founded on incorrect data. For, surely, no one will suggest that so vast a problem of terrestrial physics can be held to be solved till we have exhausted all the evidence at our command, and have shown that it largely preponderates on one side or the other.

The present article is intended to supply a hitherto unnoticed class of arguments for oceanic permanence, and these must of course be taken in connection with the other evidence which has been summarised in the sixth chapter of the writer's Island Life, and also with the admirable "Summary" of the purely physical argument in the second edition of Mr. Ô. Fisher's Physics of the Earth's Crust. It is certainly a remarkable fact that writers approaching the subject from so many distinct points of view-as have Professor Dana, Mr. Darwin, Sir Archibald Geikie, Dr. John Murray, Rev. O. Fisher, and myself should yet arrive at what is substantially an identical conclusion; and this must certainly be held to afford a strong presumption that that conclusion is a correct one.

complex one, and the evidence of so varied a nature and often so difficult to interpret, that the conclusions reached have been usually doubtful and often conflicting. This seems to have been due, in part, to the fact that no properly qualified person had, till quite recently, devoted himself to a thorough study of the whole subject, taking full account of all the materials available for arriving at a definite conclusion, as well as of the various groups of phenomena which such a conclusion must harmonize and explain. But for many years past a good practical geologist, who is also an advanced mathematician-the Rev. Osmond Fisher-has made this subject his speciality, and in a most interesting volume, of which a second and carefully revised edition, with an appendix, has been recently published, he has brought together all the facts bearing on the problem, and has arrived at certain definite conclusions of the greatest interest. The object of this chapter is to give a popular account of so much of his work as bears upon the question of the thickness and density of the earth's crust and the constitution of the interior.1

The Argument from Internal Heat.

We will first consider the nature of the evidence in favour of the view that, below a superficial crust, there is a molten or highly heated substratum. The existence of volcanoes, geysers, and hot springs irregularly scattered over the whole surface of the globe, and continually ejecting molten rock, ashes, mud, steam or hot water, is an obvious indication of some very widespread source of heat within the earth, but of the nature or origin of that heat they give little positive information. The heat thus indicated has been supposed to be due to many causes, such as the pressure and friction caused by contraction of the cooling crust, chemical action at great depths beneath the surface, isolated lakes of molten rock due to these or to unknown causes, or to a molten interior, or at least a

1 Physics of the Earth's Crust, by the Rev. Osmond Fisher, M.A., F.G.S. Second edition, altered and enlarged. Macmillan and Co., 1889. With an Appendix, 1891.

general substratum of molten matter between the crust and a possibly solid interior. The first two causes are now generally admitted to be inadequate, and our choice is practically limited to one of the latter.

There are also very important evidences of internal heat derived from the universal phenomenon of a fairly uniform increase of temperature in all deep wells, mines, borings, or tunnels. This increase has been usually reckoned as 1° F. for each 60 feet of descent, but a recent very careful estimate, by Professor Prestwich, derived from the whole of the available data, gives 1° F. for every 475 feet of descent. It is a curious indication of the universality of this increase that, even in the coldest parts of Siberia, where the soil is frozen to a depth of 620 feet, there is a steady increase in the temperature of this frozen soil from the surface downwards. Much has been made by some writers of the local differences of the rate of increase, varying from 1° in 28 feet to 1o in 95; and also of the fact that in some places the rate of increase diminishes as the depth becomes greater. But when we consider that springs often bring up heated water to the surface in countries far removed from any seat of volcanic action, and the extent to which water permeates the rocks at all depths reached by man, such divergences are exactly what we might expect. Now this average rate of increase, if continued downwards, would imply a temperature capable of melting rock at about twenty miles deep, or less, and we shall see presently that there are other considerations which lead to the conclusion that this is not far from the average thickness of the solid crust.

Before going further it will be well to consider certain objections to this conclusion, which for a long time were considered insuperable, but which have now been shown to be either altogether erroneous or quite inconclusive. In Sir Charles Lyell's Principles of Geology, Mr. Hopkins is quoted as having shown that the phenomenon of the precession of the equinoxes, due to the attraction of the sun and moon on the equatorial protuberance, requires the

1 In a recent deep boring at Wheeling, Virginia, the rate of increase was found to be greater as the depth increased.

interior of the earth to be solid, or at least to have a crust not much less than one thousand miles thick. This view was supported by Sir William Thomson and other eminent mathematicians, and so great was the faith of geologists in these calculations that for nearly forty years the theory of the earth's internal liquidity was almost wholly abandoned. But this argument has now been shown to be erroneous by the more complete investigations of Professor George Darwin, while Sir William Thomson (now Lord Kelvin) has recently shown experimentally that a rotating liquid spheroid behaves under stresses as if it were a solid. Another difficulty arises from the phenomena of the tides. It has been argued that, if the interior of the earth is liquid tides will be formed in it which will deform the crust itself, and thus, by lifting the water up with the land, do away with any sensible tides in the ocean. But Mr. Fisher has pointed out that this conclusion rests on the assumption that the liquid interior, if it exists, is not an expansible fluid; and he shows that if this assumption is incorrect it is quite possible that little or no deformation would be caused in the crust by tides produced in the liquid interior; and he further maintains, as we shall see presently, that all the evidence goes to prove that it is expansible. Moreover, in a late paper, he claims to have proved that even the deformation of the crust itself would not obliterate the ocean tides, but would diminish them only to the extent of about one-fifth.1

There remain the geological objections founded on the behaviour of volcanoes, which is supposed to be inconsistent with a liquid interior as their effective cause. We have, for instance, the phenomenon of a lofty volcano like Etna pouring out lava from near its summit, while the much lower volcanoes of Vesuvius and Stromboli show no corresponding increase of activity; and the still more extraordinary case of Kilauea, on the lower slopes of Mauna Loa, in the Sandwich Islands, at a height of about 3,800 feet, whose lake of perennial liquid lava suffers no alteration of level or any increased activity when the parent mountain is pouring forth lava from a height of 14,000

1 Proceedings of the Cambridge Philosophical Society, 1892.

feet. Again, it is argued that if the igneous products of volcanoes are derived from one central reservoir there ought to be a great similarity between them, especially between those of the same district. But this is not the case, an example being the Miocene lavas of Hungary and Bohemia, which are of a totally different character, chemically, from each other. But although the molten interior of the globe may be the common source of the heat which causes volcanic eruptions, it by no means follows that the whole, or any large portion, of the matters ejected from volcanoes are derived from it; and it is a remarkable indication of the probable truth of Mr. Fisher's theory, that, as will be shown further on, it entirely removes the two geological difficulties here noticed. At the same time it explains other geological phenomena of a striking character which the theory of solidity altogether fails to account for, as will be now briefly indicated.

The Argument from Subsidence under Deposition of

Sediment.

It has long been known to geologists that the series of sedimentary rocks, ancient as well as modern, afford repeated examples of great piles of strata hundreds, or even thousands of feet thick, which throughout present indications of having been formed in shallow water, and which therefore imply that as fast as one bed was deposited it sank down, and was ready to receive another bed on the top of it. As an example we may refer to the Paleozoic rocks of the Alleghany Mountains, which are not less than 42,000 feet thick; yet the lowest of these strata, the Potsdam sandstone, was not deposited in a deep sea, but evidently in shallow water near shore, several of the beds exhibiting distinct ripple markings, and the same is the case with the highest strata found there-the carboniferous. On this point Sir Archibald Geikie remarks:

"Among the thickest masses of sedimentary rocks-those of the ancient Palæozoic systems-no features recur more continually than the alternation of different sediments, and the recurrence of surfaces covered with well-preserved ripple-marks, trails and burrows of