EVOLUTION OF THE COLORS OF BIRDS. 61 



It is still not inconceivable that both external and in- 

 ternal forces may influence variation. There is no pos- 

 sible doubt that sexual combination is productive of 

 variations, but I think it has been shown that these are 

 not progressive. It is doubtless this factor which pro- 

 duces individual variation in species which have re- 

 mained unmodified through long geological ages. This 

 class might well be termed conservative variations. 



It is a fact well known to gardeners and breeders that 

 when a species is placed under entirely new environ- 

 mental conditions it is apt to produce variations which 

 are well marked, know^li as sports. It is such variations 

 as these, due to the influence of environment, which, it 

 would seem, are originative of new characters and might 

 be called progressive variations. According to this view 

 not only certain unusual modifications are due to en- 

 vironment, but all variations wdiich are new have been 

 thus produced. It must not be supposed that the en- 

 vironment acts definitely upon the organism, compelling 

 it to vary in a single direction. Such may be the case 

 in some instances, but not universally. Change in nu- 

 triment may produce very different results in different 

 individuals, even of the same species, and would thus 

 furnish natural selection with a very diverse, although 

 by no means unlimited, assortment to pick from. Climate 

 would often act in a more direct manner in modifying 

 organisms as will be seen later on in considering the 

 varieties of North American birds. Use and disuse of 

 parts would also be productive of definite variations. 

 According to the principles above stated, variations may 

 be classified as follows: 



I. Conservative . Occurring in all directions within 

 the limits of variation of the 

 species. Produced by sexual com- 

 bination of unlike individuals. 



