106 Arthur Holmes — Radio-activity 



sediments do not differ chemically in any essential respect from those 

 of later date. It might, however, be worth while to examine the 

 older sediments or their metamorphic equivalents for a possible nickel 

 content. Such rocks contain manganese, although in meteorites, 

 even in the stony varieties, nickel is always much more abundant 

 than manganese. 



The view that the external shell of the earth has not been in 

 a general molten condition depends on the assumption that radiation 

 from the surface was able to keep pace with the heat generated by 

 impact, together with the amount brought to the surface by con- 

 duction from the interior and the still greater amount carried by the 

 extrusion of molten tongues of the lighter and, with the aid of 

 volatile fluxes, more fusible materials. As Daly has pointed out, 1 it 

 is very doubtful whether this assumption can be justified. 



Another factor suggested by Dr. Evans which must be taken into 

 consideration is the thermal blanketing that would result as soon as 

 an atmosphere was generated. For the primitive atmosphere would 

 be exceedingly rich in carbon-dioxide and water vapour, athermanous 

 substances that would very efficiently absorb thermal radiations and 

 conserve the growing supply of heat within. It is impossible to 

 discuss quantitatively the relation between heat generation and 

 radiation in so complex a system, and it can only be said that the 

 assumption of a solid crust throughout the later stages of growth has 

 to overcome a number of difficulties which would not otherwise be so 

 serious. 



There can be no doubt that the mechanism by which granitic and 

 basaltic magmas were differentiated from average planetesimal 

 material and arranged according to their densities would work very 

 much more quickly and effectively if the earth were molten near the 

 surface during the whole period of growth, than if the temperature 

 of fusion were not reached until the freshly fallen planetesimals had 

 been buried to a considerable depth. There is, in fact, clear evidence 

 that fusion temperatures did in the early history of the earth come 

 very near the surface. The great abundance of Archaean granites 

 and orthogneisses throughout the continents proves a widespread 

 molten condition at no very great depth. It may be objected that 

 where the relations between the oldest granites and the oldest 

 sediments are indeterminable, the former are always intrusive. This 

 is, however, just what would be expected, for sedimentation implies 

 submergence, and the rocks constituting the underlying platform of 

 deposition must therefore have been lowered to depths where con- 

 siderably higher temperatures would be reached. If the temperature 

 gradient were high in these early days, the rise of temperature might 

 easily reach the point of refusion. Moreover, an additional effect of 

 sedimentation would be to cover the underlying platform with 

 a radio-active layer which, as Joly has shown, 2 would substantially 

 raise the basal temperature still further. Thus, granting a sufficiently 

 high temperature gradient in these early times, there is no difficulty 

 in understanding the irruptive contacts of the most ancient igneous 



1 Igneous Rocks and their Origin, p. 157, 1914. 



2 Radio-activity and Geology, p. 102, 1909. 



