534 Colonel R. Meinertzhagen on [Ibis, 



The breeding-season is the most important period in a 

 bird's life, and it is the environment of the breeding-area 

 which influences variation more than av inter-quarters. It is 

 the fact that a bird like the Stonechat, which in various con- 

 tinental and insular forms breeds throughout Europe, Asia, 

 and Africa, and many of whose races co-exist in the same 

 winter-quarters, that induces me to believe that the environ- 

 ment or isolation of the breeding-quarters is in the main 

 responsible for variation, and that this variation is the result 

 of natural selection working on gradual change, and not the 

 result of the perpetuation of some aceidental and sudden freak. 



I doubt very much whether mutation, or the sudden 

 a{)pearance of " sports,"' can establish a subspecies, though I 

 believe a species might evolve another species by splitting, 

 originating in mutation but eventually becoming a constant 

 and heritable germinal character. 



I also believe that wherever geographical races are con- 

 nected by intermediate forms, it is absolute proof that the 

 differences are entirely due to environment and not muta- 

 tion, except where such intermediate forms are hybrids. 



As a corollary to this, it follows that environment can affect 

 the gamete; and this seems quite a reasonable assumption, if 

 it is accepted that geographical causes influence the bird at 

 all. Surely it is just as possible for the gamete to gradually 

 change and enable new constant characters to become 

 heritable, as it is for the gamete to suddenly change and 

 embody heritable characters. 



The mutationist* will say that there is no evidence to show 

 the gamete alters, and that, to perpetuate any environmental 

 variation, the conditions to produce such variation must be 

 repeated in each successive generation. In other words, 

 acquired characters are not heritable. 



That may be true for artificially-, accidentally-, or ab- 

 normally-acquired characters, but it is not true for environ- 

 mentally-acquired characters, which are certainly heritable 

 for one generation, but which appear to react to extremes 

 of environment especially when artificially reproduced (cf. 

 Beebe's experiment on Doves), 



