﻿502 DE. WHITMAN CEOSS OX THE XATTJEAL [Aug. 1 9 10, 



old system. And, as Mr. Harker points out, the names of rocks 

 found under other subrangs have the same significance as those 

 under Toscanose. Washington 1 found that rocks called ' granite ' 

 belong to 20 different subrangs, ' trachytes ' to 23 subrangs, 

 ' andesites ' to 27 subrangs, which is neither an argument for the 

 homogeneity of the old group nor for the heterogeneity of the new 

 units of definite bounds ! 

 Mr. Harker remarks that 



' the manner in which tbe various oxides are actually combined in the minerals 

 —in the author's terminology, tbe " mode " of the rock — finds no place in tbe 

 classification.' (' The Natural History of Igneous Eocks ' 1909, p. 365.) 



The mode naturally finds no place in the magmatic classifica- 

 tion expressed by the new systematic nomenclature. But it is not 

 true that the authors of the Quantitative System did not provide, 

 by many definite suggestions, ways of stating in any given case 

 the actual mineral and textural characters found in a given rock- 

 unit. It is undesirable to enlarge on this subject here, but the 

 idea of the authors of the Quantitative System may be repeated. 

 It is that mineral composition and texture are each variable 

 qualifying features of any consolidated magma. The ends of 

 systematic petrography will be best met by using these characters 

 as qualifiers in the ways which experience may show to be most 

 effective. By these means, and only by them, I think, can the 

 appropriate descriptive terms needed for the illimitable 

 variations possible for each magmatic unit be connected with the 

 fundamental chemical character. 



The foregoing review of principles, of facts of nature, of points 

 of view, will, I hope, assist in directing discussion towards the vital 

 question, Is a natural classification of igneous rocks 

 possible? Mr. Harker outlines the 'Desiderata in an Ideal 

 System.' I can quite agree with him in this statement. In the 

 abstract I should also agree with the forceful final comments on 

 the Quantitative System embodied in these sentences : — 



' The strongest objection to the Quantitative Classification is, however, that 



it is planned entirely on a 'priori lines As von Ricbthofen aptly remarked, 



in order to establish a more natural system, we have not to make groups, but 



to find them The scbeme of Nature (is) based not on arithmetical but on 



physical and chemical principles.' {Op. supra tit. p. 366.) 



But these truisms and ideal characters of a natural system do 

 not make the system. What are the physical and chemical prin- 

 ciples practically applicable to petrographic system ? Who shall 

 find the natural groups ? If the groups do not exist, if the prin- 

 ciples are unfortunately not suitable to systematic use, then must 

 we not come to an a priori basis for that classification on which 

 specific nomenclature must rest ? 



1 ' Chemical Analyses of Igneous Eocks ' Prof. Paper No. 14, U.S. Geol. 

 Surv. 1903, p. 61. 



