SUBOCEANIC STRUCTURE 295 



separation of basic and acid magmas, and to infer a coarse stratification 

 of earth matter into a basaltic shell under and within a granite shell.*^ 

 Bowen and Harker have recently discussed the settling down of heavier 

 crystals in a molt-cn magma, and the former has given us a clear analysis, 

 which amounts almost to demonstration, of the process of gravitative 

 separation; but many will agree with Harker that the requisite condi- 

 tions of a prolonged state of fluidity in masses of sufficient size have not 

 commonly existed in intrusive bodies such as are exposed to observation. 

 The process, if it existed, was a deep-seated one in general, and the mag- 

 nitude of the results which may be attributed to it is a subject of specu- 

 lation.^° 



It is not necessary in the present discussion, and certainly it is not 

 desirable in the absence of proof, to adopt any one of the major hy- 

 potheses in this field of speculation. Acid and basic magmas appear 

 intruded separately at the surface. The separation may have been origi- 

 nal or, by differentiation of a complex magma, subsequent to their inclu- 

 sion in the earth body. When the earth attained its present size they 

 may have been distributed in large or small bodies. They may have been 

 intimately interrelated in layers, lenses, and dikes, or they may have been 

 separated into basaltic and granitic shells. Any or all of these possibili- 

 ties may have been realities. From the point of view of isostasy, the 

 problem is: How could basic and acid rock-masses assume such relations 

 in the isostatic shell as to give rise to the ocean basins and continents? 



ORIGIN OF OCEAN BASINS 



In presenting a possible solution of this problem of the origin of the 

 ocean basins based on selective fusion, I rest on the following postulates 

 regarding the conditions under which rocks melt in the depths of the 

 lithosphere : 



1. Dry melting, or melting, in the absence of water, water vapor, or 

 other catalyzers, differs from melting in their presence, it being well 

 established that the catalyzers promote solution. 



2. Although the weight of evidence is in favor of the conclusion that 

 volcanic vapors are original constituents of the magmas, it is not unrea- 

 sonable to assume that there are. large masses of rock in the lithosphere 

 which are so nearly free from vapors that when heated they melt as dry 



*» R. A. Daly : Igneous rocks and their origin, 1914, pp. 159-164. 



.T. Barrel! : The evolution of the earth and its inhabitants, 1919, pp. 33 et seq. 

 ^ S. L. Bowen : The later stages of evolution of the igneous rocks. Chicago .Tour, of 

 Geol.. vol. xxiii, Supplement, 191.5. 



A. Harker : Differentiation in intercrustal magma basins. Chicago .Tour, of Geol., 

 vol. xxiv, 1916. 



XX— Bull. Geol, Soc. Am., Vol. 31, 1919 



