nal des Savants.' He always enjoys good health, and 'eats more than I do,' says M. Chevreul, fils." His temperance grew out of a repugnance which he contracted in youth to wines and liquors, and extends to smoking.

His favorite topic is colors, respecting which, our correspondent says, "he would insist on sitting up in bed and giving a demonstration on the propagation of colors. His strong point was that the 'colors. are in us, and the cause in the things we look at' (de hors). Although he had talked a great deal during the day, there was no stopping him when once the started on the color question, or getting him to change the subject; and when we rose to leave, he protested that we were going away because his exposition wearied us.” He is as earnest and enthusiastic a student yet as if he had another hundred years before him. "No man, perhaps, has seen his country pass through so many revolutions, and has lived under so many régimes as M. Chevreul. He remembers Louis XVI. His recollections of the Revolution and the Directoire are clear, though he was not then at Paris. He can call up pictures of the glory and the dignity of the First Empire. He has lived under the First Restoration, the Hundred Days, the Restoration of 1815, the Legitimist rule of 1830, the Republic of 1848, the Second Empire of 1852, and the Third Republic-in all eleven régimes, which is tolerably good for one lifetime."

The lesson has been drawn from M. Chevreul's life of what one writer styles "the physical wholesomeness of sustained labor." Cases of extreme longevity are usually found either among persons who live in almost complete inactivity of mind and are thus subject to no wear whatever from their nervous and intellectual faculties, or else among those who spend their lives in constant, vigorous thought. Persons of the class between these, who learn and pursue some business which in time becomes largely a matter of routine and ceases to call out exertion of the powers, usually die early, or at a moderate old age. Hence, the wonderful brightness and activity which we sometimes admire among very old persons, is not so wonderful after all, but is a part of their old age, and one of the causes that have enabled them to enjoy it. And the general rule is sustained, in the case of M. Chevreul, as in the case of numerous other men who have served the world or are serving it at ages far beyond threescore and ten, that "the harmonious development of all the many-sided aspects of man is the most conducive to the health of the individual, and that the training of the brain may be as valuable as the training of the muscles.”

THE

GEOLOGY OF THE ATLANTIC OCEAN.*

BY SIR WILLIAM DAWSON,

PRINCIPAL OF MCGILL COLLEGE, MONTREAL.

I.

HE geological history of the Atlantic depression of the earth's crust, and its relation to the continental masses which limit it, may furnish a theme at once generally intelligible and connected with great questions as to the structure and history of the earth, which have excited the attention alike of physicists, geologists, biologists, geographers, and ethnologists. If we imagine an observer contemplating the earth from a convenient distance in space, and scrutinizing its features as it rolls before him, we may suppose him to be struck with the fact that eleven sixteenths of its surface are covered with water, and that the land is so unequally distributed that from one point of view he would see a hemisphere almost exclusively oceanic, while nearly the whole of the dry land is gathered in the opposite hemisphere. He might observe that the great oceanic area of the Pacific and Antarctic Oceans is dotted with islands-like a shallow pool with stones rising above its surface-as if its general depth were small in comparison with its area. He might also notice that a mass or belt of land surrounds each pole, and that the northern ring sends off to the southward three vast tongues of land and of mountain-chains, terminating respectively in South America, South Africa, and Australia, toward which feebler and insular processes are given off by the Antarctic continental mass. This, as some geographers have observed, gives a rudely threeribbed aspect to the earth, though two of the three ribs are crowded together and form the Europ-Asian mass or double continent, while the third is isolated in the single Continent of America. He might also observe that the northern girdle is cut across, so that the Atlantic opens by a wide space into the Arctic Sea, while the Pacific is contracted toward the north, but confluent with the Antarctic Ocean. The Atlantic is also relatively deeper and less cumbered with islands than the Pacific, which has the higher ridges near its shores, constituting what some visitors to the Pacific coast of America have not inaptly called the "back of the world," while the wider slopes face the narrower ocean, into which for this reason the greater part of the drainage of the land is poured. The Pacific and Atlantic, though both depressions or flattenings of the earth, are, as we shall find, different in age, character, and conditions; and the Atlantic, though the smaller, is the older, and from the geological point of view, in some respects, the more important of the two. If our imaginary observer had the means of knowing anything of the rock formations of the continents,

* From the inaugural address of the President of the British Association for the Advancement of Science, delivered at Birmingham, England, September 1, 1886.

he would notice that those bounding the North Atlantic are in general of great age, some belonging to the Laurentian system. On the other hand, he would see that many of the mountain-ranges along the Pacific are comparatively new, and that modern igneous action occurs in connection with them. Thus he might be led to believe that the Atlantic, though comparatively narrow, is an older feature of the earth's surface, while the Pacific belongs to more modern times. But he would note in connection with this that the oldest rocks of the great continental masses are mostly toward their northern ends, and that the borders of the northern ring of land and certain ridges extending southward from it constitute the most ancient and permanent elevations of the earth's crust, though now greatly surpassed by mountains of more recent age nearer the equator.

Before leaving this general survey we may make one further remark. An observer looking at the earth from without would notice that the margins of the Atlantic and the main lines of direction of its mountain-chains are northeast and southwest, and northwest and southeast, as if some early causes had determined the occurrence of elevations along great circles of the earth's surface tangent to the polar circles. We are invited by the preceding general glance at the surface of the earth to ask certain questions respecting the Atlantic: 1. What has at first determined its position and form? 2. What changes has it experienced in the lapse of geological time? 3. What relations have these changes borne to the development of life on the land and in the water? 4. What is its probable future? Before attempting to answer these questions, which I shall not take up formally in succession, but rather in connection with each other, it is necessary to state as briefly as possible certain general conclusions respecting the interior of the earth. It is popularly supposed that we know nothing of this beyond a superficial crust perhaps averaging fifty thousand to one hundred thousand feet in thickness. It is true we have no means of exploration in the earth's interior, but the conjoined labors of physicists and geologists have now proceeded sufficiently far to throw much inferential light on the subject, and to enable us to make some general affirmations with certainty; and these it is the more necessary to state distinctly, since they are often treated as mere subjects of speculation and fruitless discussion:

1. Since the dawn of geological science, it has been evident that the crust on which we live must be supported on a plastic or partially liquid mass of heated rock, approximately uniform in quality under the whole of its area. This is a legitimate conclusion from the wide distribution of volcanic phenomena, and from the fact that the ejections of volcanoes, while locally of various kinds, are similar in every part of the world. It led to the old idea of a fluid interior of the earth, but this is now generally abandoned, and this interior heated and plastic layer is regarded as merely an under-crust.

2. We have reason to believe, as the result of astronomical investigations, that, notwithstanding the plasticity or liquidity of the undercrust, the mass of the earth-its nucleus, as we may call it-is practically solid, and of great density and hardness. Thus we have the apparent paradox of a solid yet fluid earth; solid in its astronomical relations, liquid or plastic for the purposes of volcanic action and superficial movements.

3. The plastic sub-crust is not in a state of dry, igneous fusion, but in that condition of aqueo-igneous or hydro-thermic fusion which arises from the action of heat on moist substances, and which may either be regarded as a fusion or as a species of solution at a very high temperature. This we learn from the phenomena of volcanic action, and from the composition of the volcanic and plutonic rocks, as well as from such chemical experiments as those of Daubrée and of Tilden and Shenstone.

4. The interior sub-crust is not perfectly homogeneous, but may be roughly divided into two layers or magmas, as they have been called -an upper, highly siliceous or acidic, of low specific gravity and lightcolored, and corresponding to such kinds of plutonic and volcanic rocks as granite and trachyte; and a lower, less siliceous or more basic, more dense, and more highly charged with iron, and corresponding to such igneous rocks as the dolerites, basalts, and kindred lavas. It is interesting here to note that this conclusion, elaborated by Durocher and Von Waltershausen, and usually connected with their names, appears to have been first announced by John Phillips in his "Geological Manual," and as a mere common-sense deduction from the observed phenomena of volcanic action and the probable results of the gradual cooling of the earth. It receives striking confirmation from the observed succession of acidic and basic volcanic rocks of all geological periods and in all localities. It would even seem, from recent spectroscopic investigations of Lockyer, that there is evidence of a similar succession of magmas in the heavenly bodies, and the discovery by Nordenskiöld of native iron in Greenland basalts affords a probability that the inner magma is in part metallic.

5. Where rents or fissures form in the upper crust, the material of the lower crust is forced upward by the pressure of the less supported portions of the former, giving rise to volcanic phenomena either of an explosive or quiet character, as may be determined by contact with.

The underlying material may also be carried to the surface. by the agency of heated water, producing those quiet discharges which Hunt has named crenitic. It is to be observed here that explosive volcanic phenomena and the formation of cones are, as Prestwich has well remarked, characteristic of an old and thickened crust; quiet ejection from fissures and hydrothermal action may have been more common in earlier periods, and with a thinner over-crust.

6. The contraction of the earth's interior by cooling and by the

emission of material from below the over-crust has caused this crust to press downward, and therefore laterally, and so to effect great bends, folds, and plications; and these, modified subsequently by surface denudation, constitute mountain-chains and continental plateaus. As Hall long ago pointed out, such lines of folding have been produced more especially where thick sediments had been laid down on the sea-bottom. Thus we have here another apparent paradoxnamely, that the elevations of the earth's crust occur in the places where the greatest burden of detritus has been laid down upon it, and where, consequently, the crust has been softened and depressed. We must beware, in this connection, of exaggerated notions of the extent of contraction and of crumpling required to form mountains. Bonney has well shown, in lectures delivered at the London Institution, that an amount of contraction almost inappreciable in comparison with the diameter of the earth would be sufficient; and that, as the greatest mountain-chains are less than one six-hundredth of the earth's radius in height, they would, on an artificial globe a foot in diameter, be no more important than the slight inequalities that might result from the paper gores overlapping each other at the edges.

7. The crushing and sliding of the over-crust implied in these movements raise some serious questions of a physical character. One of these relates to the rapidity or slowness of such movements, and the consequent degree of intensity of the heat developed, as a possible cause of metamorphism of rocks. Another has reference to the possibility of changes in the equilibrium of the earth itself as resulting from local collapse and ridging. These questions in connection with the present dissociation of the axis of rotation from the magnetic poles, and with changes of climate, have attracted some attention, and probably deserve further consideration on the part of physicists.

In so far as geological evidence is concerned, it would seem that the general association of crumbling with metamorphism indicates a certain rapidity in the process of mountain-making, and consequent development of heat, and the arrangement of the older rocks around the Arctic basin forbids us from assuming any extensive movement of the axis of rotation, though it does not exclude changes to a limited extent. I hope that Professor Darwin will discuss these points in his address to the Physical Section. I wish to formulate these principles as distinctly as possible, and as the result of all the long series of observations, calculations, and discussions since the time of Werner and Hutton, and in which a vast number of able physicists and naturalists have borne a part, because they may be considered as certain deductions from our actual knowledge, and because they lie at the foundation of a rational physical geology.

Keeping in view these general conclusions, let us now turn to their bearing on the origin and history of the North Atlantic. Though the Atlantic is a deep ocean, its basin does not constitute so