136 Proceedings of the Royal Physical Society. 
traction to be at the rate of ^2^^ (lineal) for 180° Fahr., 
the lineal contraction will be y|^-o, and the consequent in- 
crease of density 151^ — 150^ = ¥9 V^f- Hence IMfl + 
2-0556 = (-04138 + 2-0556)== 2*09698 = the average density 
or specific gravity of the crust after refrigeration. But this 
is less than that of the immediately subjacent molten matter 
(2-1113, Par. 8), so that, if a portion of the crust should be 
completely detached by Assuring, it still would float upon 
the fluid mass. If the specific gravity of the crust increased 
on account of refrigeration in a higher ratio than the density 
in descending, the conservation of its stability would depend 
wholly upon its retention of the arched form, and if portions 
of it should by any means be separated, they would neces- 
sarily sink. In the relations, then, which these two ratios 
(provided our deductions are legitimate) bear to each other, 
we cannot but observe a most beautiful and striking example 
of design and providential arrangement. 
10. It appears sufficiently obvious that the upper surface of 
the molten interior must, to a certain considerable extent, 
be unequal, in accordance with the greater inequalities of the 
external surface of our globe ; and that the depth from, say, 
sea-level down to the internal fluid must vary considerably 
at different places. The solid crust must be nearly as thick 
below the bottom of the ocean (through several miles deep) 
as under the surface of dry land, whether we suppose the 
main oceans to have occupied nearly their present beds for 
an indefinite time, or what is more consistent with geo- 
logical facts, to have repeatedly changed them, heat being 
conveyed from the bottom of the sea to its surface by the 
process of convection with great rapidity, and the tempera- 
ture of the whole body of water, so far as the subterranean 
sources of heat are concerned, kept nearly the same. The 
figures of the isothermal planes in the interior cannot there- 
fore be those of concentric spheroidal surfaces, but must 
deviate therefrom to a certain extent, waving in a manner 
similar to, though differing in degree from, the undulations 
of the surface of the land and sea-bottom. 
11. The consequence of the contraction of the earth's 
radius from refrigeration, would (as has been already shown, 
