recent, and not the ancient, sea-floors would be the deeper. For that reason, in my opinion, the differences of depth are also to be explained by the temperature relationships. The old ocean-floors became more strongly cooled, and are therefore of greater density than the younger. If the specific gravity, say, of sima amount to 2.9, then it would change to 2.892 by an elevation of temperature of 100° C. on the basis of the cubical coefficient of expansion for granite of 0.000 026 9. Two ocean-floors with a difference of temperature of 100° C., which are in isostatic equilibrium with each other, must then show a difference in depth of 300 m., in which the warmer floor is at a higher level. It is certainly difficult to imagine, for example, that the floor of the Atlantic should have preserved its higher temperature throughout a period of time estimated at millions of years, even if the original difference of temperature may have been very much higher (1000° to 1500° C.). Still, we do not know from what source the internal heat of the earth has been derived. If it is produced by the disintegration of radium, or only partly maintained in this manner, the idea that freshly exposed deep-seated layers could show, even throughout geological ages, an increased temperature on account of the greater content of radium, is not to be entirely rejected off-hand.
If the sima is actually a viscous, fluid body, comparable with sealing-wax, it would be remarkable if its ability to flow only manifested itself in yielding in front of the drifting sial block, and if streams of independent character did not occur also. The maps give in places a direct appearance of more such local flows of sima by the distortion of chains of islands which were, apparently, formerly in straight lines. Two examples of this are shown in Fig. 25, namely, that of the Seychelles and that of the Fiji Islands. The
