PROBLEMS OF PETROGRAPHIC CLASSIFICATION 803 



as plutonic, occur side by side with rocks termed hypabyssal, and 

 some have been formed near if not at the surface. In treating the 

 hypabyssal rocks, Hatch shows that he is largely influenced by 

 Brogger's proposition to apply this term to the "Dike rocks" of 

 Rosenbusch. The complex genetic, structural, and textural 

 relations involved in this conception are well known and need not 

 be stated here. And in view of an earlier discussion of this sub- 

 ject^ I need only repeat that the use of these natural relations of 

 rocks in classification must be appropriate and consistent with the 

 facts, or the result is unnatural. 



The expression "natural system" as applied to petrographic 

 classifications should, I think, refer to other things as well as to the 

 use of factors of natural occurrence. The rocks themselves have 

 various natural characters or properties. The systematist makes 

 use of these characters in classification and should do so as appro- 

 priately and consistently as possible. In proportion as the nature 

 of rocks is rationally applied in their classification the result may 

 be termed a natural system. It is with reference to this idea that 

 the further systematic propositions adopted by Hatch will be 

 considered. They are, for the most part, not new. 



The next step is the distinction between feldspathic and non- 

 feldspathic rocks. The two groups of substances indicated by this 

 division occur in rocks in all amounts from o per cent to 100 per 

 cent for each. If this varying constitution of two exclusive groups 

 of minerals is to be used as a systematic factor in classification, is 

 it not the natural way to recognize the varying proportions of the 

 two groups ? To call all rocks distinctively feldspathic which con- 

 tain as much as 10 per cent of feldspar is to ignore the fact that 

 rocks with 10 per cent or more of pyroxene are by the same measure 

 pyroxenic. In short, this is a quite unnecessarily arbitrary and 

 unnatural procedure and opposed to the almost universal tendency 

 of modern petrographers to introduce a quantitative element into 

 system wherever it is appropriate and feasible. 



It is to be presumed that "chemical considerations" lead Hatch 

 to form three groups of rocks called acid, intermediate, and basic, 



'Whitman Cross, "Natural Classification of Igneous Rocks," Quar. Jour. Geol. 

 Soc, LXVI (19 10), 470-506. 



