PROBLEMS OF PETROGRAPHIC CLASSIFICATION 797 



from analyses of much-altered rocks. ^ With increasing experience 

 it becomes more and more clear that quantitative system names 

 should be based only on good analyses of fresh or but slightly altered 

 rocks. Termor's "calculated analyses" are clearly liable to errors 

 unfitting them for any systematic use, however valuable and justi- 

 fiable they may be in discussing the nature of the alteration that 

 has taken place. Kodurite of both types and the garnet rocks are, 

 Hke many of their associates, very interesting and unusual types 

 and, if their igneous character can be estabhshed and good analyses 

 furnished, some of them will undoubtedly fall in divisions of the 

 Quantitative System in which there are at present no other known 

 representatives, and new quantitative names may well be based 

 on them. 



Fermor is not satisfied with the result of applying the Quanti- 

 tative System to the four hypothetical rocks, and points out what 

 he regards as the defects of the system indicated by this case. He 

 considers it a serious flaw in the system that the subrang is reached 

 without reference to the important constituent normative tephroite 

 (i 2 . 93 per cent in la) and further that rocks poor in manganese may 

 fall in the same subrang. This criticism touches the fundamental 

 distinction between salic and femic components which it seems 

 unnecessary to discuss at this time. Other criticisms may best 

 be expressed in the following quotations : 



There certainly seems to be something lacking in a system which takes no 

 effective account, until the ninth subdivision of the system is reached, of a 

 constituent that is fourth in order of importance in the chemical analysis and 

 third in order of importance in the norm. 



The same criticisms apply to boiranose, but to a much smaller degree on 

 account of the much smaller amount of MnO present; but they apply in an 

 even more striking manner to the Kotakarra garnet-rock (spandite-rock) , in 

 which MnO is third in order of importance in the chemical analysis, and teph- 

 roite, Mn2Si04, is second in order of importance in the norm, of which it forms 

 nearly 25 per cent.^ 



^ Cross, Iddings, Pirsson, and Washington, Quantitative Classification of Igneous 

 Rocks, p. 166. This and all other references to the Quantitative System contained in 

 this paper are made with full approval of my colleagues. A review of all names 

 proposed since the Quantitative System was first published will be given by Dr. Wash- 

 ington in a forthcoming new edition of his Chemical Analyses of Igneous Rocks. 



^ Records, -p. 217. 



