124 CALIFORNIA ACADEMY OF HCIENCES. 



are necessary to the gradual building up of adaptations, 

 while these, like the woody and permanent branches, 

 grow continuously in importance through all, the tree of 

 life. Now, it is the office of natural selection to see to 

 the growth of these permanent branches; physiological 

 selection has to do only with the deciduous leaves." 



Having now stated the principal arguments for and 

 against physiological selection, with as little bias as 

 possible, it only remains to pass judgment upon their 

 merits. This, however, does not appear to be an easy 

 task; but as usual in such cases, it will probably be 

 found that both sides are partly in the right and partly 

 wrong. Thus, with regard to sterility, natural species 

 are as a general thing sterile inter se, while domestic 

 races are usually fertile, but these conditions are by 

 no means uniformly constant. Physiological isolation 

 might explain this state of affairs, while Mr. Seebphm's 

 hypothesis has quite an aspect of plausibility, — or both 

 views might be correct ; whereas, various secondary 

 factors have been instrumental in bringing about this 

 condition. There seems to be hardly any evidence, 

 however, that sterility with the parent stock is generally 

 one of the first variations of a diverging race, but if this 

 be not so, then the chief aim of the theory — viz. to 

 account for the early stages of divergence — fails. 



Isolation may be advantageous in preventing the 

 swamping effects of free intercrossing, although Mr. 

 Romanes has granted that the size, shape, or degree of 

 coloration, may be modified without its aid. Further- 

 more, geographical isolation is an ever-present factor, 

 while a slight change in habits on the part of certain 

 individuals would effect their isolation, psychological 

 isolation, as suggested by Galton, may also be a valid 

 factor, while the operation of physiological selection 

 itself need not be excluded. 



