Appendix II, 209 



of Mr. Spencer's criticisms— including this matter of the 

 influence of a previous sire. Here he adopts exactly the position 

 which in my article I surmised that he would ; so that, to all 

 who have read The Germ-plasm, it must have appeared that 

 I was prophesying after the event. Hence the need of this 

 explanation. 



Lastly, in the same issue of the Contemporary Review, 

 Mi. Spencer explained: — 



Mr. Darwin's hypothesis of Pangenesis implies not only that 

 the reproductive cell must contain numerous kinds of gemmules 

 derived from different organs, but that the numbers of these 

 gemmules must bear to one another something like the pro- 

 portions which the originating organs bear to one another in 

 size. The conception involves many different kinds, whose 

 numbers are in many different proportions, and I supposed the 

 difficulty alleged was, that for the influence of a previous sire to 

 be communicated from the growing foetus to the mother would 

 imply not only the transfer of the various kinds of gemmules 

 derived from him, but also maintenance of their numerical 

 proportions, and that again these gemmules, diffused throughout 

 the maternal system, would have to be transferred in these pro- 

 portions to the subsequently formed ova. No such difficulties 

 arise if the units conveying hereditary characters are of one 

 kind only. 



From this it is apparent that Mr. Spencer has misunder- 

 stood " the difficulty alleged," and that the desired explanation 

 is not yet forthcoming. I did not say anything about " kinds " 

 or " proportions " of the carriers of heredity ; my difficulty 

 is to conceive of any mechanism whereby these carriers can 

 first directly influence the somatic-cells of the mother, and 

 then indirectly reflect this influence upon her germ-cells. 

 Also, I cannot see any obvious necessity for the intervention 

 of the " embryo " in the process. 



