of weak rock, but is abnormal for a fjord. Professor Gregory notes that the Lake Melville basin extends below sea-level, but this is not peculiar to fjords, lakes of every type have basins deeper than their outlets, and many of those near the sea, and some far inland (e.g., the Great Lakes) have their bottoms below sea-level. Depth is merely a matter of how effectively the glaciers happened to erode, how far a fault block dropped, or other extraneous factors. Professor Gregory is able to classify Lake Melville as a fjord only because he makes fracturing the primary cause of fjords ; but as has been pointed out above (a) the fracturing is vastly more ancient than the period of fjord formation, and (b) fractures are as commonly found associated with river valleys and lake basins as with fjords. The fact that earlier writers applied the term fjord to Lake Melville, a fact stressed by Professor Gregory, should not mislead us. It was common practice in earlier days, especially among geologists who were not specialists in physiography, to use the term “ fjords ” very loosely for narrow sea inlets of all kinds and for lake basins closely associated with them. Even inland lakes with steep-walled basins were sometimes called “ fjords.” Thus the name “ fjord ” has been applied to bays and lakes on the coasts of Maine, Massachusetts, Nova Scotia, Cape Breton Island, Denmark, and to lakes in Switzerland, none of which are now classed as fjords.
I have stated in the above terms my reasons for rejecting the contention that Lake Melville is a fjord, not because I believe this theoretical matter pertinent to the present issue, but merely to make clear the fact that Professor Gregory's whole memorandum is based on a theory of fjord origin which has been rejected as untenable by most physiographers, on a history of the Labrador coast which is in conflict with that worked out for the whole north-east coast of America by those geologists and physiographers best acquainted with the region, and on a standard of fjord topography which is at variance with that generally recognized by specialists in land forms. As an eminent authority, Professor Gregory is most assuredly entitled to hold and defend these several views, and to demonstrate, if he can, their superiority over the views more generally accepted throughout the geologic fraternity. But coast-lines cannot be delimited, or territory assigned to sovereignty upon any such debatable grounds.
LAKE MELVILLE NOT A BAY OF THE OCEAN.
True arms of the sea may result from the extensive submergence of former river valleys (producing rias), glacial troughs (producing fjords), or other types of depressions. Whatever their origin they have certain features in common. The valleys, troughs or other land features are sufficiently drowned by marine submergence to change completely the general appearance of the region. The sea has free access to the depressions ; it may find such access through narrow inlets, or large inflow of water from rivers may freshen the headward portions of the sea arms, but the sea does not have to make headway against a definite outflowing current of land water. The mean level
of the sea arm is substantially that of the open sea nearby. Tides are not greatly affected by river action, so fail to develop in any marked degree, if at all, the behaviour characteristic of tidal rivers.
In none of these respects does Lake Melville resemble a sea arm. There is no extensive marine submergence of the basin. The lake is deep, but most of the submergence actually existing would remain if the sea were a thousand miles distant and thousands of feet lower. Of the thousand feet of depth recorded for the lake only the upper 150 is due to the entrance of the sea ; the rest would persist if there were no sea. Even with this addition the floor of the basin is still visible over large areas as a low plain bordering the lake. The essential form of the basin and its narrow outlet are little obscured by such sea-water as finds its way into the depression. Access of sea-water to the basin is not free ; the incoming flood must overcome the outflow of the largest river on the east coast of America north of the St. Lawrence, a river draining a basin so vast that all of England could be put into it, with plenty of room to spare. Against this outflow the sea must struggle for miles along a narrow and crooked river valley. The mean level of the lake is above the mean level of the sea, even when the river outflow is not at its maximum. Tides cannot reach the lake without first undergoing the striking modifications characteristic of river tides. The lake is not directly connected with the sea, but as shown in a preceding section, is separated from it by a narrow tidal river 20 miles in length. For all of these reasons it would be altogether improper to classify Lake Melville as a bay or arm of the sea.
LAKE MELVILLE IS A TRUE LAKE.
A careful consideration of all the evidence makes it clear that Lake Melville must be classed with the countless other lakes which diversify the glaciated area of north-eastern America. Deep lake basins, while far less numerous than those of moderate depth, are none the less well recognised as normal features in glaciated districts. The Hamilton drainage system includes, in addition to Lake Melville, Grand Lake with a depth of 540 feet, and Lake Winikapau, far in the interior, over 400 feet deep. Lake Superior has depths slightly over 1,000 feet, Lake Michigan over 800 feet, and Lake Ontario over 700 feet. Kindle has referred to Lake Mjosen in Norway, with a depth of more than 1,000 feet. Nor is there anything peculiar in the fact that the bottom of Lake Melville is below sea-level. Of the various lakes mentioned above, all but one have their bottoms hundreds of feet below sea-level, notwithstanding the fact that most of them are far inland. The great depth of Melville is not only normal for a glacial lake, but it finds a ready explanation in the presence of a belt of weak sandstones, first etched out by streams to form a broad lowland in pre-glacial times, then easily deepened by ice erosion.
Salt water and tides in lakes are of frequent occurrence where such lakes are near the coast. Lake Kennebecasis in New Brunswick, with a level
distinctly above that of mean sea-level * is entered by the sea at high tide, and is itself tidal. Inasmuch as Lake Kennebecasis is directly connected with the sea by a very short tidal inlet it is far more closely allied to bays than is Lake Melville, and while such classification is certainly incorrect, it is understandable why it should sometimes be called a bay. This is done in certain maps and atlases, and I have myself been guilty of the error. Nevertheless, it is more widely known as a lake, is regarded by the Canadian Department of Marine and Fisheries as an expansion or lake of the St. John River, and the fishery in the lake is regarded and treated in the Department's Regulations as a river fishery.†
Lake Pontchartrain in Louisiana is separated from the sea by a narrow channel eight miles long, through which both salt water and tides enter the lake basin. Yet David Starr Jordan cites it as a type example of lakes formed by the damming effects of river deposits.‡ The Secretary of War of the United States has stated that the War Department regards Lake Pontchartrain as “ a body of water distinct and separate from the Gulf of Mexico, and hence a waterway, the navigable portions of which lie wholly within the limits of a single state.”§ It may be added that although this lake is mostly fresh, sea fish find their way into the basin with the salt water, and have been observed there by Jordan mingling with fresh water species. The lake is accessible to small ocean-going vessels drawing nine feet. Both the long narrow channel leading into the lake, and the name, Rigolets, remind one curiously of the entrance to Lake Melville. In this same region Lake St. Catherine, while nearer the sea, is fresher than Pontchartrain although entered by salt water and the tides, Lake Maurepas is farther inland, yet subject to tidal undulations and to some access of salt water. Grand Lake, like Lake Melville, is separated from the sea by an irregular channel 20 miles long, and is generally fresh, yet salt water and the tides penetrate into it. Farther west, on the Gulf Coast, Calcasieu Lake and Sabine Lake, through which large rivers of the same names discharge are separated from the sea by narrow channels a few miles in length. Both lakes are brackish from sea water, and both are tidal, and it may be added that both basins differ in origin from those previously mentioned. All of the foregoing lakes in the Louisiana-Texas area are true lakes in every sense of the word, and are accessible to small boats entering from the Gulf of Mexico. Furthermore, “ all of these lakes are classed as inland waters by the Steamboat Inspection Service, i.e., the pilot rules for inland waters or for western rivers apply on these lakes as distinguished from the high seas where the international rules apply.”**
* The difference is at least two feet, because differences of 1.96 and 2.04 feets were determined by two lunar months' tide-gauge readings during the autumn level of the river, the gauge in the lake being well down toward the tidal inlet.
† “ Special Fishery Regulations for the Province of New Brunswick,” 1924, p.17, Sec.32.
‡ “ High Lights of Geography, North america.” Jordan, D.S., and Cather, K.D.,p.102, 1926.
§ Personal communication from United States Engineer's Office, New Orleans, La.
** Personal communication from U.S. Coast and Geodetic Survey, Washington, D.C.
PHYSICAL CONTRAST BETWEEN HAMILTON RIVER AND HAMILTON INLET.
No one familiar with the loose and variable way in which names are frequently printed on maps and charts can attach any significance to the fact that the name Hamilton Inlet has occasionally been so printed as to cover the tidal lower portion of the Hamilton River system. The distinction between the true inlet and the true river system must depend, not on the whim nor even on the best judgment of the cartographer,* but rather on the fundamental physical features described in preceding sections. It is submitted that those physical distinctions have been shown to he real and practical, but it may emphasise the validity of those distinctions to apply a
very simple and non-technical test to the whole area here under discussion. This is to imagine the sea to be entirely withdrawn from the area, as would be the case for example if sea-level were to drop, say, 1,500 feet. The result would be accomplished by a drop of but one-fifth this amount, but to avoid any confusion we may assume the figure of 1,500 feet in order to carry the sea far below the lowest part of the Lake Melville basin.
Fig. 4A is an outline of the present hydrography of the region, in so far as it relates to the Lower Hamilton River, Lake Melville, the Narrows
* In preparing maps and charts the United States Geological Survey, The United States Coast and Geodetic Survey, and presumably similar government bureaus in other countries follow the practice of naming physical features as they understand them to he named locally. Experience shows that not infrequently their understanding is a mistaken one, due perhaps to insufficient inquiry on the part of the surveyor.
and Hamilton Inlet, and is based on surveys by Low and Kindle of the Geological Survey of Canada, and on Chart 420 of the Canadian Naval Service. Fig. 4B is an outline of the hydrography as it would appear if the sea were to abandon the area entirely, consequent upon a 1,500 foot drop in sea-level. It is based in part on contours drawn from soundings and in part directly on soundings. For Lake Melville and the Narrows the data are found on Chart 420 and is abundant. If we assume that the ponded lake waters will have a depth of 35 feet where they escape through the narrow channel at Pike Run, at the upper end of the Narrows (an assumption which cannot be far wrong), the 20-fathom subaqueous contour, already drawn on Chart 420. except where close inshore, will give with more than needful accuracy the shore-line of Lake Melville under the new conditions.* Goose Bay, however, will be held at a slightly higher level by the narrow and shallow channel which connects it with the rest of the lake.
For Hamilton Inlet proper the data are far less abundant, although sufficient for our needs. Soundings (Admiralty Chart No. 375) are too few in number to make possible the drawing of accurate contours, but a study of these soundings makes it clear that after the drop in sea-level there would remain no large water bodies to require contouring. Near the head of the funnel-shaped inlet or embayment there is a deep area revealed by soundings of thirty-four, thirty-three, fifty and forty-six fathoms. Eastward there is a belt several miles broad in which the deepest soundings are twenty-two and twenty-three fathoms. Still farther eastward there are evident channels deeper than twenty-three fathoms. From these figures we may be sure that with the drop of sea-level all water would drain off eastward down to the twenty-two fathom level. Below this there might persist in the head of the embayment a small lake (L) occupying the “ deep ” above referred to. If there is a channel deeper than twenty-two fathoms, not revealed by the soundings, the lake might be smaller than shown in Fig. 4B, or it might be wholly replaced by a normal river channel, but in no case could it be very melt larger than it is represented.
Eastward from the lake there appears to be a depression or channel twenty to twenty-five miles long, running close to the south side of Saddle Island (S), thirty to forty fathoms deep throughout most of its length, and
forty to fifty fathoms at its extreme eastern and north-east of George Island (G). In places this channel may shallow to depths of twenty-five fathoms, as depths of twenty-five and twenty-six fathoms are recorded at two places, or it may be that the deep channel continues in the areas without soundings in this section of the Inlet's floor. Soundings toward the south-east from
*Unpublished soundings brought to my attention since this was written indicate the presence of a 16-fathom shallow or bar in the Narrows just above Rigolet. The effect of this would be to hold the water above the bar at a slightly higher level than I have assumed. Consequently the Lake and Narrows under the new conditions would resemble the present Lake and Narrows even more closely than I have indicated in Figure 4B.