THE EARTH'S INTERIOR— ADAMS 267 



tory the volume-change, being measured directly, is independent of 

 the degree of isotropy of the material. In view of the fact that the two 

 methods agree remarkably well for nonporous substances such as siHca 

 glass and also for the porous rock, limestone, the writer prefers to accept 

 the results of direct volume-change measurement and the conclusions 

 based thereon, at least until measurements by the linear method are 

 made on specimens of diabase cut in three mutually perpendicular 

 directions, and until further measurements by both methods on the 

 same samples of rock become available." 



From observations on the tidal deformation of the surface, the earth 

 has long been laiown to be as rigid as steel, and from seismologic 

 data we find that from crust to core the rigidity increases steadily 

 with increasing depth. It is very clear, therefore, that the earth as a 

 whole is "soUd"; but whether its substance, particularly in the inter- 

 mediate zone, is crystalhne or glassy is more difficult to decide. 

 From what information we have it does not seem possible that any 

 silicate glass of a composition favorable for remaining permanently 

 in the glassy state can support the requisite wave velocities. Although 

 there may be, and probably are, shallow zones or limited regions of 

 glassy material, the weight of evidence, in the writer's opinion, seems 

 to favor crystallinity for practically the entire silicate part of the 

 earth. This is a subject for which admittedly there are decided 

 differences in the viewpoints of various investigators. Lack of space 

 prevents a complete discussion of the subject at this time. 



Many circumstances, concerning which there is at the moment 

 insufficient time for discussion, indicate that the early history of the 

 earth was as follows: The primitive molten magma consisting mainly 

 of magnesium iron silicates with smaller amounts of other oxides, 

 including water, together mth a considerable amount of metallic iron, 

 first separated into two layers — molten iron below and silicate magma 

 above. The silicate layer then began to crystallize at the bottom. 

 As the solid layer increased in thiclaiess, the minor constituents, in- 

 cluding water, were concentrated to a greater and greater extent in 

 the remaining liquid. Finally, when the liquid layer had been reduced 

 to a thickness of a few tens of kilometers, and was much richer in the 

 originally nunor constituents, the crust of the earth was formed. 



One of the most cogent reasons for believing that the earth is 

 crystalline is that m no other way can we easily account for the fact 

 that the crust differs so markedly from the interior. Granting that 

 the earth was once molten and well stirred, we apparently must admit 

 that the separation into zones on so large a scale took place either by 

 the falling of a heavy insoluble liquid to the bottom (thus producing 

 the iron core) or by the residuum of a process of crystallization, this 

 residuum becoming the crust. 



» Footnote added in proof. Recent measurements by Birch and Bancroft (Journ. Geol., vol. 46, p. 126, 

 1038) confirm the compressibility results obtained at the Geophysical Laboratory. 



