XIII. ] ORIGIN OF CRYSTALLINE ROCKS. 295 



in the elements of the solidifying layers, such as is seen in many 

 slags, and in certain eruptive rocks. (Ante, page 186.) Add to 

 this notion that of the separation of the fluid or, rather, viscid 

 mass into two or more layers of different composition and 

 density (ante, page 3), and we might have generated from 

 them, by their solidification under the above conditions, the 

 various types of stratiform feldspathic, hornblendic, and chrys- 

 olitic rocks, which would afterwards be penetrated by injections 

 from the yet liquid portions below. If now we imagine the 

 various plutonic rocks thus formed, both stratified and unstrati- 

 fied, to be the subjects of epigenic or pseudomorphous changes, 

 by which some beds or masses were converted in serpentine or 

 into steatitic or chloritic rocks, while others were changed into 

 limestone, quartzite, or iron-oxide, we shall have as clear a con- 

 ception as it is possible to form of the vaguely denned views of 

 Naumann, Bischof, and their school, as to the origin of the 

 crystalline rocks as we now find them. 



Naumann, while denying the sedimentary origin of the great 

 mass of crystalline schists, admitted, however, the conversion 

 of younger uncrystalline sedimentary strata, in certain cases, 

 into crystalline gneisses and mica-schists, resembling those of 

 the primary formations, and like them subject to epigenic 

 changes. That such crystalline rocks have ever been formed 

 from the alteration of palaeozoic or more recent sediments, 

 except locally (pages 18, 298, and 310), is, however, more than 

 doubtful, as will appear from the examination of the supposed 

 examples of this conversion in the preceding pages of this 

 paper, and also in the following one on the Geology of the 

 Alps. In connection with these two papers are given the views 

 of Gumbel (page 305) and of Favre on this important question. 



These crystalline rocks, whatever their origin or mode of 

 formation, appear to be in all cases older than the palaeozoic 

 sediments. They belong to at least three or four geognostically 

 discordant series, and are moreover occasionally associated with 

 fragmentary rocks, which render it impossible to admit for them 

 any other than an aqueous sedimentary origin, in accordance 

 with the view already defined on page 286.] 



