without having the extra burden of having its sea-shore under a different administration which might impose a burdensome tariff. The natural outlet from the belt as far back as the watershed is to the Atlantic. The Indians of the interior go to the seashore as at Davis Inlet, for trade in the spring (Low 1895, p. 48), and they go to the coast in severe winters (Newfoundland and Labrador Pilot, 1907, p. 643). The Livyeres or permanent residents on the coast go inland in the winter for hunting, visiting lines of traps, which, according to A. P. Low (1895, p. 43, 1909, p. 144, and Jt. App., p. 2608), are 50 miles in length. Dr. W. T. Grenfell (Jt. App., p. 2571) says that the “ furriers ” of Sandwich Bay go 350 miles inland. On many parts of the coast there is clearly no firewood on the shore, and the residents might be cut off from the only local supply of fuel. I can think of no case in which such a narrow strip along the shore is owned by one state while the interior belongs to another. The Zanzibar Protectorate, which is ten miles wide, along the East African Coast, is the closest analogy ; but it and the land behind are both under the direct control of the Colonial Office, and there seems to be no real analogy either historical or geographical, between these two cases.
a. The Term Coast in Central Geography.
It may be urged that the term coast means the shore line, but such does not seem to me the correct geographical meaning. The term coast has been largely used, especially in the 17th and 18th centuries, for a border-land region, and not for a mathematical line, or for a narrow strip along the meeting place of land and sea. When Herod, as the Bible states, slew all the children in the coasts of Bethlehem, it would not have served his purpose merely to have slain those on a line on the outskirts of Bethlehem. The term coast there obviously means the tract of country including Bethlehem and its neighbourhood. Most authoritative British Dictionaries recognise the use of the term coast for a broad belt of country. The term is said to come from the Latin costa, a rib, and has been used for side or rib, and would therefore mean a side of the Labrador Peninsula. The Encyclopædic Dictionary, 1889, Vol. II, gives as the first meaning “ a side ” ; as the fourth, “ frontier limit or border of a country ” ; as the fifth, “ the seashore ” ; and as sixth, “ a district, a country, a region.” The New English Dictionary, Vol. III, 1893, after pointing out the origin of the word coast from costa, its resemblance to) côte, and its primary meaning as a rib or side, gives as its fourth meaning “ the seashore ” ; as the fifth, “ the borderland” ; and as the sixth, “ a tract or region of the earth.”
Johnson's Dictionary of 1755 gives a restricted definition of the word coast ; but his original Dictionary is famous for its freakiness, and the edition of 1818, to quote that nearest 1825, gives as the second meaning, “ 2, Border, limit or frontier of a country,” and remarks that this sense was annoticed by Johnson.
The term coast is still widely used in. geography for a region as in “ the Gold Coast,” and the “ Coastal Cordillera ” of South America, which are often 60 miles back from the shore.
b. The Term Coast in Canadian Geography.
Its similar use in Canada may be illustrated by Coast being the name of a sub-division of British Columbia which extends for 150 miles inland, by the Coast Range of western Canada extending at least 100 miles inland, and as in Labrador the Coast River, lat. 55½° N., does not itself reach the sea, from which it is 25 miles distant at its lower end and 90 miles at its source.
The distinguished Canadian authority on Labrador, Mr. A. P. Low, applies the term coast to the tract of country extending a considerable distance back from what he (Low, 1895, p. 21) distinguishes as “ the coast line.” North of Nain, he says, p. 22, the Coast Range is much higher, being from 4,000 to 6,000 feet. He refers to the northern mountain range as confined to the coast region, so that the coast region includes the whole of that mountain range, and must therefore extend at least 50 miles inland.
On p. 23 he states that the northern half of the Atlantic coast rises in a chain of mountains higher than any other part of the Peninsula. On p. 25, he speaks of “ the highlands of the coast cutting off the drainage of the interior and forcing it to flow northward into Ungava Bay.” So according to Low the watershed is determined by the coastal highlands. Low, therefore, extends the coast inland to the watershed and to include the whole of the northern chain.
2. ALONG THE COASTAL MOUNTAINS.
A second possible line would be one joining the highest summits of the mountains which lie inland from the shore and passing along the crest of the former mountain chain. If the ancient chain were still complete its crest might have furnished a suitable frontier between the coastal belt and the interior. The chain is now most nearly complete in the northern part. “ Taken as a whole, the Nain-Kiglapait-Kaumajet-Torngat Mountain group is a distinct entity, a distinct topographic province of the Labrador Peninsula ” (Jt. App., p. 2537). Even in the northern part from Cape Chidley to Nain it is broken through by a series of deep valleys, so that a line joining the highest mountain summits would be a very inconvenient and geographically undesirable boundary. That type of frontier was claimed by Chile in its dispute with the Argentine, and was rejected by King Edward's Award.
3. THE WATERSHED.
a. Its Convenience.
The only convenient western boundary of the “ distant entity ” and “ distinct topographic province ” is the watershed, which separates the rivers that flow across the mountain belt from those that cross the western plains to Hudson Bay. South of Nain the mountain belt is less complete. Prof. Daly remarks, “ South of Nain there is no well-defined existing mountain-chain,” but the continuity of the foundation is shown by
Prof. Daly's figure (1909, p. 88). Fragments of the ancient chain remain as the Kokkoke Mountains with Mount Muckamee (alt. 4500 ft.), north of the western end of Lake Melville (see Admiralty Chart No. 375) and the Mealy Mountains. The intervals between the southern remnants of the mountain chain are so great that even if the position of the original mountain crest could be determined, it would form an even less satisfactory frontier than north of Nain. The most suitable boundary between this mountainous dissected coastal belt and the plains to the west is the main watershed, with the possible exception of the upper part of the Hamilton Basin.
The term watershed means the line separating the drainage as it comes from the German “ wasserscheide ” or “ water separation.” It is sometimes incorrectly used for the slope of the land from the divide. Watershed means precisely the same as “ height of land ” as used in Canada. The “ height of land ” in Labrador separates the fiord-riven coastal region from the plateau across which many rivers flow westward to Hudson Bay. The watershed for most of Labrador is the most convenient and natural geographical boundary. For owing to the depth of the fiord valleys all communications in northern Labrador necessarily pass along the valleys to the coast. As remarked previously (p. 22) the Indians trade down these valleys to Davis Inlet, and the permanent residents on the coast during the winter work inland along miles of traps which, according to Low, extend 50 miles in length, and probably reach to the “height of land.” The Admiralty Pilot remarks that “ in summer travelling can only be done by sea ” (Newfound-land and Labrador Pilot, 1907, p. 643).
The suitability of the watershed in Labrador as the boundary is indicated by the numerous authorities who have adopted it, amongst whom may be mentioned the late Dr. Robert Bell, formerly Director of the Geological Survey of Canada in his map (Scott. Geogr. Mag., Vol. XI, 1895, opp. p. 360). He remarks (ibid. p. 336) that “ Newfoundland claims about 140,000 square miles, the remaining 420,000 belonging to Canada ” ; so that he regarded a quarter of the peninsula as not belonging to Canada.
That the “ height of land ” was intended to be the western boundary of the Dependency of Newfoundland is suggested by the line selected for the frontier from the St. Lawrence. It was defined as inland due north from Blanc Sablon to the 52nd parallel, a distance of over 40 miles. There the defined line was near the “ height of land ” between the drainage into Hamilton Inlet, and that into the St. Lawrence. That the intention was to follow the watershed from that point is shown by the frequent adoption of that line as the boundary in the maps of that period.
b. The identity of the Rocks on the two Sides of the Watersheds.
It may be represented that in a plateau country the watershed is not an important or logical boundary ; but on such a plateau it is usually the natural line and the line of greatest economic importance, as it is the main factor in the direction of trade and traffic. Timber is naturally floated
downstream from the divide, and all bulky produce tends to go the same routes. It has been objected that this line is artificial as the rocks on both sides are identical. That objection would also apply to the watershed in Scandinavia, which is also a vast plateau of rocks of the same age and general nature as those of Labrador. There is the same unity of geological composition between Norway and Sweden as between Labrador east and west of the watershed.
VI. ANALOGY OF SCANDINAVIA AND LABRADOR.
This reference to Scandinavia recalls the instructive analogy between the peninsula and Labrador. As the geology and geography of Norway are better known they throw light on the origin of the Labrador fiords, and on the natural geographical divide in such a country. Scandinavia, like Labrador, consists of a foundation of rocks belonging to the same geological date, and most of the country has also remained above sea-level through most of geological time. Scandinavia also consists of a fiord-riven coastal belt facing the Atlantic, and of a long gentle slope away from the ocean toward an inland sea. As the present structure of both Labrador and Scandinavia was impressed on them from the Atlantic the sequence of the corresponding belts is reversed. Scandinavia begins on the west with an island fringe known as the Skargaard, which consists of an enormous number of islands ranging from mere rocks to islands of a considerable size. The mainland rises from the coast, especially to the north, in abrupt precipices, and is indented by long fiords. Further south a low rocky platform often separates the plateau front from the sea, and the inlets in this platform are fiards. The coastal part of the main plateau is cut up by an intersecting network of trough valleys into isolated mountainous blocks. From the watershed the plateau extends inland and bears hills of moderate elevation and innumerable lakes. This plateau descends by a long slope traversed by numerous rivers which are much longer than those flowing to the ocean, and this slope ends in an inland sea, the Baltic.
The structure of Labrador is essentially the same. The Atlantic coast has its guard of islands ; to the north the plateau front descends precipitously into the sea, and the coast is indented by long fiords ; to the south there is a low lying rock platform with fiards. The meeting of fiord valleys cuts up the coastal belt into isolated blocks. In contrast to the broken topography of the coastal region the plateau to the west has a relatively level surface with low hills, broad valleys, and innumerable lakes. The plateau descends in a long slope drained by many rivers, and ending in the inland sea of Hudson Bay.
The geographical development of Scandinavia along the fiord-riven region was slow owing to the small amount of cultivable land, and the extreme difficulty of land communication, which rendered the country dependent upon the sea. The Norwegians are therefore a sea-going people, and occupation spread slowly inland on the working of mines and forests.
The development of the slope towards the Baltic was much easier and towns were established far inland, while the settlements in Norway were still confined in the coast. The boundary which has been adopted between Norway and Sweden is for most of the distance the watershed between the rivers flowing to the Atlantic, and to the Baltic. Such departures from the watershed as have been made south of 60° were made to secure a suitable frontier where the watershed was inconveniently sinuous.
On the southern coast the area with the fiords which indent the land from the sea between Scandinavia and Denmark is included in Norway. The Scandinavian analogy is therefore in favour of the adoption in Labrador of the watershed as the general frontier.
J. W. GREGORY.
The Peninsula of Labrador consists of two main sections—an eastern belt facing the Atlantic which is high and mountainous, and is deeply indented by many arms of the sea (fiords and fiards). The western section is a plateau with a comparatively gentle undulating surface sloping downward to Hudson Bay and Ungava Bay. The high mountains and arms of the sea in the western section give it a distinct geographical entity. This belt has been called, as by some Canadian geographical authorities, the coastal region, and that term is in accordance with geographical and Canadian usage, as may be illustrated by the wide areas assigned to “ Coast ” in British Columbia. The Hamilton Inlet has been generally accepted as a fiord, and as the longest fiord in Labrador. Its entrance is not a ria, but a fiard ; and the inner (Lake Melville) is the essential part of Hamilton Inlet, and to it the name inlet was first applied. As a deep basin filled with seawater below a thin layer of freshwater, it is an arm of the sea. Its branch, the Double Mer, is accepted as a fiord, and Lake Melville has still better title to the name fiord.
The Labrador inlets were neither cut by ice nor by ancient rivers. Their arrangement shows that they were due to the fracturing of the eastern coast of Labrador during the earth-movements which upheaved the Labrador coastal mountains and led to the foundering of the floor of the North Atlantic basin. The inlets are true fiords ; their arrangement is essentially different from that of river valleys, the normal arrangement of which is illustrated by the river valleys on the Labrador plateau.
Labrador is strikingly similar in structure to Scandinavia, both peninsulas which in each consist of blocks of ancient rocks are found on both sides of the watershed. The watershed is the main boundary between the fiord-river mountainous region of Norway and the gentler slope of Sweden to the Baltic.