, and he there found along with the common varieties of gneiss a number of patches of fine-grained red syenite.

Charles island is wholly formed of Archæan gneisses. The prevailing rock in the western part is a fine-grained light-gray, or pink mica-gneiss, associated with medium to coarse-grained mica-hornblende granite-gneiss; the latter cutting and altering the light-coloured gneisses.

Resolution island has never been visited by a geologist, and consequently its rocks can only be described from observations made while passing it in the ships. The rocks everywhere appear to be crystalline Archæan, a red variety predominating.

The southern shores of Hudson strait from Douglas harbour to the mouth of George river in the southeast part of Ungava bay were examined by the writer in 1897, and a detailed statement concerning them is given in the report of that year. The remaining portions of this side of Hudson strait were examined on the voyage of the _Neptune_, the part westward from Douglas harbour to Cape Wolstenholme while sketching the coast-line from the ship, and the greater part of that between Cape Chidley and the mouth of George river by Mr. Caldwell in a boat during the absence of the _Neptune_ to the north in 1904. These examinations connect with the work of the writer beyond Cape Wolstenholme, and thus practically finish the examination of the north and west sides of the Labrador peninsula.

The rocks of the south coast of the strait westward from Cape Chidley, as examined by Mr. Caldwell, show that large masses of red and pink mica and mica-hornblende-granite, in a more or less foliated condition, occupy the greater part of the coast area. These granites are newer than the other crystalline rocks associated with them, which they have inclosed and altered. The older rocks are largely of basic igneous origin, and vary in composition from anorthosites almost free from bisilicates, to hornblendic and chloritic schists containing very little feldspar. A series of light-coloured, high quartzose gneisses is also found, and probably represents altered bedded rocks associated with the basic igneous ones.

The basic rocks and the light-coloured gneisses are penetrated by many large dikes of pegmatite from the granites. In many places these dikes are very quartzose, and where they cut the basic masses often contain pyrite and give indications of other minerals. Associated with the lighter gneisses are large long masses of rock, which carry in places considerable quantities of graphite, in others an impure iron ore. The study of these bands has not been sufficient to pronounce as to whether they are veins or beds.

Ancient crystalline rocks occupy the entire coast-line from the mouth of George river at Ungava bay to Cape Wolstenholme at the eastern entrance to the strait. Mica-granite and, in less quantities, mica-hornblende granite, both more or less foliated, occur along the greater length of this coast. In many places these granites are associated with other gneisses, which are usually lighter in colour, finer in texture, and contain more quartz than the eruptive granite. These lighter gneisses usually are garnet-bearing, the crystals of that mineral being often of large size. In a greater number of places the gneisses of these two series are so closely mingled as to render a separation impossible; but there are localities, notably along the west side of Ungava bay, where such a separation can be made, and the lighter coloured gneisses are seen to be cut and foliated by the intrusion of the coarser and garnet-free granite-gneisses. At the mouth of Payne river and about the mouth of Hopes Advance bay the change from unaltered clastic rocks to the light-coloured gneisses is plainly seen in a number of places. The unaltered series consists of impure dolomites, sandstones, cherts and bedded iron ores similar to the series of rocks found in the central parts of Labrador and along the east coast of Hudson bay. This series bears a close resemblance to the iron-bearing rocks of Lake Superior, and there is little doubt that they are of the same age. In former reports they have been termed so-called Cambrian, but by the new classification they represent one or more members of the Huronian. These rocks are associated, as elsewhere, with great outbursts of basic igneous matter in the form of sills, dikes and irregular masses. Where the newer granites have cut and inclosed masses of this series, the different rocks forming it are seen to have undergone considerable alteration. The bedding has been disturbed, so that the strata lie at angles approaching the vertical, and have been broken, and minutely penetrated by quartzose injections, both along and across the bedding planes. Foliation and schistosity have been induced, and the arrangement of the chemical constituents has been altered so that new minerals are formed. The impure cherty limestone is changed to hornblendic schists, the impure sandstone and quartzite to garnet-bearing quartzose gneisses, and the cherty iron ores to a gneissic rock consisting of layers of quartz and specular iron.

An examination of a number of the contacts between the granites and the Huronian rocks shows an alteration, from a slight crumpling and baking to highly tilted and contorted crystalline schist and gneiss. The accompanying basic igneous rocks, originally fine-grained traps or diabase, are changed in like manner to strongly foliated hornblendic and chloritic schists, usually freely penetrated by quartz veins holding quantities of pyrite and other minerals.

In passing westward from Ungava bay the granite predominates along the coast, and the areas of the altered bedded rocks and their accompanying traps are much smaller, and are so intimately intruded by the granite that it is very difficult to separate them. The gneisses of the altered Huronian rocks can only be guessed at, by their texture, light colour and the presence of garnet in them. The ancient traps and greenstone masses are more easily separated from the complex, but they are so penetrated by the granites that it is impossible to trace them except on a large scale detailed map, which would require many years’ work to complete.

Large masses of these basic rocks occur along the coast in several places between Cape Hopes Advance and Douglas harbour, most noticeably about Wakeham and Fisher bays, where quartz veins are numerous, and carry considerable quantities of sulphides.

The examination of the coast to the westward of Douglas harbour was only such as could be made from the ship, and lacks all the detail of the eastern portion. Red granite gneisses appear to occupy the greater portion of the coast, with areas of dark basic rocks at intervals. At Sugluk bay, where a closer examination was made, a medium to coarse-grained pink to red mica and mica-hornblende granite-gneiss was most abundant. This granite inclosed bands of a lighter coloured quartzose-gneiss, and also intruded large masses of dark-green altered diabase.

The granite-gneisses occupy the coast to within a few miles of Cape Wolstenholme, when the prevailing rock is a rusty fine-grained sillimanite-gneiss containing scales of graphite and considerable pyrite in small grains disseminated through the rock. These dark gneisses are inclosed and penetrated by the granite-gneisses, and probably represent portions of the ancient bedded series.

Beyond Cape Wolstenholme, gneisses occupy the eastern coast of Hudson bay to within a short distance of Cape Smith, where a high ridge of trap runs inland in a northeast direction and has a width of several miles. Further south the gneisses again predominate along the coast to the Portland promontory in latitude 58° N.

GEOLOGY OF THE NORTHWEST SHORES OF HUDSON BAY.

The following account of the geology of the northwest shores of Hudson bay has been compiled from observations made during the trip of the launch from Winchester inlet to Chesterfield inlet in September, 1903. These are supplemented by the notes made in May, 1904, while making a track survey from Cape Fullerton to the entrance of Chesterfield inlet. The observations to the north of Fullerton were made by Mr. Caldwell, in April and May, 1904, while on his surveying trip to the head of Wager inlet; to these are added observations by the writer made on a boat trip along the coast later in that spring, on the way to and from Southampton island, when the rocks of the mainland were examined as far north as Yellow bluff.

The rocks seen along the shore between Chesterfield and Winchester inlets are largely a flesh-red to pink mica-hornblende granite-gneiss, often only slightly foliated, and varying in texture from medium to coarse-grained. These are associated with broken bands of dark-gray or red gneisses, usually very quartzose, and containing a considerable quantity of mica and hornblende, the latter often partly decomposed to chlorite. These gneisses have evidently been cut and broken by the intrusion of the granite-gneisses. Many veins of pegmatite cut all these rocks; it consists chiefly of red or violet feldspar with much quartz, and in some places large crystals of black hornblende.

The granite-gneisses are also most abundant inland, as was seen along the lower part of Chesterfield inlet and in the country forty miles inland from Winchester inlet.

The granites occupy the shores and islands between Winchester inlet and the west side of Island bay, about half way to Fullerton, when they give place abruptly to a series of dark schists. These schists are largely micaceous, but there are also frequent bands containing considerable quantities of hornblende, and these are more basic than the more common mica-schists, which always contain quartz in varying amounts. These mica-schists appear to have been clastic rocks associated with bands of trap, all having been altered and foliated by the granite intrusion. The schists are very regular, and have a constant strike of N. 10° W. Many of the basic bands contain varying quantities of pyrite, but it was never seen in sufficient abundance to constitute a mine. All the schists carry dark-red garnets, some of which are regular in their crystallization and of good size.

The islands about Fullerton harbour are formed partly of granite and partly of these dark schists. On this eastern contact of the granite with the area of schistose rocks, the latter have been greatly disturbed by the intrusion, being squeezed, contorted and broken by granite masses, as may be seen from the illustration. Areas of coarser basic rocks now occur in the rocks along the coast, usually in the form of coarse gabbro, but often in a more altered condition as coarse hornblende-gneisses. The granites are the prevailing rocks along the coast as far north as the mouth of Wager inlet, and have generally a red or pink colour. Associated with them and evidently altered by their intrusion are patches of gray quartzose gneiss, and less frequently areas of old basic intrusive rocks.

At Whale point, where the rocks were closely examined, the oldest rock was represented by a medium to fine-grained, gray and pink, very quartzose gneiss. This had been cut by a coarse diabase, and both had been foliated and broken by the intrusion of the newer granite. Dikes of a newer diabase cut all the other rocks.

The granites prevail about Wager inlet, but there are more and larger areas of the dark basic rocks about that great bay than to the southward, making it a more promising field for economic minerals, especially as these basic rocks generally carry sulphides, and Rae reported free gold to have been found about the head of the inlet.

All the geological information concerning the western coast of Hudson bay to the northward of Wager inlet is contained in the narratives of the voyages of Parry and Dr. Rae. The explorations of Rae ended at Repulse bay. He reported only Laurentian rocks along the coast, with granite-gneisses predominating, these being accompanied by considerable areas of greenstones, showing the rocks in the northern part to be very like those along the southern shores of Roes Welcome.

Parry explored the west side of Fox channel from Frozen strait to Fury and Hecla strait. The result of his observations has been summarized by Dr. Dawson as follows:

‘The geological specimens brought back were examined by Prof. Jameson, and the detailed maps of the expedition include indications of the character of the rock at so many places, as to afford the means of tracing the geological outlines with very considerable accuracy. Granitic and gneissic rocks occupy the whole of the southern part of the east shore of Melville peninsula, and are continued northward behind a low track of limestone country, forming a range of mountains in the centre of the peninsula to Hecla and Fury strait. They also form the south shore of this strait, and most of the islands in it, and apparently the whole eastern shore of the adjacent south part of Cockburn “island.”

‘The rocks referred above, in a general way, to the Archæan, probably include areas of Huronian. Jameson mentions as among the prominent varieties of rocks derived from this region, “Granite, gneiss, mica-slate, clay-slate, chlorite-slate, primitive-trap, serpentine, limestone and porphyry.” In association with these the following minerals occur: “Zircon and beryl, also precious garnet, actinolite, tremolite, diallage, coccolite, rock crystal, calc-spar, rhomb-spar, asbestos, graphite or black lead, specular iron ore, magnetic iron ore, chromic ore or chromate of iron, titanic iron, common and magnetic iron pyrites.” Some of the “transition rocks” noticed by Jameson should probably also be classed with the Archæan, and in addition to several of the minerals above mentioned, in these were found tourmaline (schorl) and molybdenite.’

The coast between the mouth of Chesterfield inlet and Churchill, was examined by Tyrrell, and the following summary of the geology is taken from his report:—

‘On the low flat shores of Hudson bay between Seal river and Cape Esquimaux few rock exposures occur, but those seen consisted of granites and gneisses of typical Laurentian aspect. For forty miles north of Cape Esquimaux no rock in place was seen, and thence northward to Baird bay some of the points consisted of granite and gneiss, though the shore generally consisted of Huronian rocks.’

‘The largest area of Huronian rocks found in this district extends more or less continuously for 120 miles along the west coast of Hudson bay, from near Baker’s foreland to a point 45 miles north of Cape Esquimaux.’

‘The rocks constituting this system may be divided into three more or less distinct groups, viz.: The Marble island quartzites, the greenish quartzites and graywackes, and the more or less highly altered, and often schistose diabases and gabbros.’

‘The Marble island quartzites are composed of hard white quartzite, consisting of more or less rounded grains of quartz, of moderately regular size, cemented together by interstitial silica. They are very distinctly bedded in thick and thin beds, and the surfaces of the beds are often covered with beautiful ripple-markings. The heavier beds also often show distinct false bedding. They are usually in a more or less inclined attitude, but they were nowhere seen to be very much crumpled or squeezed into minute folds. Their total thickness was not determined.’

‘These quartzites were first noted by Dr. Bell from Marble island, and although this island was not examined by the writer, rocks of undoubtedly similar character to those described by Dr. Bell, were seen in many places along the shore, and consequently the name is here retained.’

‘In one place near the _cache_ on the west side of Hudson bay, a thickness of sixty feet of this quartzite, in a nearly vertical attitude, was seen almost in contact with the Laurentian gneiss, there being but a narrow, drift-filled gap between the two. This would indicate either the existence of a fault, or that here the quartzites are the base of the Huronian, or that the gneiss represents an eruptive rock which has risen through or into the Huronian subsequent to the deposition of the quartzite.’

‘Dark-green eruptive rocks, chiefly diabases, often very much squeezed and altered, are largely developed in the Huronian, composing a considerable proportion of the rocks of this system. On the west coast of Hudson bay these rocks are cut by many veins of white quartz, highly charged with iron and copper-pyrite.’

‘Associated with the massive diabases, and often indistinguishable from them except on close examination, are many beds of fine-grained, often schistose graywacke, or greenish quartzite, which appear to have been caught up in, or surrounded by the eruptive rocks.’

GEOLOGY OF ISLAND GROUP II.

This group is comprised of the great island of Baffin, with Bylot island lying off its northeast corner, and the many smaller islands which lie as a fringe around both.

Geological specimens from the east side of Baffin were collected by the expedition under Ross and Parry, and were described by Dr. McCulloch. They consisted of loose specimens collected in two localities, and give little information. Specimens collected by Parry on the same coast were described by Koning as gneiss and micaceous quartz rock, also some ambiguous granitic compound in which hornblende seems to enter as a subordinate ingredient.

Dr. P. C. Sutherland, in 1853, describes the east coast of Baffin island between Lancaster sound and Cumberland gulf as follows:

‘On the opposite shore (south) of Lancaster sound, at Cape Walter Bathurst, the crystalline rocks are again recognized, and from this point they occupy the whole coast south to Cumberland strait and probably considerably beyond it. To this, however, I believe there is one exception, at Cape Durban, on the 67th parallel, where coal has been found by whalers; and also at Kingaite, two degrees to the southwest of Durban, where from the appearance of the land as viewed from a distance, trap may be said to occur on both sides of the inlet. Graphite is found abundant and pure in several islands situated on the 65th parallel of latitude, in Cumberland strait, and on the west side of Davis strait.’

C. F. Hall brought home a considerable collection of rocks and minerals picked up during his explorations about Frobisher bay and the southeast coast of Baffin island. These were named by Prof. B. K. Emerson, and consist of ordinary Laurentian rocks, including granite, gneiss and schists. The minerals were magnetite, apatite, bornite and pyrite from Frobisher and Cyrus Field bays. Lower Silurian limestones were found in a small outlier at Silliman’s Fossil Mount near the head of Frobisher bay. This locality was visited in 1897 by a party from the Peary Arctic expedition of that year. In the course of a few hours they obtained fifty-four species of fossils from this locality, which were later named by C. Schuchert.

Dr. Franz Boas describes the nucleus of the mountain masses of Baffin island to be everywhere gneiss and granite, with Silurian limestones about the region of the large lakes of the interior and along the low lands of the west coast.

Dr. R. Bell visited the north shores of Hudson strait in 1884 and 1885, and again in 1897, when he made a close examination of the coast from the neighbourhood of Big island to Chorkbak inlet near Gordon bay. Dr. Bell describes the prevailing rocks of the southern shore of Baffin island as consisting of well stratified hornblende and mica-gneiss, mostly gray in colour, but sometimes reddish, interstratified with great bands of crystalline limestones, parallel to one another and conformable to the strike of the gneiss, which in a general way may be said to be parallel to the coast in the above distance. The direction, however, varies somewhat in different sections of the coast.

‘The distinguishing feature in the geology of the southern part of Baffin land is the great abundance, thickness and regularity of the limestones associated with the gneisses. At least ten immense beds, as shown on the accompanying map, were recognized, and it is probable that the two others, discovered in North bay, are distinct from any of these. There would, therefore, appear to be twelve principal bands as far as known, to say nothing of numerous minor ones, between Icy cape and Chorkbak inlet. The limestones are for the most part nearly white, coarsely crystalline, and mixed with whitish feldspar.————The limestones usually contain scattered grains of graphite, and among the other minerals which commonly occur in the various bands are mica, garnet, magnetite, pyrite and hornblende.’

‘Although white is the prevailing colour of these limestones, this, in some localities, is replaced by light-gray and occasionally by mottled varieties.’

‘The limestone bands have not suffered greater denudation than the gneisses, and they form hill and dale alternately with the latter.————Owing to the scantiness of vegetation in Baffin land, the white colour of the limestones on the sides and tops of the hills and ridges renders them very conspicuous in the landscape. Seen from a hill-top at a distance of fifteen or twenty miles, they might be taken for glaciers.’

‘As to the total thickness of the twelve bands of crystalline limestone which have been mentioned as occurring in this part of Baffin land, the available data on the subject are not sufficient to form a correct estimate, but on adding together their probable approximate widths it seems to be no exaggeration to place their possible total volume, great as it may appear, at about 30,000 feet, or on an average of 2,500 feet for each of the principal bands, taking no account at all of the smaller ones.’

From his observations made along the coast to the eastward of Big island in 1885, and from the finding of crystalline limestone fragments by Hall in Frobisher bay, Dr. Bell concludes that the crystalline limestones extend eastward to Resolution island, giving a very extensive development of the Grenville series of the Laurentian in the southern part of Baffin island.

At present we know that the limestones of the typical Grenville series are only the highly crystalline equivalents of some of the Huronian limestones. This probably is the case in Baffin island, where these rocks with some of the accompanying gneisses represent a highly metamorphic phase of portions of the Huronian, while other of the gneisses are the foliated state of the granite masses which caused the alteration of the limestones. This would correlate the rocks on the north side of Hudson strait with the altered Huronian rocks of northern Labrador, where in places similar crystalline limestones occur.

The Huronian rocks of Labrador are marked by the number of repetitions of the strata caused by thrust faults in all the areas examined, and this repetition of measures by similar faults may account for the number of bands of limestone found in the southern part of Baffin island.

The crystalline rocks appear to form the southwest coast of Baffin island for some distance beyond King cape on the east side of Fox channel, when they give place to a wide area of low lands extending nearly to the head of Fox channel, where the crystalline rocks again form the higher lands to the north and east of Fury and Hecla strait.

On the late voyage of the _Neptune_ the rocks of the east side of Baffin island were examined at Ponds inlet, on the islands on both sides of Cumberland gulf, and at Cape Haven and Frenchmans cove on Cyrus Field bay. In other places the ship passed sufficiently near the shores to allow of a good idea being formed of the rocks by the aid of powerful glasses.

Examinations of the rocks were made at Button point, the southeast part of Bylot island, on the north side of the entrance to Ponds inlet; also in the vicinity of Salmon river some thirty miles up the inlet and on its south side, and at Erik harbour on the same side near the mouth of the inlet. At all these places typical Laurentian gneisses and schists were obtained. Among the specimens brought home from these localities is a light-coloured coarse-grained augen-gneiss consisting largely of white and pink feldspar, with thin bands of biotite and little quartz. Another seeming variety of this rock is a well-banded fine-grained mica-gneiss composed of pink and white bands of feldspar separated by thin bands of mica. Associated with these are bands of very quartzose gneiss varying in colour from light to dark from the varying proportions of mica present. These gneisses are usually found containing a considerable number of dark-red garnets; and they probably represent a metamorphic series. A fine to medium-grained rock, usually somewhat foliated, and composed largely of dark-red feldspar with much mica, little hornblende and quartz, cuts the foregoing gneisses, and probably was the granite which altered them by intrusion to their present state. The basic intrusive rocks are represented by dark-green diabase, or its alteration products, dark hornblendic and chloritic schists and gneisses. Taken as a whole, this series of specimens would answer for any of the typical Laurentian regions of northern Canada.

At Cumberland gulf the rocks were examined at Kaxodluin on the south shore, some twenty miles from Blacklead station; also at Blacklead and at Kekerten islands. At Kaxodliun light and dark-coloured mica schists and gneisses were found, cut by a light-pink mica-granite-gneiss. The dark schistose rocks were decomposed near the surface, and contained a considerable amount of disseminated pyrite. Between this place and Blacklead the ship followed the shore-line closely, so that the prevailing dark, rusty gneisses were distinctly seen.

The most abundant rock on Blacklead island is a coarse-grained, pink mica-granite-gneiss, containing large feldspar crystals. This cuts, and is foliated with, coarse, dark mica-schists, and finer-grained lighter-coloured quartzose gneisses. Some of the dark schists contain flakes of graphite, and this mineral is said to be abundant in places on the islands and shores of the gulf farther to the westward, where attempts have been made to work some of the mica and graphite deposits, without much success.

At Kekerten similar gneisses are found, along with large masses of diabase and greenstone, somewhat decomposed near the surface, where it weathers reddish.

At Frenchman cove at the head of Cyrus Field bay, the prevailing rock is a coarse-grained, red mica-granite-gneiss, associated with bands of coarse mica-schist.

At Cape Haven station near the northern entrance to the bay, pink and gray mica-gneiss prevails, and is cut by many large dikes of red pegmatite composed largely of perthite, with some quartz and mica. Schists forming one of the islands of the harbour contain many well-developed crystals of pyrite, up to an inch cube.

The northern and eastern sides of Bylot island appear to be wholly formed of crystalline rocks, without any of the capping limestones found upon the other islands of Lancaster sound.

ISLANDS OF GROUP III.

This group contains the large islands of Bank, Victoria, Prince of Wales, North Somerset and King William, all situated south of Lancaster sound and west of Prince Regent inlet. North Somerset alone was visited by the _Neptune_; all geological information concerning the others being from the observations made by the several parties engaged in the search for the Franklin expedition.

Dr. G. M. Dawson collected this information from the narratives of these search expeditions, and printed a concise summary of it in his report on the northern portions of the Dominion, from which the following notes have been taken:—

‘Archæan rocks are found only on Prince of Wales and North Somerset islands, where a spur from the great mass of crystalline rocks forming the northeastern mainland extends northward through the peninsula of Boothia and forms the land on both sides of Peel sound.

‘The granitoid rocks are again found on the west side of North Somerset, where they form the eastern boundary of Peel sound. Boulders of the granite are found at a considerable distance (100 miles) to the northeastward of the rock _in situ_, as at Port Leopold, Cape Rennell, &c. The general characters of the granitic rocks in the north and west of North Somerset are thus described by Capt. M’Clintock: “Near Cape Rennell we passed a very remarkable rounded boulder of gneiss or granite; it was six yards in circumference, and stood near the beach, and some fifteen or twenty yards above it; one or two masses of rounded gneiss, although very much smaller, had arrested our attention at Port Leopold, as then we knew of no such formation nearer than Cape Warrender, 130 miles to the northeast; subsequently we found it to commence _in situ_ at Cape Granite, nearly 100 miles to the southwest of Port Leopold. The granite of Cape Warrender differs considerably from that of North Somerset, the former being a graphic granite, composed of gray quartz and white feldspar, the quartz predominating, while the latter, a North Somerset granite, is composed of gray quartz, red feldspar and green chloritic mica, the latter in large flakes. Both the granite and gneiss of North Somerset are remarkable for their soapy feel.’

‘To the east of Cape Bunny, where the Silurian limestone ceases, and south of which the granite commences, is a remarkable valley called Transition valley, from the junction of sandstone and limestone that takes place there. The sandstone is red, and of the same general character as that which rests upon the granitoid rocks of Cape Warrender and at Wolstenholm sound. Owing to the mode of travelling, by sledge on the ice, round the coast, no information was obtained of the geology of the interior of the country, but it appears highly probable that the granite of North Somerset, as well as that of the other localities mentioned, is overlaid by a group of sandstones and conglomerates, on which the Upper Silurian limestones repose directly. A low sandy beach marks the termination of the valley to the northward, and on this beach were found numerous pebbles, washed from the hills of the interior, composed of quartzose sandstone, carnelian and Silurian limestone.’

‘Cape Granite is the northern boundary of the granite, which retains the same character as far as Howe harbour. It is composed of quartz, red feldspar and dark-green chlorite, and is accompanied with gneiss of the same composition.’

‘The granitoid rocks extend across Peel sound into Prince of Wales island in the form of a dark syenite, composed of quartz, greenish-white feldspar passing into yellow, and hornblende.’

ISLANDS OF GROUP IV.

Archæan rocks are found only in the eastern part of this group, on the large islands of Ellesmere and North Devon. They rise from beneath the newer rocks on the south side of Hayes sound a few miles north of Cape Sabine, and then occupy the remainder of the eastern coast of Ellesmere and that of North Devon. This area appears to form a wedge-shaped mass expanding southward, so that on Jones and Lancaster sounds they extend a considerable distance to the westward, until they become capped by limestone, and then gradually sink below the level of the sea.

Both the Laurentian and Huronian divisions of the Archæan are represented in the area. A series of bedded rocks consisting of several thousand feet of sandstones, limestones and other sediments occupies the coast and islands of the east side of Smith sound, from Cape Atholl northward to Foulke fiord. On the west side the northern limit of these rocks is Cape Isabella, from which they occupy the shore of Ellesmere for upwards of twenty miles to the south, the southern limit not having yet been determined on that side.

These bedded rocks are associated with dark coloured traps and diabase, which are present in the form of sills between the bedding; as dikes cutting the bedded rocks and as large intrusive masses. Dr. Sutherland classified these rocks as the equivalents of the Tertiary sandstones of Disko on account of their lithological resemblance and from the occurrence of traps with both. The southern junction of these bedded rocks with the granites and gneisses forming the Greenland coast to the southward was not seen, but at Foulke fiord and at Cape Isabella the northern contact is quite plain. In both places the bedded series, for some considerable distance from the contact, has been tilted and fractured, while near the contact the sandstones and limestones appear to have been changed into quartzite and crystalline limestones by the injection of great masses of granite. This granite seen at Cape Sabine and Cape Herschel is an ordinary Laurentian granite, and in no way resembles the acidic rocks of Tertiary or Post-Tertiary age, which they should do if the bedded series were of the age assigned to them by Dr. Sutherland. The sandstones, limestones and their associated traps bear a close resemblance to portions of the Huronian series found on Hudson bay and in the interior of Labrador. There is also a similarity between their contacts with the Laurentian granite and some of the contacts found in those more southern localities. No fossils have as yet been found in these rocks, and until such are found it is thought best to remove this series from the Tertiary and place it in the Huronian.

On the past voyage the coast of Ellesmere island was lost sight of about twenty miles south of Cape Isabella, and no land was again seen on the west side of Baffin bay until Philpots island, lying off the east end of North Devon, was reached, where the ship passed sufficiently close to the small outlying islands to show that they were composed of Laurentian gneisses and granites. From thence similar rocks were seen forming the southern shores of North Devon as far as the west side of Croker bay, where they begin to sink slowly to the westward, and are capped by a considerable thickness of flat-bedded limestone, which rests unconformably upon the rounded surface of the older rocks. The Laurentian rocks finally dip below the sea a few miles to the westward of Cuming creek.

The specimens from the Laurentian area, which extends southward from Hayes sound to Cape Isabella, were collected at Capes Sabine and Herschel. The specimens from both localities are very similar, the prevailing rock being a moderately coarse-grained granite, of a dark-red colour, composed largely of red feldspar and bluish quartz, with a small quantity of biotite in diminutive scales. These rocks are only slightly foliated in a few places.

The specimens from the Laurentian measures beneath the Silurian limestones at Cuming creek show a greater variety. A red gneiss, varying in texture from fine to coarse, predominates. It is composed largely of feldspar, with quartz and considerable biotite. It cuts a lighter-coloured, more quartzose gneiss, and also bands of dark mica-diorite-gneiss.