W. Cross — Petrographic Classification. 659 



the Quantitative classification, and adds : " There are many 

 of us who fail to see in this elaborate system anything but an 

 admirable card classification of analyses." The same idea has 

 been expressed before in the remark that the system classifies 

 analyses, not rocks. 



Now the authors of the C. I. P. W. system attempted a 

 chemical classification of igneous rocks and spent long study 

 on it. I must personally express my surprise and regret 

 that a petrographer of Lindgren's standing should feel warranted 

 in making a belittling and by no means lucid criticism of 

 this work without a word of justification. His comment fails 

 to explain (1) what he finds admirable in the system as a 

 card classification of analyses, (2) how a classification of 

 analyses, using the data presented by them, can be admirable 

 if it does not classify the rocks analyzed, (3) why the sys- 

 tem does not classify the rocks, (4) how chemical data should 

 be used to classify rocks so that his objection may not apply. 



Like Daly's criticism this comment touches abstract principles, 

 in this case of chemical classification, but it rests partly, I think, 

 on a misconception of the C. I. P. W. system. 



The main interest a petrographer, as such, can have in 

 classifying rock analyses is as to their usefulness to him, as 

 complete or incomplete, good, bad, or indifferent, as Washing- 

 ton has done. He is interested in a good analysis only when 

 it represents the composition of a known rock. 



The fundamental importance of chemical composition of an 

 igneous rock, bearing, together with variable, and in part 

 extraneous physical factors, primarily on its magmatic and 

 genetic relations and later on its mineral composition, is beyond 

 any need of discussion. To be sure, an Irish petrographer 

 makes the claim that the chemical composition represented by 

 analysis is not fundamental because the chemical composition 

 of a specimen depends entirely on its mineral composition, 

 while the latter depends partly on physical conditions. The 

 classificatory value of chemical composition is stated by Daly, 

 after an admirable review of the difficulties of accurate 

 mineralogical classification, in these words — " the basis for an 

 ultimate classification is now universally found in the chemical 

 analysis (total analysis) of the rocks" (op. cit. p. 9). 



In discussing rock classification by chemical composition it 

 is desirable to first consider what a complete analysis of an 

 igneous rock represents. 1. It represents as nearly as we can 

 ever determine it the composition of the magma the consolida- 

 tion of which has produced the hyaline, hypohyaline or crys- 

 talline rock. 2. It is a concise report, exhaustive in its way, 

 made by an expert, on the fundamental material property of a 

 rock. But this report is far from satisfactory in form when 



