212 L. V. Pirsson — Classification of Igneous Rocks. 



or nothing. In actual practice, since this difficulty must be met, 

 those using this system in such cases fall back upon the texture of 

 the rock, for Rosenbusch holds that the textures are characteristic 

 of the modes of occurrence, and that from the former the latter 

 can be told.° He confesses however that this determination of 

 the geologic character from the texture (structure) is unsafe, but 

 feels that this should have no influence on the naming of the rock, 

 since a granite-porphyry is a granite porphyry, whether it occurs 

 in the normal form of a dike, or the border facies of a stock of 

 granite (plutonic), or, finally, as a central facies of a mass of 

 quartz-porphyry (effusive).^ One sees from this then, that, 

 whereas in theory the mode of occurrence is the basis of classifi- 

 cation, this must really be largely replaced by texture. Since the 

 petrographic result is the same, it would seem more logical to 

 adopt a uniform practice and use texture in all cases. 



Another defect is that so little weight is placed on quantitative 

 relations in the mineral groups. The value of this factor was 

 felt by Rosenbusch and applied in places, as in the theralite- . 

 shonkinite group, which is stated to have a mafic character, but it 

 finds no formal expression in the system. It obtains to some 

 extent an underlying recognition in the division into groups of 

 some of the rocks on the genetic-magmatic basis. 



It is to be observed that in this system, as in others, the division 

 lines must be arbitrarily drawn ; thus, whereas a rock consisting 

 of orthoclase with a little hornblende is a syenite, and one com- 

 posed of andesine with hornblende is a diorite, a rock containing 

 equal parts of orthoclase and andesine must leave the petrog- 

 rapher in doubt as to its classification. The same is true in 

 divisions by texture, and even at times in the geologic mode of 

 occurrence, as pointed out in a previous section. 



Modifications of the Geologic-Petro graphic System. — 

 The Norwegian geologist Brogger has made some 

 changes in the geologic-petrograpliic system, as presented 

 in the foregoing section. He prefers first to di\dde the 

 rocks into large families on chemical grounds (G-ranite 

 family, Syenite family, etc.) and then to apply the method 

 of geological occurrence to each family to form plutonic 

 and effusive groups/ He adds a third group which he 

 calls the hypahyssal as transitional between the other 

 two. The idea is to include the minor intrusions, which 

 in the form of dikes, sills, laccoliths, etc., appear above 



^ Elemente der Gesteiiislehre, 3d Ed. p. 74, 1910. 

 « Ibid. 



^ G-rorudit-Tinguait-Serie. Eruptivgest. Kristiaiiia-gebietes. Videnskap. 

 Skrift. I Math-nat. KL. 1894. No. 4. p. 125. 



