THE INTEEIOR OF THE EARTH. 231 



example, as might be appropriate if the lava were the residue of an 

 inherited original state and were merely differentiated by fractional 

 crystallization as it passed toward solidification. 



While these contrasted views of the history of magmas are nat- 

 urally connected with views of the genesis of the earth, they are not 

 limited to this connection. They are inherent in the very relations of 

 solid and liquid matter; they have a more or less important place 

 irrespective of the earth's genesis ; they would raise even keener ques- 

 tions than they do if the earth were supposed never to have had a 

 genesis, but to have always existed. 



An element of no small importance to a revised concept of the 

 interior of the earth has arisen from geodetic studies on the dis- 

 tribution of densities within the earth. As the geodetic point of view 

 is to be presented by its foremost exponent, Dr. Hayford, it is per- 

 missible for me merely to refer to certain geologic bearings. 



On the assumption that the earth was once in a molten state, the 

 inference is unavoidable that a perfect state of isostatic equilibrium 

 Avas originally assumed by the surface, and that its primitive con- 

 figuration was strictly spheroidal. The material must have been ar- 

 ranged in concentric layers according to specific gravity, and each 

 layer should have had the same density at every point. All such 

 reliefs of the earth's surface as have since arisen as well as all such 

 differences of specific gravity as now exist in the same horizon must 

 have been superinduced upon this originally perfect isostatic state. 

 With good reason therefore these inequalities have heretofore been 

 supposed to be relatively shallow. It is difficult to account for them, 

 then, even hypothetically. On the hypothesis that the earth grcAV up 

 by heterogeneous accretions, it is an equally natural inference that 

 differences of specific gravity extend to great depths. In an en- 

 deavor to find out the bearings of geodetic data on the distribution 

 of densities. Dr. Hayford tested four assumptions, all of which he 

 found measurably compatible w^ith his geodetic data. From these he 

 derived the respective compensation depths of 37, 76, 109, and 179 

 miles, these being the horizons to which differences of density ex- 

 tended and below Avhich they vanished or became negligible. Now 

 all these depths are notably greater than had been assigned as prob- 

 able depths of differentiation in the traditional molten earth. On the 

 other hand, the highest figure, 179 miles, was derived from a curve 

 drawn specifically to represent the probable distribution of densities 

 in an earth of planetesimal growth. The distribution represented by 

 this highest figure fits the geodetic data quite as well as either of the 

 other assumptions of distribution, though drawn on a strictly natu- 

 ralistic basis. If it could be said that geodetic data demonstrate that 

 the actual differentiation of specific gravities extends to dejiths of ^ 



