﻿,"301 THE NATURAL CLASSIFICATION OF IGNEOUS HOCKS. [Allg. I9IO, 



which on general mineralogical grounds would be ranged in Perfe- 

 mane, appears in Washington's Tables in Salfemane. 



Dr. J. W. Evans believed that a natural classification of igneous 

 rocks must be founded on genetic principles, but admitted that 

 our knowledge of the processes involved in the differentiation of 

 igneous magmas was still too incomplete to enable . us to frame a 

 satisfactory permanent system of classification on these lines. It 

 might, he thought, be possible to demonstrate the existence of rock- 

 types which occurred with greater frequency than the intervening 

 links, and might serve as the centres of natural groups that would 

 prove to have a genetic significance ; but the necessary detailed 

 information as to the composition and extent of rock-masses was as 

 yet insufficient for the purpose. In the meantime, what might be 

 called an arithmetical classification, with arbitrary lines of division 

 based on the proportions of different constituents, might, he thought, 

 be of considerable value as a kind of pigeon-hole arrangement of 

 analyses for convenience of reference. The silica percentage had 

 in fact been widely employed in this way. In that case there was 

 only one substance employed as an index. If, on the other hand, 

 the proportions which one group of constituents bore to another 

 were adopted as the basis of the classification, it was important 

 that the increase or decrease of the associated substances in the 

 ■differentiation of rocks should be determined by similar conditions. 

 Even if it were admitted that crystallization was the predominant 

 process in the differentiation of a magma, the eutectic theory 

 would give us little assistance except in the more basic rocks, for 

 up to the present it had been worked out only in anhydrous 

 magmas. The presence of water profoundly affected the order of 

 crystallization of the rock-forming minerals, and the succession 

 enunciated by Hosenbusch was, he believed, an expression of the 

 differences in the affinity of rock-constituents for water. Those 

 with the greatest attraction for water, the ' hydrophil ' constituents, 

 if the expression might be allowed, were the last to separate from a 

 water-bearing magma : they included silica, and the alkali felspars 

 and the felspathoids. The two last-named groups, however, probably 

 existed in the magma, in the form of silica and the alkaline alumi- 

 nates which were highly soluble in water. This selective action 

 of water appeared to operate, both in segregation by crystal- 

 lization and in magmatic differentiation in the fluid state. If 

 the hydrophil constituents had been grouped together by the 

 American Quantitative Classification, it might have been of some 

 service to petrologists. Unfortunately the ' salic ' minerals of its 

 authors included, not only these substances, but anorthite and any 

 molecular excess of alumina over the alkalis. Anorthite, however, 

 played a part totally different from that of the alkali felspars, 

 though it occurred in solid solution with albitc as a result of com- 

 munity of crystalline structure. It was, as a matter of fact, 

 •essentially a basic silicate, which, unless it was present in very 

 small amount or was 'brought down' by albite, crystallized out 



