74 



PB0CEEDr5rGS OF THE &EOLOGICAI. SOCIETT. 



may call type-forms to justify us in the assumptioii of the existence 

 of species for purposes of classification, and I tmst that I have 

 indicated a natoral and logical principle on which to proceed. The 

 relations, then, of the groups may be expressed graphically in the 



annexed diagram. 



In the above sketch of what appears to me a natural system of 

 classification, I take no notice, as above stated, of the geological age 

 of the rock, which is regarded as of primary importance by many 

 continental investigators, who consider Pre-Tertiary igneous rocks 

 always separable from those of later date. For this distinction I 



