464 Canadian Record of Science. 



ern volcanic rock composed of Sanidine and ordinary Aug- 

 ite, will have quite a different chemical composition from 

 one containing large amounts of Anorthoclase and Acmite, 

 although both, being compounds of an alkali feldspar and 

 pyroxene, would be Trachytes if classified according to 

 mineralogical composition. In the same way we have in 

 the Augite Syenites a series of rocks presenting great 

 diversities in chemical composition as well as in petro- 

 graphical relationship. At present, however, a purely 

 chemical classification presents many difficulties. A know- 

 ledge of the exact chemical composition of many rocks is 

 wanting, while for practical purposes it would be impossible 

 to adopt any method of classification which requires a com- 

 plete chemical analysis of a rock before its name and pro- 

 per position could be ascertained. Mineralogical composi- 

 tion and structure must still be important factors in any 

 scheme which is to be generally adopted. 



In the accompanying table, however, a classification 

 according to chemical composition has in a general way 

 been secured. On the left we have rocks rich in alkalies, 

 principally potassa. Going towai-d the right in the table 

 we have, first, rocks in which this alkali is largely replaced 

 by soda (the Leucite rocks, however, forming an exception) 

 then those in which this alkali is associated with progress- 

 ively larger proportions of lime, while on the extreme right 

 are rocks which are free from all alkalies, but in which lime, 

 magnesia and oxide of iron are present in large amount. 

 Speaking generally, moreover, it may be said that the 

 rocks decrease in acidity from left to right, the principal 

 exception being the small group of Nepheline, Leucite and 

 Melilite rocks. 



In order to bring out these chemical relationships as clearly 

 as possible and place the several groups of rocks in posi- 

 tions where their affinities are more clearly shown, I have, 

 at Prof. Eosenbusch's suggestion, given to several groups 

 of rocks positions other than those which they occupied in 

 his former table, or in the last edition of his book. The 

 Nepheline, Leucite and Melilite rocks, for example, instead 



