BEL ATION SHIP OF SPECIES. 99 



the insect -eating proclivities of Fringilla a difference may have 

 arisen in the germ-plasm of this group as compared with 

 Acanthis and kindred genera, which causes the infertility 

 between them, but that otherwise the relationship between 

 Fringilla, Ligurinus, Acanthis, Serinus, &c, is co-equal. I do 

 not think this view probable, however (although it might 

 account for their readiness in pairing), but imagine the one I 

 gave when commenting on this to be more correct. 



It is hard to keep from what appears to be arguing in a 

 circle, and arriving at the commencement of our argument 

 again, and at the apparently opposite conclusion that the germ- 

 plasm changes through diet, while the structural change may not 

 occur, and that therefore the latter is, in such cases, the truer 

 test of relationship, and that the usual methods of classification 

 are correct. In this way, one might imagine a germ-plasmatic 

 change, let us say, through diet, but there would seem to be no 

 apparent necessity that there should invariably or of necessity 

 be any great corresponding structual change, unless environ- 

 ment through adaptation ordered otherwise ; and one might 

 argue that this is indeed what one apparently finds in the above 

 quoted cases of Fringilla and Acanthis. Allowing this, one may 

 quite reasonably further contend that Ligurinus evoluted out 

 further back than the separation of Fringilla and Aca?ithis, but 

 through the similarity of food keeps a similar germ-plasm, but 

 on account of some physical labour develops a diverse structure 

 (powerful beak) for crushing larger seeds. The development 

 being diagrammatically something as follows : — 



Fig. 1. 

 (Inner lines show germ-plasmatic, and outer lines structural 

 similarities or changes.) 



_ -_=_=*5=g====Nf A. Ligurinus. 



4^========= TTtT^rrT -^- . '.".'.'.:".'. ~.~ ==-~ B. Acanthis. 



C. Fringilla. 



Where A and B retain the same fertility-producing plasm, 

 from which C has diverged, structural changes being vice versa. 

 And from this one might assert that Fringilla and Acanthis 

 were more nearly related to each other than either to Ligurinus. 



i2 



