460 WHITMAN CROSS 



difference between the trachytoid and the granitoid structures, 

 since they believe that certain mineralizing agents ("agens 

 mineralisateurs ") are necessary to the latter development. 



This view of rock structures makes the shape of the min- 

 eral grains all important and casts aside the formal relation- 

 ship prominent in the porphyritic structure as of little value. 

 The equidimensional grain and the elongated microlite are placed 

 in fundamental opposition to each other. 



The mineralogical composition of rocks is applied for their 

 classification in a qualitative way, similar in some respects to 

 that adopted by the German petrographers, but with the all 

 important modification that only the minerals of the second period 

 of consolidation are considered. Such a principle may be desig- 

 nated as subjective, extremely unnatural and highly artificial. 

 There is in this system no attempt to express the chemical com- 

 position of the rock in terms of its minerals, for in some cases 

 all the minerals of a rock are used in its classification, where 

 there was no first period of consolidation, and in other cases 

 only a small portion of the constituents, as in porphyries with 

 abundant phenocrysts and microlitic groundmasses. Since in 

 porphyries this portion of second consolidation bears no definite 

 quantitative relation to the mass as a whole it must often happen 

 by this system that rocks of widely different chemical composi- 

 tion will be brought together and, conversely, that rocks of the 

 same chemical character and even of the same magma will at 

 times be separated. For example, certain intrusive quartzose 

 hornblendic diorite-porphyries of the Rocky Mountain region, 

 in which hornblende and plagioclase are developed entirely in 

 phenocrysts, would fall among the microgranulites while their 

 granular equivalents would be quartz-diorites. It is also clear 

 that all but granular rocks would be classified in this scheme 

 by their most obsure constituents, often to the neglect of every 

 prominent megascopic character, and systematic petrography 

 would become purely a microscopical science. It is interesting 

 to recall at this point the principle announced by these authors 

 and quoted above that a rational classification of rocks must be 



