26 1 A. Vaughan — Corrugation of the Earth's Surface. 



depths to which observation has been carried ; so that this property 

 may be roughly said to define what is meant by the crust. 



Let us now carefully consider what would be the efi"ect of allowing 

 a large heated mass to cool. We will first take the simplest case, 

 namely, a sphere of homogeneous material. Considering the mass 

 as made up of consecutive concentric shells, the outermost shell, on 

 account of its greater exposure and closer proximity to the cooling 

 influences, will cool faster than the next inner ; but, since cooling 

 necessarily produces contraction, and since the amount of contraction 

 is dependent on the loss of temperature, the external shell will 

 contract faster than that underlying it. It obviously follows that 

 the outer shell exerts a squeezing force upon the interior, and, by 

 compressing the mass into a smaller volume, increases its density. 

 The interior in its turn prevents the outer shell from contracting 

 to the full extent proper to its loss of temperature, and thus sets up 

 a state of strain in the outer shell. Thus, progi-essively, each shell, 

 by contracting upon the next inner one, increases the density of the 

 interior, and is itself at the same time thrown into a state of strain. 



A familiar illustration of this process is the brittleness induced in 

 most metals by rapid cooling, and to minimize this state of strain 

 recourse is had to the process of annealing. This reasoning seems 

 to apply with equal force to the case of the earth. Each shell, 

 as it contracts from loss of heat, compresses the included mass and 

 increases its density ; continued contraction of shell upon shell pro- 

 duces cumulative results in the same direction. 



From this very effect also the amount of conti-action will increase ; 

 for it is found that the coefiicient of expansion of any material 

 increases with its density. This, perhaps, affords an explanation 

 of the fact that, whilst the specific gravity of the surface rocks 

 averages about 2-5, that of the whole globe reaches 5'5. 



It may, perhaps, be diffidently suggested that in a molten mass 

 there would be a tendency to lessen the effect of differential con- 

 traction by a rearrangement of material in such a way that each 

 successively smaller shell should have a greater coefficient of ex- 

 pansion to recompense it for a smaller loss of temperature. 



Further, the effect of gravity upon such a molten mass would be 

 to cause the accumulation of the heavier material towards the centre. 



These two causes — the effect of gravitation on a molten mass 

 and the subsequent effects of contraction — must be looked to as 

 furnishing a probable cause of increased density in the interior. 



To the compressing force of contraction we must also look for 

 an explanation of the fact, established by astronomy, that the whole 

 globe is solid. For, judging by the increase of temperature towards 

 the centre, the rocks of the interior must be far above their melting 

 point, and can only be prevented from melting by subjection to 

 great pressure. 



It must be doubted whether superincumbent pressure can have 

 any great effect ; for, were the whole globe composed of concentric 

 homogeneous ellipsoidal shells, each such shell of gravitating matter 

 must be necessarily self-supportive, and could only exert pressure 

 upon the interior by undergoing contraction. 



