SYSTEM A TIC PETROGRAPHY 47 1 



crimination of the group named "Dike rocks." In this light 

 this group is certainly not co-ordinate with the other two of the 

 same rank as defined. 



In the subdivision of the three classes of " Massive rocks " 

 Rosenbusch applied mineral composition as a factor, producing 

 Families. The quantitative composition, either chemical or 

 mineral, received no expression, so that, for example, anorthosite 

 and the most highly pyroxenic gabbro or norite are found 

 together in the gabbro family. Moreover in the porphyritic 

 Dike rocks only the phenocrysts are considered in determining 

 the systematic position of a given rock. Thus a porphyry having 

 the chemical composition of a granite is referred to syenite- 

 porphyry in case its excess of silica chances to be confined to 

 the groundmass, while had quartz phenocrysts been present it 

 would have been called granite-porphyry. In the Effusive 

 rocks, Rosenbusch hesitates to apply the same rule consistently. 

 The families of these rocks are defined in very general terms as 

 the "equivalents" of certain granular rocks and described as con- 

 taining certain phenocrysts in a groundmass of variable appear- 

 ance. 



As in the earlier system, all feldspar-free rocks of the deep- 

 seated class are united as Peridotites. The peculiar character of 

 the Dike rocks as a division not co-ordinate or co-extensive in 

 range with the Deep-seated or Effusive rocks appears in the fact 

 that mineralogical groups corresponding to the granites, syenites, 

 and diorites, only, are recognized. 



Geological age is acknowledged by Rosenbusch to have been 

 assigned a higher value in classification than belongs to it, but it 

 is retained, in the Effusive class, and the use of duplicate terms 

 perpetuated. 



H. Rose?ibusch, i8g6. — The third edition of the "Mikro- 

 skopische Physiograpliie der massigen Gesteine," issued in 1896, 

 contains no essentially new systematic features. The principles 

 above set forth are reaffirmed, and, save for the elaborated dis- 

 cussions of magmatic differentiation, which show more plainly 

 than in the preceding edition the strong influence of hypo- 



