160 
ME. EOBEET MALLET ON VOLCANIC ENEEGY. 
great currents of circulation must have arisen within the spheroid, by reason of the 
more rapid cooling at the poles than in equatorial regions, warmer currents proceeding 
superficially from the equator towards the poles, and colder currents returning about 
the axis of rotation, and dividing and ascending about the plane of the equator. 
But the effect of these in equalizing the temperature of the superficial and deep 
parts of the spheroid would be small; it would, notwithstanding their influence, be 
always greatly hotter at the centre than at any part of the surface or near it. The 
exterior couches becoming more and more viscous as refrigeration proceeded, a thin 
solid crust would at length be formed over the whole spheroid, thickest at the poles. 
This crust, while thin , though solid through a certain thickness, would be very 
nearly at the same high temperature as the viscous and the fluid matter beneath it. 
It would therefore still continue to part with heat by radiation at a rate not very far 
different from what the surface of the liquid spheroid did prior to formation of any crust. 
49. But it is certain (as wall be hereafter shown) that the coefficient of contraction 
per degree is less, for such materials as rocks are formed of, when in the solid than in the 
viscous or liquid state, and decreases as a function of the temperature down to the 
mean temperature of our present atmosphere. 
It would depend upon the rate of surface-cooling, compared with that of the viscous 
matter below, and upon conditions of radiation &c., for which we have no sufficient 
data, whether the relations between the rate of cooling and of the contraction of the 
exterior and interior might not give rise to tensions within the thin crust ; but upon 
the whole the effect would appear to be to produce tangential pressures also within 
the thin crust. To the compressions producible by these the thin crust would readily 
give way by wrinkling or folding over, or equally give way to tension by breaking to 
pieces. The density of the solidified matter in this thin crust was no doubt greater than 
that of the viscous layer beneath it, and the density of the last greater than that of the 
liquid matter still deeper ; but the difference in density must have been small (as we 
shall hereafter see), the difference of temperature between such layers being so. Nor can 
we attribute any great difference in density to any abrupt molecular change in volume 
known to accompany the passage from the viscous to the solid state in the materials of 
rocks. [The differences in density observed in the passage of mineral compounds from the 
vitreous to the crystalline state, and vice versa, are not great, and in those cases the 
cooling has been comparatively sudden. In the production of Eeaumur’s porcelain, 
the change by slow heating is preceded by abrupt cooling as glass. In our crust there 
was not abrupt cooling.] 
50. The thin crust, even if more or less dislocated, would therefore probably still remain 
upon the surface, sustained by the viscous bed beneath and that by the liquid spheroid, 
and would not piecemeal sink down into the latter. 
But should we assume it to do so, there appears the utmost improbability in the 
supposition, which originated with Poisson, and seems to be adopted by Thomson, 
that the solidified sheets broken from such thin crust and constantly renewed would 
