having become an estuary and an arm of the sea. As Loch Melville extends to the depth of almost 1,000 ft. below sea-level, and the Report by Mr. H. W. Jones (Jt. App. pp. 2334, 23512) shows that the bulk of the water in it is seawater, I regard Lake Melville * as an arm of the sea and not a freshwater lake or a river.
It may be pointed out that the range of tide in Lake Melville is small, the amplitude (i.e. half the height between high and low tide levels) being about ·83 of a foot (W. J. Stewart, Jt. App., p. 2319). The tide at the inner end of Lake Melville in Goose Bay, has a rise and fall of 2·6 ft. for springs and 1·34 ft. for neaps (H. W. Jones, amplitude of 1·3 ft., ·67 ft., Jt. App., p. 2358). This tidal range is greater than that in the Baltic, where on the Swedish coast near Stockholm and on the opposite part of the Finland coast at Helsingfors it is less than an inch (Baltic Pilot, Pt. II. 1913, p. 21), or than in the Mediterranean, as Naples is said to have no regular tide (Mediterranean Pilot, Pt. II., 1905, p. 260), and that on the south-east coast of Sicily is one foot (ibid., Pt. I., 1913, p. 519).
The salinity of Lake Melville is much higher than that of parts of the Baltic. The lowest salinity recorded in H. W. Jones' table, Jt. App., p. 2361, for Lake Melville is 10·57,† whereas in the northern part of the Baltic it is less than three (i.e. three parts to the 1,000) and as far south as the latitude of Stockholm and around the Aland Islands, the salinity is only from 5 to 6 per 1,000 (Petterson, Scott. Geogr. Magaz., Vol. X., pl. X., 1894).
On consideration of all the evidence available to me I regard Lake Melville Basin, the Narrows, and the entrance to them as part of one arm of the sea.
IV. THE NATURE OF THE LABRADOR INLETS.
The nature of the Labrador inlets has an important bearing on the width of the coastal region. Two suggestions have been put forward as to their origin. The older view is that they were ancient valleys excavated by rivers and enlarged and deepened by ice during the glacial period. That view seems to me quite inconsistent with their shapes and distribution. The two characteristic geographical features of these valleys are their youth, as shown by their steep, flat, wall-like sides and their arrangement in rectilinear series, the members of which often cross at regular angles like a network. Their youth is especially shown in the northern series, where they occur as deep, narrow valleys with precipitous parallel sides.
A. THE FIORD VALLEYS NOT ICE-CUT.
That these valleys were not carved out by ice follows from the fact, which is I think admitted by all the authorities, that they were earlier than
*A. S. Packard (1891, map opp. p. 90), marks it as Melville Bay.
† He records surface fresh water in the upper part of “ lake Melville ” (Jt. App. p. 2352) & in Carter Basin (Ibid., p. 2361) : but under the brackish water of Mulligan's Bay lay 88% sea water (Ibid. p. 2352)
the Ice Age, and that those in the northern area were never even filled with ice. Robert Bell (Scottish Geogr. Mag. XI. 1895, p. 340) remarked that “ the mountains forming the northern part of the chain have not been glaciated like the rest of Labrador, and they present steep sides with jagged crests rising out of the cold sea.” Prof. Daly confirms the view that the northern valleys were never more than partially occupied by ice ; according to him the ice only reached the height of 2,000 ft. above sea level in Nachvak fiord, where the walls rise 3,400 ft. high. If, therefore, the upper part of each of these valleys was never occupied by ice, they cannot have been made by ice. The utmost that can be claimed for ice action in respect to them is that it deepened pre-existing valleys.
B. NOR OLD RIVER VALLEYS.
The essential characteristics of these valleys are equally inconsistent with their origin as cut by rivers which slowly excavated them by wear throughout geological time. Rivers may excavate deep canyons, but such canyons are short-lived. Frost, wind, rain and streams, and the chemical decay of rocks when exposed to the atmosphere together attack and wear back the upper walls of a canyon, which is thus widened into a broad rounded valley. If these fiord valleys were due to the action of rivers which, as Low suggests, had begun their excavation in very ancient times (Low, 1895, pp. 26-7, Jt. App., p. 2600, which he dates even as early as pre-Cambrian, Jt. App., p. 2601, quoted from his 1895 memoir), then it is impossible that the valleys should still be canyon-like valleys. Moreover, if these valleys were of great geological antiquity some of the hollows should include deposits laid down in the long period since their excavation. Most of such deposits might have been swept away by ice, but some would probably have been left, like the “ Cambrian ?” sandstones beside Lake Melville and Double Mer.
Practically all those who have described the coastal valleys of Labrador represent them, or at least many of them, as fiords. Fiord valleys are always young, for when rivers occupy them their walls are soon worn into the irregular slopes of ordinary land valleys, and the fiord character is destroyed.
Rivers and ice may have modified and deepened the fiord valleys of East Labrador, but they did not initiate them, and some other explanation of their formation is needed.
C. THE ARRANGEMENT OF THE FIORD VALLEYS.
The fact that the fiord valleys are independent of river action is indicated by their plan and distribution. The fiords at the northern end of the series near Cape Chidley run east and west. They are continuous through the land as tickles or straits, which separate the northern blocks as islands. These “ tickles ” have the characteristics of young valleys. It is most improbable that they were cut by recent rivers, which would not have been likely to saw their way through this massive mountain range, while Hudson Strait was open to them. It may be suggested that they were cut before the
formation of Hudson Strait, but the youthful character of the tickles shows that they are younger and not older than Hudson Strait.
South of this east and west series are a number which run inland from north-west or north-east ; two of them so nearly meet that the northern end of Labrador is a peninsula joined to the mainland by a narrow isthmus. The east to west direction of the fiords is predominant from Cape Chidley to Cape Mugford. From Cape Mugford southward to Nain is an intermediate area, in which east and west valleys are combined with others trending from about W.N.W. to E.S.E., and others from north-east to south-west.
From Nain to the Isle of Ponds (lat. 53½°) the shore trends W.N.W. to E.S.E., and the fiord valleys run in three directions. One series is parallel to the coast-line ; a second series cuts inland at an angle of about 60° and this series includes Loch Melville and the front of the Mealy Mountains ; the third series has a trend of east and west.
From the Isle of Ponds southward to Cape Charles the coastline resumes its N.N.W. to S.S.E. direction and the predominant inlets have the same trend from Nain to the Isle of Ponds, or they are combined with others trending approximately east and west.
The direction of these fiord lines must be due to some process which has cracked the coastal region of Eastern Labrador along three chief directions. The east and west valleys are parallel to Hudson Strait ; this direction is dominant in the northern end of Labrador, but is less marked toward the south. The second important series, trending from E.N.E. to W.S.W. includes Hamilton Inlet, which is in line with the long straight course of the St. Lawrence past Rimouski to Quebec, and is also parallel, amongst other important geographical features in this region, to Nova Scotia, and to the peninsulas and arms of the sea in south-eastern Newfoundland. The members of the third series are either parallel to the coast line from Nain to the Isle of Ponds, or are at about 60° to it. This regular distribution of the inlets is quite independent of the fundamental structure of the country. The inlets cut across the rocks regardless of their strike. That the fiord valleys cut across “ the Labrador trend ” at high angles, and often at right angles, is shown by Prof. Daly's map (1909, p. 88), and that Hamilton Inlet (i.e., Lake Melville) lies across the strike of the rocks, is shown by Mr. Kindle (1924, p. 16).
D. THE DIRECTING LINES OF SOME CANADIAN LAKES.
The independence of geological structure shown by the fiord valleys is similar to that of many Canadian lakes. This fact is shown by maps published by the Geological Survey of Canada for western Quebec and eastern Ontario, as by M. E. Wilson, Kewagama Lake Map Area, Quebec, Geol. Surv. Can., Mem. 39, 1913, pp. 18, 19 ; by W. H. Collins, the Onaping Map Area, Geol. Surv. Can., Mem. 95, 1917, pp. 16-18, and by M. E. Wilson, Timiskaming, Geol. Surv. Can., Mem. 103, 1918, pp. 35, 36 and 38. The Authors in presenting these maps remark that the linear valleys are
independent of the rocks, of the geological structure, and of the action of either rivers or ice, and must be due to earth movements ; they however point out that there is very little evidence to connect them with faults.
E. THE FIORD VALLEYS FORMED ALONG CLEFTS IN UPLIFTED ROCKS.
The directing lines in these cases may be explained not as ordinary faults, but as clefts produced in the tension of the country during a long gentle uplift which ruptured the rocks. The arrangement of the valleys thus produced resembles that of cracks in a pane of glass which has been broken by bending, or in cooling enamel or in a drying block of cement. The shrinkage of a drying block of cement produces cracks, some of which are parallel to the edge, and others that run in from the edge. If the block is homogeneous and the cracking is produced simply by drying, the cross cracks will tend to be at right angles to the faces ; but if the block while drying is undergoing bending, the cracks will pass in at an angle determined by the stresses due to the bending.
This tectonic origin of the fiord valleys of Labrador was recognised by Packard (1891, p. 18), who attributed them to “original lines of fracture and faults.”* The zigzag course of the Bowdoin Canyon below the Grand Falls on the Hamilton River is regarded as of a similar nature. The canyon consists of half-mile lengths, trending alternately to east and to south-west (Low, 1909, p. 154). Prof. A. W. G. Wilson of Montreal (Jt. App., p. 2605) attributes this course to the influence of “ two series of joint fractures” : i.e., to clefts due to tension.
The existence along the Labrador coast of many minor clefts that where probably made at the time of the formation of the major clefts which are now the fiord valleys, is shown in Prof. Daly's photographs. The cleft in the Bishop's Mitre is probably of a similar nature as I remarked (Fiords, 1913, p. 284).
Fig. 3, near the mouth of Nachvak.
„ 6, N.W. of Gulch Cape, near entrance to Nachvak.
„ 14, Cliffs north of Cape Mugford.
„ 15, the same (2 clefts)
„ 16, Cape Mugford (2 well marked clefts ; not a fault).
„ 17, Cliffs north of Mugford Tickle (beds in low arch ; clefts faintly marked). Clefts better shown in same place, fig. 21.
„ 18 & 19, Kaumajet Mountains at Mugford Tickle (clefts).
* Packard further remarks (1891, p. 18), “ We must believe that the same causes that produce the deep fiords likewise account for these deep fissures and depressions in the summit of the water-sheds. It is evident that any amount of glacial action, however long sustained and vast in its operation, can never account for these rude, irregular, often ' geoclinal ' troughs which follow lines of fracture and faults, lying along the axis of elevation of mountain chains, or at nearly right angles to them.”
In my book on Fiords I have endeavoured to prove that the fiords of the world are all due to the rupture of the earth's crust during the uplift of areas of hard rocks. The inlets of Labrador appear to me explicable only by the fracturing of the coastal belt during its uplift and the foundering of the adjacent parts of the Atlantic Ocean.
The directions of the main lines in Labrador are dominant in several parts of the North Atlantic borders, and the St. Lawrence—Hamilton Inlet line is a continuation of that of the main part of the coast of East Greenland.
F. DIFFERENT PLAN OF THE LABRADOR RIVER VALLEYS.
The direction of the fiords in the coastal belt of Labrador seems to me fundamentally distinct from the arrangement of a series of rivers. The natural arrangement of river valleys is well illustrated by the Labrador Peninsula. Rivers usually rise on the main highland and flow radially to the coasts. The most extensive highland in Labrador is an area at about 1,800 ft. above sea-level in the centre of the Peninsula ; from that highland many rivers, such as the East Main, Big River, &c., flow westward to James Bay. Others go north to Ungava Bay, such as the Kaniapiskau River, the Whale River, and the George River. The Hamilton River flows E.S.E. until it is diverted E.N.E. by falling into the Hamilton Inlet Valley. From the highland numerous rivers flow southward to the St. Lawrence ; some go south-westward until they are diverted eastward through Lake St. John and the deep Saguenay Fiord, or reach valleys which carry their waters to James Bay or through the Ottawa River to the St. Lawrence.
These radial rivers of Labrador have the normal arrangement of rivers draining such a plateau. They present a striking contrast to the rectilinear network and short deep clefts of the valleys and fiords of the Atlantic coast-land. The fracturing of the Labrador coastal belt appears primarily due to its uplift combined with the subsidence of the North Atlantic Ocean in comparatively recent geological times. That movement accordingly gave the Labrador coastal region its geographical unity.
It is necessary next to consider the natural inland boundary of that coastal belt.
V. THE THREE POSSIBLE BOUNDARIES.
Three lines have to be considered as the possible western boundary of the Dependency of Labrador.
1. THE SHORE STRIP.
The first is a line one mile in from the shore ; but the allocation of a one-mile strip of shore throughout a long coast to one state, while the interior is left to another, would appear to be a very inconvenient and unprecedented arrangement. The development of the interior of Labrador is sufficiently handicapped by its climate, its inaccessibility, and the poverty of its soils