that, in a molten globe, the rocks which became solid should sink into the fiery menstruum; that permanent solidification must have accordingly commenced at the centre and ended at the surface, leaving some internal lakes or pockets of lava to keep up volcanic action. Now, such reservoirs of molten rock should have a small size in order to receive a roof in opposition to the laws of hydrostatics; and, if much of the original fluidity of their contents is still preserved, notwithstanding their immense losses of heat, especially during volcanic eruptions, we are compelled to make a very high estimate of the time required for the solidification of a fused globe 8,000 miles in diameter; and we must conclude that the time since the earth was covered by a permanent crust is but a very small part of that which has elapsed since terrestrial matter began its career in a gaseous or in a molten condition.
Though deprived by recent investigations of much of the support first claimed for it by Hopkins, the doctrine of the almost total solidity of the internal earth is still held by many; and, in the hands of Thomson and Tait, it is made to contribute to the evidences of the youth of our planet. They consider that, with such a solid and inflexible constitution, the terrestrial structure must have retained almost immutably, to the present day, the exact shape impressed on it in the beginning. As the figure which it now bears differs little from that of equilibrium, and as the polar compression is nearly the same as that which the present diurnal movement would occasion in a molten world, it has been concluded that, during geological history, the length of our day has changed little, though it would be increased by tidal friction one per cent, in the course of 20,000,000 years. From such considerations, the period since our earth assumed its terraqueous character has been estimated at not more than 10,000,000 years. Yet, in the vague use of this round number for marking the career of our globe, there is shown a wish rather to fix a limit to geological time than to adhere to strict mathematical precision. So little does the earth deviate from a figure of equilibrium, and so imperfect are the means for ascertaining the exact amount of the deviation, that it would be hazardous to say whether the age obtained by this course for the terrestrial crust is nearer to 1,000,000 or to 10,000,000 years.
Far more satisfactory would be the issue in dealing with a case like the hypothetical one which Prof. Tait introduces for illustration. Alluding to the ancient world he says: "Suppose, for instance, that it had not consolidated at less than 1,000,000,000 years ago. Calculation shows that at that time, at a moderate computation, it must have been rotating twice as fast as it now rotates; that is to say, the day must have been twelve hours instead of twenty-four. Now, if that had been the case, and the earth still fluid throughout or even pasty, the double rate of rotation would have produced four times as great centrifugal force; and the flattening of the earth's poles and the bulging out of the equator would have been much greater than we find them to be."