EVOLUTION OF THE COLORS OF BIRDS. 125 



Thirdly, some specific characters, at least, are ap- 

 parently of no conceivable use, but these may be due 

 to the direct action of the environment or to the use 

 and disuse of parts. 



As regards the bearing of the above discussion upon 

 the evolution of the colors of birds, in the light of the 

 evidence adduced, it would hardly seem justifiable to use 

 physiological selection, unless the facts admitted of no 

 other rational explanation. It is always an easy loop- 

 hole for escape in accounting for any difficult character, 

 and for this very reason should be treated with especial 

 caution. 



GULICK ON ISOLATION. 



One of the most valuable contributions made of late 

 years to the literature of evolution is the series of 

 articles upon isolation by the Rev. J. T. Gulick. His 

 views coincide with those of Mr. Romanes upon many 

 important points, while their theories overlap to a cer- 

 tain extent. Without considering in detail the relation- 

 ship between their respective views, it may be said that 

 they both agree that many specific characters are use- 

 less, and that isolation is a necessary factor in their 

 origination, but Gulick does not emphasize physiological 

 isolation as Romanes has done; while he asserts, con- 

 trary to Romanes, that not only species, but larger 

 groups as well require isolation as a factor in their 

 origin. His central idea is this: no two groups of 

 individuals of one species contain exactly the same pro- 

 portion of variations in their aggregate number; accord- 

 ingly, if two or more miscellaneous groups are isolated, 

 inasmuch as free intercrossing is no longer practicable 

 to preserve the average, each division will necessarily 

 start off on an independent line, irrespective of its 

 environment. Futhermore, natural selection without 

 isolation could not produce divergent, but only linear 



