﻿Vol. 66.~] CLASSIFICATION OF IGNEOUS ROCKS. 497 



immediate comparison. If a method can be found for analysing 

 the differences of composition expressed by the norms, the latter 

 ma}* serve as the basis of classification. 



The 'American' Quantitative System. — The foregoing 

 discussion was written before I had an opportunity of reading the 

 masterly treatise on the ' Natural History of Igneous Rocks ' by 

 Alfred Harker. For that work I have profouud admiration, and 

 feel that the science of petrology has been materially advanced by 

 this mature and progressive treatment of the major problems 

 of igneous rocks. I can cordially agree with the author in most of 

 his conceptions as to the natural relations of igneous rocks. In 

 general, I subscribe to his optimistic view (op. cit. p. 362) that 



' we have almost within grasp a fundamental principle analogous with that of 

 descent, which lies at the root of natural classification in the organic world,' 



but unfortunately see no ground for the hope that we can apply 

 that principle in petrographic system. The laws of magmatic 

 descent operate on such complex solutions and under such a 

 variety of conditions, that factors of necessary simplicity and of 

 clear application in logical system seem to me impossible of 

 realization. The future must decide which view is correct. 



It was no part of the original plan of this paper to advocate or 

 defend the quantitative system proposed by my colleagues and 

 myself, often referred to as ' the American system.'* But it has 

 been so severely criticized by Harker in the work just quoted, in 

 connexion with a general plea for the ' natural ' classification of 

 igneous rocks, that a brief reply to certain points seems not to be 

 out of place. 



I will pass over such comment on the system as comes from the 

 general objection to a quantitative as opposed to a natural system. 

 The preceding discussion shows why the desirable natural system 

 seems to me unattainable. The fact that Mr. Harker is able to 

 make no definite proposition of systematic value, after his ex- 

 haustive study of the natural relations of igneous rocks, seems 

 confirmatory of this view. The hope that in eutectics there may 

 be found a natural principle suitable to classification seems futile, 

 in view of the probable complexity of that relation and of other 

 considerations, as has already been pointed out. 



The general criticism of the Quantitative System that it 

 separates similar things is inherent in any system with precise 

 definition of units. The current mineralogical system is every- 

 where approaching that situation with no definite, consequent basis 

 for its precision. 



As to the basis of the system, the norm, I find a strange reluctance 

 on the part of Mr. Harker to acknowledge that it has any merit 

 or significance, while certain features are repeatedly referred to in 

 terms of condemnation, as wofully artificial. The basis is alluded 

 to as chemical in - a disguised form,' and as ' a circuitous manner 



