GABBEO AND DIORITE-PORPHYEIES. 245 



In some cases the small plagioclases are clouded brown in the same manner 

 as the larger ones already described. The phenocrysts are light-green 

 augite and paramorphs after (?) olivine. The augite phenocrysts contain 

 much magnetite in minute grains, which are sometimes arranged zonally. 

 One of the thin sections contains several clusters of grains of grass-green 

 augite, slightly pleochroic, which form borders around other minerals. In 

 one instance they inclose grains of quartz, calcite, and pale-yellow garnet, 

 the grains of garnet being mingled with those of augite. 



A question now presents itself which is interesting because of its 

 importance in discussions regarding systems of classification of igneous 

 rocks, all of which systems are designed to be as natural as possible. The 

 question is, Which of two lines of relationship is to be followed in a 

 particular instance? We have described the subaerial breccias and lava 

 flows together in a group, and then given with a description of the 

 dikes and sheets of intrusive rocks of similar character. Shall we continue 

 the description of the remainder of the rocks occurring in dikes, which 

 exhibit a wider range of variation than those just described, or shall we 

 follow the dikes continuously a few feet farther into the granular core and 

 take up the consideration of their more crystalline forms? In the first 

 instance we have a natural grouping based on similarity of occurrence and 

 of outward petrographical habit — that is, of general aspect derived from 

 their phase of crystallization, which is combined with variations in mineral 

 and chemical composition. In the second case we have a natural connec- 

 tion based on actual continuity of mass, and when this is not directly 

 traceable in the field the connection is one of identity of chemical composi- 

 tion. With these constant factors are combined the variable ones expressed 

 by differences in mineral composition and in degree of crystallization. In 

 nature these relationships either exist as accomplished facts in regions of 

 extinct volcanoes or they are in process of development in regions of active 

 ones. In the former they exist, not in two groupings such as we have 

 depicted, but as one great complex system of relationships, involving varia- 

 tions in chemical composition and crystalline structure and still more 

 intricate variations in mineral composition. While in regions of active' 

 vulcanism we may readily conceive of a number of processes of rock- 

 making being in action in different places at one time, in the vast com- 

 plexity of rocks resulting from the working out of these processes it is the 



