PETROGRAPHY, 347 



theory of their origin from two different subterranean localities, 

 but upon the assumption of their origin at different depths of 

 the crust. He held as a general principle that subterranean 

 magmas are distributed in the crust according to their specific 

 weight, the lighter magmas rich in silica occupying the crust- 

 cavities in the higher zones, the heavier basic magmas occurring 

 at lower horizons. Durocher, in 1857, gave a similar explana- 

 tion of the chemical and mineralogical differences in rock- 

 magmas. 



On the other hand, Poulett-Scrope (1825), Darwin (1844), 

 and Dana (1849) attributed the varieties of eruptive rocks to 

 the subsequent division and differentiation of a homogeneous 

 primitive magma. Justus Roth (1869) also regarded all 

 plutonic rocks as having been derived from a uniform 

 primitive magma, and explained their present differences of 

 constitution as a result of the different rates of cooling. 

 Iddings more recently remarked on the fundamental mineral- 

 ogical affinity of the different rock varieties in an eruptive 

 district, and compared such resemblances with the blood 

 relationships of organisms. Although most geologists at the 

 present day incline to the opinion that the different facies of 

 eruptive rock represent portions of a single primitive magma, 

 there is still great variance of opinion regarding the mode of 

 division and differentiation. 



The experiments of Spallanzani, Hall, and Bischof showed 

 that by means of regulating the process of cooling, or by 

 the application of different degrees of pressure, fused silicate 

 mixtures could be obtained in glassy, slaggy, or crystalline 

 rock-form. By Daubree's experiments it was ascertained that 

 the conditions requisite for the artificial reproduction of 

 granite-grained eruptive rocks were a moderate temperature 

 and the presence of water vapour. Again, the experiments of 

 Fouque and Levy seemed to show that the younger eruptive 

 flows with porphyritic structure had solidified slowly from an 

 igneous magma. It has proved a very complex and difficult 

 question to find out what determines the particular sequence 

 in which the rock-forming minerals separate from a viscous 

 magma. Fournet and Bunsen showed that the minerals by 

 no means separated from the magma in the order of their 

 fusing-points. After various attempts to solve the problem by 

 direct methods, it was then approached indirectly: keeping in 

 view the essential constituents of any particular rock, attempts 



