660 W. Cross — Petrograjphic Classification. 



the power of chemical composition to influence the mineral 

 composition of the rock is considered. It is apparently this 

 power which is the natural factor in chemical composition 

 which should be used in systematic classification of rocks. 



The customary statement of an analysis follows an old con- 

 ventional form, giving simply the sum total of certain acid or 

 basic radicals found in the material analyzed. The petrog- 

 rapher, in his attempts to interpret the significance of the 

 proportions of a single analysis or of the differences in a series 

 of analyses, as bearing on mineral development, has restated 

 the analyses in various ways. They have been translated into 

 a percentage statement of the elements. The molecular pro- 

 portions of substances stated by analysis have been calculated 

 and expressed in percentages. It is clearly possible, moreover, 

 to transform all rock analyses into terms of a certain set of 

 mineral molecules, selected for the purpose, and following 

 some invariable rule for the calculation, as is necessary to 

 secure comparable results. Such a set of molecules is the 

 norm on which the C. L P. W. system is based. 



This primary character of the norm was not pointed out 

 with sufficient clearness by the authors of the C. 1. P. W. sys- 

 tem in their original publication. They were too intent on 

 explaining the reasons for the norm ^ehosen, its practicability, 

 and relation to the mode of the rock. Nevertheless the 

 abstract character of a norm is evident from the discussion of 

 the question as to the particular norm best suited to the pur- 

 pose of the C. I. P. W. system. 



The interpretation of analyses has also been attempted by 

 means of various ratios between single constituents or groups 

 of constituents and these factors have been used to character- 

 ize rock types established on mineral composition. In fact 

 the use of chemical composition in a systematic way most fre- 

 quently attempted by petrographers has been not a true chem- 

 ical classification but rather a chemical characterization of pri- 

 mary mineral types, themselves established on a very faulty 

 basis. 



The chemical classification of igneous rocks may be effected 

 by the use of the data presented in the analyses as stated by 

 the chemist, as by percentage of silica, or of any consistent 

 restatement or interpretation of those analyses, as in Rosen- 

 busch's use of Atomzahlen. If chemical composition be 

 chosen as the first and mineral constitution as the second fac- 

 tor in a system, it seems natural to apply the first factor in 

 some way which expresses as nearly as possible some definite 

 relation to the second. The C. I. P. W. system is an attempt 

 to do this, recognizing that variable physical conditions attend- 

 ing the formation of the rock from the magma modify in dif- 



