who did not fail, however, to attribute the actual excavation of the fjords to glacial erosion. In a similar manner Reusch for the Norwegian fjords, and Andrews for those of New Zealand, make a clear distinction between the rôle of faulting in determining lines of weakness favourable to rapid stream and glacial erosion, and the role of glaciers in giving to the fjords their present form and depth.
In 1895 Shaler, in discussing changes of sea-level, accepted the glacial origin of fjords and stated that since glaciers may cut their channels below the surface of the sea, the flooding of a glacial trough may be accomplished as the ice melts, without any sinking of the land or rising of the water-level. This same view, that fjords do not indicate past changes of level, was adopted by Hubbard in a brief review of the fjord problem which he published in 1901, by Daly in his account of the Labrador fjords, and by Andrews in discussing the fjords of New Zealand. It is further elaborated by Gilbert in his report on glacial studies, forming the third volume of the Harriman Alaska Series, where the reader will find a discussion of the physics of glacial erosion below sea-level. Marshall in his “ Geography of New Zealand,” and Tarr in his report on “ The Physiography and Glacial Geology of the Yakutat Bay Region, Alaska,” are among other students of fjords who attribute their excavation to ice erosion.
Members of the non-glacialist group are by no means in agreement among themselves as to the origin of fjords. They agree on one thing only—that ice did not excavate these deeply submerged canyons. Some consider fjords the product of normal stream erosion followed by a partial submergence which permitted the valleys to be drowned. This was the view expressed by Dana, who first emphasized the restriction of fjords to high latitudes, but did not suggest for them a glacial origin. Upham definitely rejects the glacial explanation, and follows Dana in considering fjords as drowned normal river valleys. Brigham and Hull seem to incline to the same view, the former speaking of “the common-sense conclusion that they are river valleys made tidal by drowning,” but both recognize that fjords have been to some extent modified by glaciers. Hirt in a review of “Pas Fjord-Problem,” Dinse in a more elaborate study of “Die Fjordbildungen,” and Grossman and Lomas, in a report of the Faroe Islands, tend to assign to glaciers but a moderate rôle in modifying pre-existing valleys ; while J. W. Tayler and Fairchild definitely reject the glacial theory of fjord formation, Fairchild specifically invoking coastal subsidence to account for the fjord embayments.
Among those students who admit that ice erosion played an essential part in fashioning fjord valleys, there are a number who either expressly require coastal subsidence, or else tacitly assume that subsidence is necessary for the drowning of the glacial troughs. Robert Brown, writing on “ The Formation of Fjords ” in 1869 and 1871, required the combined action of glacial erosion and coastal subsidence. The same view is supported by Remmers in his “ Untersuchungen der Fjorde an der Küste von Maine,” and by Güttner
in an essay on “ Geographische Homologien an den Küsten ” published in 1895. Those writers, assuming the necessity of subsidence without specifically discussing the point, include Penck in his “ Morphologie der Erdobcrfläche,” de Lapparent in his “ Traité de Geologie,” Gallois in his account of “ Les Andes de Patagonie,” Le Conte, in his “ Elements of Geology,” and Hobbs in his “ Earth Features and Their Meaning.”
Formerly many observers were inclined to regard every fjord as either a rift valley formed by the dropping down of a narrow strip of the earth's crust between two parallel faults, or as a gaping chasm opened along a single fault. This tectonic theory of the origin of fjords, once much in vogue as an explanation for all valleys, is now generally regarded as obsolete. Statements of the tectonic theory in which ice is credited with a very minor role in clearing out crushed and broken rock left in the fault cleft, or in the moderate widening of an open chasm, will be found in a short paper by Gurlt entitled “ Uber die Entstehungsweise der Fjorde,” published in 1874; in Peschel's “Neue Probleme der vergleichenden Erdkunde als Versuch einer Morphologie der Erdoberfläche,” dated four years later ; and in Kornerup's account of the fjords of south-west Greenland. A more modern supporter of the tectonic origin of fjords is Steffen in a shore paper on “ Der Baker-Fjord in West-Patagonien.” But by far the most elaborate thesis in support of the tectonic theory is J. W. Gregory's recent book on “ The Nature and Origin of Fjords.” This serious attempt to rehabilitate a much-discredited theory of fjord original contains extensive references to the literature of fjords, but frequently misinterprets the view held by the authors quoted. In a critical review of the book the present writer has endeavoured to show that Gregory's arguments are based upon a misconception of what the glacial theory of fjords implies, and upon an uncertain and variable interpretation of the tectonic theory.
Readers who wish to follow the discussion of the fjord problem further will be interested in an essay by Nordenskjold on “ Topographisch-geologische Studien in Fjordgebieten,” and in a shorter paper by Werth entitled “ Fjorde, Fjärde, and Föhrden.” Both contain many references to the literature of the subject, and Werth's paper explains the differences between typical fjords, the allied forms in low rocky coasts like south-western Sweden some-times called “ fiards ” (Plate XX), and the “ föhrden ” of the Baltic shores of Denmark and Schleswig-Holstein, similar to fiards but lacking their rocky shores. The relations of these three sub-types of fjords are also considered by Penck, Dinse, and Hubbard. An early paper by Ratzel discusses at some length the essential characteristics of fjords.
Without, at this time, entering into any elaborate discussion of the several theories of fjord formation, it may be said that the interpretation which would regard fjords as partially submerged river valleys fails to account on any rational basis for the restriction of true fjords to glaciated high latitudes, for the identity in form between fjord-valleys and the glacial troughs of glaciated high altitudes, for the almost uniform violation of Mayfair's law by tributary valleys which enter main fjords with discordant junctions, and for the occurrence of submerged fjord basins which, were the land to stand higher,
would become lake-basins not distinguishable from those of typical glacial troughs. Special pleading and strained reasoning have suggested a variety of possible explanations for each of these characteristic relationships, some of which might apply in one given instance, others in another. Glacial over-deepening of pre-existing river valleys alone offers a single explanation adequate to account at once for all of the specified relationships in all of the observed cases.
The tectonic theory of fjords is based on a misunderstanding of the significance of the known occurrence of fault-lines in certain fjords, and of the rectangular pattern of other fjords, which suggests an intersecting fault pattern. There can be little doubt but that crushed zones along faults, and infaulted strips of weak rock, have often determined the position and pattern of fjord-valleys. It is, however, an error of reasoning to jump to the conclusion that faults make fjords. As already noted, the glacial theory of fjord origin fully recognizes the fact that the pre-glacial valleys later transformed into fjords were often excavated along ancient fault-lines. Stream erosion naturally took advantage of the weak belts determined by faulting, forming fault-line valleys ; but not until ice occupied these pre-glacial stream valleys and profoundly changed their shape and their depth, were the forms which we called fjords produced. To prove the presence of a fault-line through a fjord is, therefore, to prove nothing as to the tectonic origin of the fjord. The tectonic theory, moreover, affords no rational explanation of the restriction of fjords to high latitudes, nor of the identity in form between fjord-valleys and unsubmerged glacial troughs, between fjord-basins and trough lake-basins. In the glacial theory alone do all of the phenomena cited, including the relation of fjords to faults, find a logical interpretation.
The foregoing paragraphs are sufficient to show that I regard the fjord problem as one of considerable importance, and that I have given much attention to it in its proper place. A more detailed analysis of Professor Gregory's ideas on fjords was published by me in 1915,* and is proof that I am not averse to discussing at length the theories which Professor Gregory supported some years ago, and which are now incorporated in his Memorandum and applied to the Labrador coast. But while such a discussion might be of interest to the physiographer, I do not see how it could be pertinent to the present issue. Certainly Newfoundland's claim to the Labrador coast will not hang on the slender thread of a special theory of fjord formation rejected at the present time by the great majority of physiographers. Surely the coast-line must be drawn, for geographical and territorial purposes, on the basis of fundamental physical distinctions widely recognized and accepted as valid, and not on the basis of theories which are still disputed by specialists. In law and in diplomacy, in the engineering profession and in the common walks of life, the essential phenomena of tidal rivers and sea inlets are well understood, and are made the basis of important acts ; but what place do fjords, and fiards, and theories of their origin hold in the practical affairs of man?
* JOHNSON, DOUGLAS. “The Nature and Origin of Fjords,” Science, N.S. XLI, 537-543, 1915.
(b) But even if we were to admit that the theories of fjord formation were pertinent to the present issue, and that the fine distinctions between fjords, fiards, föhrden, and rias of various types should be mastered by the practical man as well as by the professional physiographer, we would still face a curious dilemma. All these theories and distinctions relate to the origin of the basins occupied by the water, and not to the nature or behaviour of the water bodies as affected by the special forms of the basins.
Thus, suppose we accept Professor Gregory's theory that fjords are basins formed primarily by fracturing of the earth's crust. Identically similar fractures occur in the interior of the continents as well as on the margins. Every geological and physiographic feature assigned by Professor Gregory to Lake Melville can be duplicated repeatedly far inland in North America. The resulting basins hold fresh water in the interior and are therefore called lake basins. On the coast they may be entered freely by the sea, when they will be called fjords, fiards, föhrden, or rias by different students, according to the details of form exhibited, or the theory of origin accepted as most plausible. If the sea enters but partially and with difficulty by a narrow passage, the student must first decide whether the water body in the basin is more closely related to lakes or to sea arms, before he knows whether to call the whole feature a lake basin or a fjord. It is a curiously illogical procedure to say ; “ fractures produce fjords ; fractures produce the Lake Melville basin ; hence Lake Melville is a fjord ; therefore it is not a lake.” If one started with the even more valid premise : “ Fractures produce lake basins,” the same reasoning would lead to an opposite conclusion.
EVOLUTION OF NORTH-EASTERN AMERICA.
In order that there may be no misunderstanding of my own position, let me say specifically that in my opinion the fracture origin of the Labrador coast, as set forth by Professor Gregory in his Memorandum, is in the highest degree improbable. It is contrary to the known history of the whole eastern section of Canada as deciphered by a number of competent geologists working on the ground. All the geologic facts which Professor Gregory adduces in support of his theory are better accounted for in the well-established sequence of events worked out by Daly, Kindle, Goldthwait and many others. This sequence places the observed fractures far back in geologic time, and limits their effect upon present topography chiefly to an indirect influence upon erosion by providing infaulted strips of weak rock which rivers and glaciers could readily excavate. The basins were not formed by the sinking of the floors between fractures as assumed by Professor Gregory, but by normal stream erosion etching out the infaulted weaker rocks long after all topographic effects of the fracturing had been obliterated. For since the fracturing there have elapsed long periods of denudation, during which the land was reduced to a nearly level plane close to sea-level. Uplift of the plane permitted rivers to entrench themselves anew, and it was in this erosion period that the infaulted strips of weak rocks were quickly removed to give unusually broad
valley lowlands, while youthful valleys were carved in resistant rocks. (Professor Gregory unfortunately confuses the old and doubtful theory of “ Pre-Cambrian Valleys ” with the well-established and almost universally accepted present view that all the valleys date from Tertiary or later time ; and from this confusion draws wholly unwarranted conclusions as to the origin of the youthful valleys.) More recently, glaciers spread over the northern part of the country disturbing the drainage in the manner already described in an earlier paragraph. Where the ice passed through valley low-lands underlain by weak rocks, it found conditions especially favourable for easy excavation of deep basins which were occupied by water as the ice melted away. Infaulted weak rock strips were more abundant in the interior than on the coast, and their erosion by streams of water and ice produced countless broad river valleys and lake basins. The uplifted and dissected plane was gently warped, without any appreciable fracturing at this period, so that the margins of the continent were encroached upon by the sea. Valleys and lowlands were “ drowned ” giving sea inlets or bays where the sea entered freely and largely extinguished the land topography ; giving partially drowned tidal rivers or lakes where the former topography was but slightly affected. Of the true sea inlets those deeply cut and profoundly altered by glacial erosion to give certain specific forms are today called fjords ; those of less relief moderately affected by glaciation are called fiards ; and those which have suffered little or no changes due to glacial erosion are called rias. The same types of basins found inland give glacial troughs if ice erosion has profoundly affected the form of the original valley, glaciated valleys if the effects of ice erosion are moderate, and normal valleys if the ice produced little or no change. Wherever such inland valleys have over-deepened places in their courses lakes are formed.
LAKE MELVILLE NOT A FJORD.
Thus, whether the Lake Melville basin is a true lake basin, or is a fjord, fiard, or ria, depends wholly on whether its essential features are little affected by marine submergence, or are largely extinguished by the sea. In my opinion Lake Melville undoubtedly is a true lake. Even if its basin were deeply submerged by a free entrance of the sea, it could not properly be called a fjord, since according to all descriptions it clearly lacks the essential topographic features of a true fjord. It shows, for example, striking contrasts with the true fjords found farther north on the Labrador coast. Instead of trough walls descending steeply to the water, the lake is for long stretches bordered by lowland plains reaching inland from the shore for several miles before high land is encountered. The catenary cross-profile of the true fjord appears to be lacking, and the broad and irregular form of the basin is unusual for a fjord. Streams from the highlands do not descend directly to the lake from the ordinary type of hanging valleys but have their fall over a probable fault-line scarp far back from the supposed fjord waters. All this is normal for a lake basin developed by erosion of an infaulted block