EVOLUTION OF THE COLORS OF BIRDS. Ill 



To overcome these three difficulties he has proposed 

 a new factor, physiological selection or the segregation 

 of the fit. It is a known fact that isolation favors the 

 production of new species. Oceanic islands in partic- 

 ular attest to this. According to Mr. Romanes, how- 

 ever, the advantage of geographical isolation is not 

 always offered to incipient species, which often branch off 

 in the midst of the parent stock. It is a noteworthy cir- 

 cumstance, however, that of all parts of an organism 

 the reproductive organs are most susceptible to change, 

 and if a variation in these parts should occur among a 

 number of individuals making them sterile with the 

 rest of the species but fertile among themselves, they 

 would be physiologically isolated even in the midst of 

 their associates. In the words of the author, this view 

 enables us to regard many, if not most, natural species 

 as the records of variation in the reproductive systems 

 of ancestors. 



In considering the three difficulties in the way of natu- 

 ral selection the author adduces considerable evidence to 

 show that the difficulties are real. He asserts that the 

 fundamental or primary diflference between species is 

 sterility, this being a common distinguishing feature 

 in nearly every instance. Other differences consist in 

 innumerable distinctions of structure, color or markings, 

 which he calls secondary differences. Inasmuch as 

 these secondary differences are never exactly alike in two 

 species while the primary difference is always the 

 same, he demands an explanation of the constancy of 

 this distinction, asserting that all previous theories 

 have dealt only with the secondary differences. Even 

 where geographical isolation has originated species it is 

 not possible to account for the almost universal sterility 

 existing between them. " The consequence is," he 

 says, " that most evolutionists here fall back upon a 



