THE UNKNOWN FACTORS OF EVOLUTION. 87 



ontogenic variations in different stages of development, and 

 under what circumstances do ontogenic variations become 

 phylogenic ? 



This brings us to an analysis of ontogenic variations in the 

 plane of time as provisionally expressed in the following 

 table : — 



Origin of Variations during Life History. 



A. Ontogenic Variations. Theories of Causation. 



{a) Gonagenic, i.e., those arising in Theoretically connected with patho- 



the germ-cells, including the ' Blasto- logical, nutritive, chemico-physical, ner- 



genic ' in part of Weismann, the ' Pri- vous influences, as implied by Kolliker 



mary Variations ' of Emery. and others, including doubtful phe- 

 nomena of Xenia and Telegony. 



{b) Gamogenic, i.e., those arising dur- Theoretically connected with influ- 



ing maturation and fertilization, in- ences named above, also with the com- 



cluding the ' Blastogenic ' in part of bination of diverse ancestral characters, 



Weismann, ' Secondary,' or ' Weismann- ' Amphimixis ' of Weismann. 

 ian variations' of Emery. 



(<:) Embryogenic, i.e., those occurring Theoretically connected with exten- 



during early cell division, including the sive anomalies due to abnormal seg- 



' Blastogenic ' and 'Somatogenic' in mentation and other causes, as observed 



part of Weismann. in the mechanical embryology of Roux, 



Driesch, Wilson, and others. 



(d) Somatogenic, i.e., those occurring Connected with reactions between 



during larval and later development the hereditary developmental forces of 



after the formation of the germ-cells. the individual and the environment. 



B. Phylogenic Variations. 



Variations from type, originating in any of the above stages which become 

 hereditary. 



The above table illustrates limits which certainly should 

 not be sharply drawn between the successive stages of onto- 

 geny, although intermediate focal points of real distinction must 

 exist. The four terms proposed are not in the sense of the 

 ' blastogenic ' and ' somatogenic ' of Weismann, for there is no 

 implication of his petitio principi, namely, of the separation 

 of the hereditary substance or specific germ-plasm from the 

 body-cells. Even before somatogenic separation has taken 

 place we have little or no reason to believe that all the blasto- 

 genic, gonagenic, or gamogenic variations which may have 

 arisen from various causes will become phylogenic. 



If we carry our analysis into the ^ plane of fitness' the first 

 point which arises is whether variations are normal, including 



