undulations entering them such marked peculiarities that “ river tides ” require special treatment in works upon the tides, while in works on rivers and hydraulics special treatment is commonly accorded “ tidal rivers.” Indeed, “ tidal rivers ” frequently form the subject of special treatises, like those by Calver, Unwin, Bouncieau, Mengin, Partiot, Wheeler, and others. Among the peculiarities of river tides are :—
(a) An unusually marked retardation in the advance of the tidal wave, clue in large part to the opposing force of the river current and to the ascent of the river slope.
(b) A progressive diminution in the vertical range of the tide, due in part to the ascent of the river slope.
(c) A distortion of the form of the tidal wave, with resulting inequality in duration of rise and fall of the tide.
(d) A marked increase in duration and velocity of the ebb current, and a decrease in the duration and velocity of the flood current.
The tidal surveys carried on in the Narrows in 1923, and reported in full in Volume IV of the Joint Appendix, pp. 1913-1957, were, in my opinion, sufficiently accurate in character and carried on for a period sufficiently long, to demonstrate beyond any doubt that all four of the above peculiarities of river tides are strikingly developed within the Narrows. Thus the rate of advance of the crest of the tidal wave, which in Hamilton Inlet proper averages two and one-third miles per minute, drops suddenly to an average of little more than one-third of a mile a minute within the Narrows (Joint Appendix, Second Proof, TV, p. 1929. Final Print, Vol. V, p. 2335). The vertical range or amplitude of the tide is increasing slightly as the wave advances up the Inlet as far as Tikoralak Head ; but on entering the Narrows it drops off quickly to about 78 per cent. of its normal value at Rigolet (Final Print. p. 2335), while farther inland towards Lake Melville it is still less (as noted above, this is due only in part to river slope). The distortion of the tidal wave in the Narrows is evident from the fact that the period of rising tide averaged but five hours and fifty-six minutes, whereas the fall required on the average six hours and twenty-nine minutes (Final Print, p. 2335). In duration the ebb current everywhere predominates over the flood current, the maximum excess period of the ebb amounting to one hour and seven minutes in mid-channel on the surface (Final Print, p. 2332). In velocity also the ebb current predominates everywhere over the flood current, the measured velocities of the ebb averaging 80 per cent. greater than the flood (Final Print, p. 2332). Thus in every detail the peculiarities well known to characterize tides in rivers are strikingly manifest in the Narrows, affording convincing confirmation of the fact that the Narrows is a true river.
Long period observations might alter slightly the precise values given above ; but I do not see how they could possibly change the essential facts pertinent to the issue before us. If the object were to determine with
highest possible precision exact values for each of the above-mentioned factors, the long-period observations recommended by Vice-Admiral Learmonth would be necessary. But where the object is, as here, merely to determine whether or not the peculiar features of tidal rivers are present in the Narrows, no such necessity exists. In fact, without any surveys one could, having before him pertinent information regarding the physical features of “ the Narrows,” safely conclude that this lower portion of the Hamilton River system must exhibit the peculiarities described. It appears, indeed, that the “ river slope ” through the Narrows, a feature involved in the peculiarities of tidal rivers, was not only predicted, but its value was estimated with considerable accuracy, before any surveys were undertaken (Joint Appendix, Second Proof, TV, p. 1956, Final Print, Vol. V, p. 2359). The surveys merely demonstrate the development in the Narrows of tidal river characteristics to an unusually marked degree—so marked that no conceivable changes in rates of river discharge, or in other seasonal variations of physical conditions could possibly change the essential relations.
In respect to the Vice-Admiral's observations that the peculiarities observed in the Narrows and Lake Melville are also found in sea inlets “ of this configuration and depth,” “ where a connecting waterway with a moderate tidal range expands into a large deep basin, into which a very extensive drainage falls from the distant watershed of the interior,” and in part depend on “ the depth, width, and sinuosities of any waterways offering obstruction ” to the travel of the tidal undulation, it is to be noted that the physical features thus described are not found in true sea inlets, but only in tidal rivers and estuaries including tidal lakes. It is because the Vice-Admiral makes navigability the test of a sea inlet, that he is able to class tidal rivers as inlets, and then to compare the Narrows with them.
When a land area has been subjected to glaciation, the river systems have impressed upon them certain features which cause physiographers to apply to them the term “ disturbed drainage.” Some rivers are turned from their former valleys into new courses, often flowing from one broad, pre-glacial valley or lowland across a former divide (usually via minor side valleys or ravines) into another broad, pre-glacial depression. Erosion in time transforms the new course through the former dividing range of hills or mountains into a gorge or canyon which contrasts strangely with the broader depressions occupied by other parts of the same stream. Where the new course leads over cliffs or ledges, waterfalls develop. Glacial debris obstructs stream valleys, ponding or laking the waters, and often diverting the streams into erratic courses. Glacial erosion over-deepens parts of the valley floors to produce lake basins in the broad valleys, smaller “ deeps ” in very narrow valleys or channels. Where belts of weak rocks favour extensive erosion, broad and deep lakes are apt to develop. Where a narrow gorge compresses the ice within limited compass, thus accelerating ice-movement and increasing erosive power, local deeps may alone result.
Labrador is such a glaciated land, and its drainage system well displays the features described above. The Narrows seems to be a typical example
of a gorge cut transversely across a former divide and connecting two broader lowlands, one occupied in part by Lake Melville, the other largely submerged to form Hamilton Inlet proper; Lake Melville is one of many similar lakes which repeatedly interrupt the river courses. The falls in the Hamilton River are just as normal characteristics of “ disturbed drainage ” as are the lakes. In the Narrows. “ deeps ” interrupt the average seaward descent of the valley floor, giving depressions greater than are found in adjacent parts of Hamilton Inlet proper. These are typical of narrow gorges traversed by ice-currents at some stage of the glacial period, and call to mind the much greater deeps in the Hudson gorge through the Highlands of New York, where depths between 765 and 950 feet below sea-level were scoured in solid rock by the ice current. (These were later partly filled by glacial debris,
but their initial depths were greater than any known in the lower course of the river, or in adjacent parts of the sea floor.) In short, not a feature of the Lake Melville-Narrows district is unusual or without its counterpart in other lake basins and river valleys in the glaciated regions of America. Were they located farther inland, not a doubt would arise as to their true character. It is only if we allow ourselves to be confused by the presence of salt water and tidal phenomena, common enough in lakes and rivers near the sea, that any question as to the essential nature of Lake Melville and the Narrows can trouble us.
It is probably safe to say that if Lake Melville alone were farther inland, and the Narrows remained precisely as they are, no one would question that Hamilton River was continuous to the mouth of the Narrows, and that the sea began near Tikoralak Head. Fig. 3 shows the conditions essentially
as they would then appear. Can it be doubted that with these conditions the river would clearly be indicated as a unit to the point where it enters Hamilton Inlet proper, opposite Tikoralak Head ? The slightly expanded mouth of the river is clearly differentiated from the greater expansion of the Inlet, not only by a distinct change in the course of the shores, but also by the fact that the river does not enter the apex of the Inlet, but a little farther east, distinctly to one side of the Inlet's axis. And if the Narrows would clearly constitute the tidal lower portion of Hamilton River were there no lake, it cannot be held to belong in a different category simply because a lake happens to interrupt the river's continuity. The interposition of a lake in a river's course effects no change in the essential nature of that river, either above or below the lake. The lake may regulate the rate of flow in the outlet stream, and dampen the effect of floods ; but that stream is a river in precisely the same degree, whether or not the lake be present.
LAKE MELVILLE IS NEITHER A FJORD, NOR A BAY INDENTING
THE COAST. IT IS A TRUE LAKE.
In his interesting memorandum on “ Geographical Considerations as to the Canadian-Newfoundland Boundary in Labrador ” Professor J. W. Gregory devotes the greater part of his text to a discussion of the “ fjord problem ” as related to the Labrador coast in general and to Lake Melville in particular. With all due respect to this distinguished authority, it does not seem to me that this discussion is relevant to the present issue. It seems irrelevant for two reasons :—
(a) First, because it relates to fine distinctions between, and controverted theories of origin concerning, different kinds of inlets (fjords, fiards, rias, etc.) which are of much interest to the physiographer, but which are of no practical value to the geographer, lawyer, or diplomat charged with the duty of distinguishing between marine and terrestrial water bodies ; and
(b) Second, because fjords, no matter what theory of origin one accepts, are never the product of marine agencies, and have no genetic relation to the sea ; they are basins of independent origin which are called fjords, fiards, etc., if occupied by true arms of the sea, but are called by other names if occupied by terrestrial water bodies. Thus, even if one were to accept Professor Gregory's theory respecting the origin of fjord basins, we must still determine first whether Lake Melville is a lake or an arm of the sea before we can apply to it the term fjord.
A further word of explanation may serve to clarify these two points.
THEORIES OF FJORD FORMATION.
(a) The origin of fjords has long been disputed. In various countries geologists and geographers have contributed a voluminous bibliography to the subject, which I have reviewed at some length in my volume on “ Shore Processes and Shoreline Development.” In order to show the complexity of the subject I cannot do better than repeat what I said there regarding the fjord problem, in words written more than ten years before the Labrador boundary question was brought to my attention. Incidentally the quotation will reveal my doubts regarding the correctness of Professor Gregory's ideas on fjord formation, and will serve to emphasize the fact that whether or not the term fjord is applicable in any given case depends wholly on whether or not it happens to be occupied by a true arm of the sea.
Fjord Shorelines.-Perhaps no type of shoreline has given rise to so much discussion as has the fjord shoreline. We may note in the first place that geologists and geographers may be divided into two main groups whose ideas regarding the origin of fjords are mutually opposed. The first group may be designated as the “ glacialists,” because in their opinion all the phenomena peculiar to fjords may be explained as the result of extensive glacial over-deepening of pre-glacial river valleys near the sea. The second group, or “ non-glacialists,” reject the theory of ice erosion, and attempt to account for the phenomena of fjords in other ways.
According to the glacial theory, fjords are partially submerged glacial troughs. The troughs of glaciated mountains far from the sea are similar to fjords, except that the former have not been drowned by marine waters. In both cases the troughs were formed by extensive glacial over-deepening of former river valleys. The pre-glacial valleys guided the glaciers which later came to occupy them, and by confining the ice streams to the narrow limits imposed by the valley walls, insured a maximum efficiency of glacial erosion. The glacial theory asks no questions as to what determined the courses of the pre-glacial valleys ; but it is fully recognized that among other causes ancient fault lines must be considered, since a fault may give a crushed zone which is weaker than the unfractured rock, or may bring a belt of weak rock into such position that subsequent valleys will soon be excavated along it, parallel to the fault. This would satisfactorily account for the fact that many fjord shorelines have a more or less angular pattern.
Esmark was the first to advocate the glacial origin of fjords, almost a century ago. The fjord valleys of New Zealand were ascribed largely to ice erosion by von Haast in 1865, while Helland a few years later, in discussing the fjords of Norway and Greenland, gave the best exposition of the glacial theory as applied to the interpretation of fjords which had appeared up to that time. Helland seems to have anticipated Shaler in recognizing the ability of glaciers to excavate their channels below sea-level, and to have given a fairly good account of the essential significance of hanging valleys some twenty years before Gannett's classic statement. The influence of rock fractures on the orientation of fjord valleys was recognized by Brögger,