The corresponding subject of the sub-oceanic features off the eastern coast of North America has been dealt with in considerable detail by Professor Spencer and Mr. Warren Upham, who have shown that these features of the sea bottom on the western side of the North Atlantic Ocean are quite analogous, if not quite similar, to those worked out by Professor Hull on its eastern side.

After the papers which have appeared in the Journal of the Victoria Institute by Professor Hull, it is not necessary for me to recapitulate the facts that have been ascertained or the general conclusions derived from them. Sufficient here will be a reminder that the Continental Platform underlying the sea on the western coasts of Europe has been found by isobathic soundings to be indented at places approximately opposite to present river-valleys by deep ravines or submarine valleys, the bottoms of which descend from the present littoral to the deep sea bottom outside the Continental Platform, cutting through what has been called the "Great Declivity," by Professor Hull, and the "Sub-Oceanic Slope" by Mr. Hudleston, while at the same time these depressions in the sea-bed widen out seawards in a manner similar to those of existing river-valleys.

It has therefore been concluded that these submarine depressions are none other than former extensions or prolongations of the river-valleys of the present land surface, formed by sub-aerial erosion when the continental area was elevated sufficiently, and this in late Geological or early Pleistocene times.

Although this explanation, which has been very clearly and strongly advocated by Professors Hull and Spencer, may appear to many obvious, and the only possible explanation of the remarkable phenomena in question, it has been objected (1) that the required change of relative level of land and sea is too great to accord with known geological facts; and (2) that the submarine depressions are not proportionate in all cases to the size and importance of the rivers to the mouths of which they are respectively opposite, and by which they are assumed to have been formed.

With the first of these objections I do not at present propose to deal, since I now wish to call attention to the second difficulty only, namely, the disproportion at present existing between some of the submarine depressions and the neighbouring rivers. I will therefore confine myself to the consideration of the question whether the elevation of land

surfaces to the extent required for the sub-aerial erosion of depressions cutting through the Continental Platform to its base would anywhere give such a change of position of greater and lesser erosive power as would sufficiently account for the admitted disproportion of some submarine depressions to the respective nearest rivers by the prolongation of which it is concluded they have been cut down to their present depth.

The examples most prominently brought forward of this disproportion are those of the small depressions opposite to the great rivers, the Loire and the Gironde, and the great depression called La Fosse de Cap Breton opposite to the smaller river, the Adour.

The Fosse de Cap Breton is in the bed of the great southeast angle of the Bay of Biscay, the Gulf of Gascony. It commences close to the shore-line adjacent to Cap Breton, about ten miles north of the present mouth of the River Adour, on the coast of the Landes, and extends westwards as a depression in the sea-bottom for a distance of about 100 miles. At a distance of six miles from the land it has a depth of 1,000 feet from the surface of the sea, and at ten miles from the shore-line a depth of 1,200 feet. At fifteen miles from the commencement of the depression another submarine valley from the mouth of the Adour opens into it on the south side, and then the Fosse rapidly deepens, assuming, in the words of Professor Hull, "the form and features of a grand cañon, bounded by steep, sometimes precipitous, walls of rock from 4,000 to 6,000 feet in height, and ultimately opening out on to the floor of the ocean at a depth of about 1,500 fathoins (or 9,000 feet)."

The sea bottom on the north side of the Fosse de Cap Breton is remarkably different from that on its south side. On the north the Continental Platform, commencing with a width of thirty miles, widens as it extends northwards, until it attains a breadth of about 150 miles off the coast of Brittany, while on the south side it is very narrow, at one place only six miles wide, and nowhere along the entire length of the Fosse of 100 miles is the platform more than twenty miles in breadth. Thus the Fosse de Cap Breton is approximately parallel with the north coast of Spain and at right angles with the southern part of the French shore of the Bay of Biscay, the coast of the Landes.

The adjacent coasts correspond in their physical features most strikingly with the sea floor on each side of the Fosse.

As is well known, the Landes is an extensive plain very little above the sea-level, spreading inland for a distance of 100 miles, and extending along the Bay of Biscay from the Adour to the Gironde, a distance of about 150 miles. The coast of Spain on the other hand is mountainous, the western extension of the Pyrenees, the Cantabrian mountains, fringing the shore-line, with their spurs forming head lands.

Corresponding, too, with the physical features of the two coasts is, as might be expected, their geological structure. The flat French coast area is formed of Quaternary accumulations overlying Miocene and Eocene strata, while the rocky, hilly, and mountainous Spanish coast lands are composed of strata of Secondary and Palæozoic Age.

Of the rivers flowing into the Gulf of Gascony, or southeast angle of the Bay of Biscay, the principal is the River Adour, which, although called by Dr. Blanford a comparatively trivial stream, is better, I think, described by Professor Hull as "a fine river." It has a breadth at the city of Bayonne of 800 feet, and from its source in the Pyrenees to its mouth measures at least 200 miles, and has several important tributaries, as the Nive, which joins it at Bayonne, the Oloron, the Gave de Pau, and on the north, the Midouze, which give the Adour a large drainage area extending for fully 120 miles along the northern side of the Pyrenees. About fifteen miles to the south of the mouth of the Adour, the River Nevelle flows into the sea at St. Jean de Luz. This at present is an unimportant river, but a little farther to the south and joining the sea at the very angle of the Bay of Biscay there is a river that deserves more attention than it generally receives. This is the River Bidassoa, the boundary river between France and Spain, and differs from the Adour and the Nevelle in draining the south side of the western part of the Pyrenees. It is even now a considerable stream, and the alluvial flats seen as Fuentarabia, at its mouth, is approached shows that in quite late Quaternary times it was a much greater river. These three rivers, the Adour, the Nevelle, and the Bidassoa, now pour their waters into the sea at the head of the Fosse de Cap Breton.

We may now inquire whether, under the conditions that must have obtained with an elevation of the contiguous lands to 9,000 feet above their present levels, the output of ice, and afterwards of water, together with the action of the sea at the bight of the Bay of Biscay, would not be of sufficient erosive power to produce the Fosse de Cap Breton,

while at the same time there was no such enormous erosive power possessed by either of the then existing rivers, the Loire or the Gironde.

With an added elevation of 9,000 or even of 7,000 feet not only the higher and mountainous districts of France and Spain, and especially the Pyrenean region, but the whole of the area so raised would be brought under glacial conditions. The Pic de Nethou in Maladetta, Mont Perdu, and a few other summits in the Pyrenees about 10,000 feet in elevation, have glaciers at the present time, and there would be many more glaciers on the Pyrenean mountains were it not for the fact that their summits are not favourably grouped for the accumulation of great glacier-producing masses of snow. Under the conditions supposed, the whole range of the Pyrenees, including all its spurs and offshoots, would reach far above the snow-line. This mountain region is of considerable breadth, for the main sierra of the Pyrenees is buttressed, as it were, on each side by mountains for a distance of from fifteen to twenty miles from the axis of the range, giving a breadth, therefore, of from thirty to forty miles of mountains extending in length from the Bay of Biscay to the Mediterranean Sea.

Although the summits of these mountains are not well adapted for the accumulation of snow and the formation of glaciers, this is not the case with those parts of the region now well below the snow-line, since there are innumerable valleys of great capacity amidst surrounding lofty hills and mountains, which if above the snow-line would retain snow and so accumulate sufficient material to produce very many and very large glaciers. Winds from the west and southwest, that is from the ocean, would prevail as now and bring with them enormous amounts of vapour and air charged with evaporated water from the warm surface waters of the equatorial seas and the scarcely less warm waters of the deflected equatorial current flowing north-eastward as now from the American continent, but, in consequence of the elevation of the West Indian, or the Antillean region, having a less northerly course. This vapour and water-gas would be rapidly condensed by the great cold of the mountains and plateaux, and the mountains being of great elevation, the atmosphere would be compelled to give up a very large proportion of its water. There would, consequently, be over the whole of the European south-west region an unusually great precipitation, and thus the material for the

production of vast glaciers and ice-sheets would here be furnished in profuse abundance.

It ought also to be borne in mind that with the elevation of the American or western continent also, or of its central portion, the Antillean region, there would be no so-called "Gulf Stream" as now to bring warmth-giving waters and consequent warmth-giving winds to the north of Europe. This deprivation would intensify the glacial conditions consequent upon elevation in the areas to the north, and the great cold so produced there would react on the temperature of mid and southern Europe. Thus an additional refrigerating influence must be taken into account. The result of the whole would be the covering of the Pyrenean region with an ice-cap or continuous glacier of great thickness through which only the more acute summits would penetrate. This vast body of ice, gradually descending to lower levels, as the Greenland ice-cap does at present, would form an ice-sheet covering all the lower levels. We are thus compelled to conclude that with an elevation of 9,000 feet, or even of 7,000 feet, the entire region of what is now southern France and northern Spain would have a climate of quite Arctic cold, and would be covered by a vast and continuous capping of ice.*

Glacial conditions, with the elevation postulated, must have extended over the whole of France, but they would be greatly intensified in the Pyrenean region and especially in its western half, from its elevation, and the much larger body of vapour-charged air that would be there intercepted and the consequent enormous amount of snow that would there be precipitated.

The great ice-sheet spreading to the north of the Pyrenean range would move in the direction of least resistance. When the surface features of southern France are considered it is at once seen that the direction of least resistance to an advancing ice-sheet from the northern side of the western half of the Pyrenees would be to the west, or in the direction of the sea, over the region now occupied by the drainage areas of the Adour and the Garonne, which includes the extensive low plain of the Landes, at present little above the level of the sea, to which it extends. The whole of

* On this subject see paper by Professor E. Hull on "Another possible cause of the Glacial Epoch." Trans. Vict. Inst., vol. xxxi, with plate, p. 141 (1897-8).