﻿Vol. 66. J CLASSIFICATION OF IGNtiOUS ROCKS. 489 



V. Factors of Mineral Composition. 



The actual mineral constitution of an igneous rock (its mode) is 

 the second of its features in importance, because it is very largely 

 an expression of the fundamental chemical composition and also 

 determines many of the most obvious characters of the rock. If a 

 natural classification cannot be based on the chemical composition, 

 we must turn next to the mineral characters in our search for 

 appropriate factors of systematic arrangement. 



At the outset it must be recognized that mineral composition is 

 not a simple function of the chemical constitution of the magma. 

 Magmas of complex character may form various mineral combina- 

 tions under differing conditions of crystallization. Any attempt 

 to express chemical character through mineralogical classification 

 must, therefore, be only an approximation. 



The minerals of igneous rocks are much more numerous than the 

 chemical elements of corresponding importance in them, and many 

 minerals have a known quantitative range in rocks, from a trace to 

 strong predominance. Hence systematic divisions on mineral com- 

 position are necessarily arbitrary or artificial, and the difficulties of 

 constructing a logical system by this means are even greater than 

 on a chemical basis. 



The relationship of mineral composition to factors of genesis or 

 occurrence is complex. Some constituent minerals of igneous rocks 

 are formed only under a limited, and others under a much wider, 

 range of chemical or physical conditions. Moreover, many rocks 

 contain minerals formed at different places under different con- 

 ditions, as illustrated by intrusive and effusive masses. These 

 facts are of great importance in the natural history of the rock, but 

 are not practical bases of systematic classification. 



It is so generally admitted that all mineralogical classification of 

 igneous rocks is arbitrary, that this point might well be dismissed 

 here, were it not for the desirability of referring to current usage. 

 It was once supposed that certain approximately proportionate 

 combinations of minerals constituted groups which were much 

 more abundant than the rocks of intermediate composition, and 

 were thus in a measure natural units, such as granite, syenite, 

 diorite, etc. Although no one may now maintain this position, the 

 fact is tacitly ignored by many petrographers that these and all 

 other terms based on mineral composition cannot be precisely 

 defined, as is necessary for accurate and universal concordant 

 classification of rocks, without recourse to an arbitrary quantitative 

 factor. The necessity for sharp definition is becoming more and 

 more evident as petrographers study the rocks of certain series in 

 detail; and here and there the arbitrary quantitative line is being- 

 introduced into definition, but without a consistent logical basis. 

 The outcome can but be confusion worse confounded. 



The distinction between felspathic and non-felspathic 

 a. J. G. S. No. 263. 2 l 



