1 20 TJic Amcricaii Geologist. August, i898 
varied portions of these two magmas. The supposed mode 
of (Hfiferentiation must explain not only the birth of a series 
of rocks of the same family, but also the origin, at great dis- 
tances, of certain series, and on the contrary, also the possi- 
bility of remarkable differences between two neighboring cen- 
ters. It must, further take account of igneous rocks found at 
great depths and of those that solidify at the surface of the 
earth. With the former it touches the problem of contact 
metamorphism, and that of the striking resemblance of 
the crystalline schists to the massive granitoid rocks, 
and of their apparently sinmltaneous production in certain 
deep-seated zones of the earth's crust. This borders on the 
question of exomorphous (regional?) metamorphism. Along 
with endomorphous metamorphism is found the phenomenon 
of the absorption, by certain granitoid rocks, of large amounts 
of their surrounding rock walls and the radical transforma- 
tion of their magmas by such local absorption. 
In connection with the classification of rocks found at the 
surface the problem of vulcanism is involved. Indeed the 
study of solid magmas spread out on the surface cannot be 
separated from that of volatile emmanations, which are of a 
formidable volume and of a weight comparable with that of 
the fused products. These fumaroles, these clouds of vapor of 
water, of chlorides, of sulphides, of different carburetted ex- 
halations, continue long after the cessation of the local par- 
oxysms. They testify at least to the connection of metalliferous 
veins and concretions with volcanic eruptions. 
Therefore, exomorphous (or concact) metamorphism lead- 
ing to the production of the crystalline schists at the expense 
of beds incontestably stratified; the probable existence of deep- 
seated zones where general metamorphism gives birth simul- 
taneously to the schists and gneisses and to the granitoid 
masses; endomorphous metamorphism modifying profoundly 
from place to place these granitoid masses by partial solution 
and incorporation of the surrounding rock ; extended transfer- 
ence of certain elements by fumarole action — such are the facts 
which a theory of the differentiation of eruptive magmas must 
necessarily take account of, without losing sight of the differ- 
ent modes of action of the two different fundamental magmas. 
3. These facts scarcely appear to be susceptible of being 
