PROBLEM OF ORIGIN OF SPECIES 51 



and advances inward, but usually more rapidly along the shaft 

 than at the sides, thus resulting in two checkers with more or less 

 pointed tips. 



The direction of change again coincides with that of embryonic 

 development, the tip of the feather, where it begins, being first in 

 order of development. 



In many checkered rock pigeons we may find in the dorsal (inner) 

 'feathers of the bars undivided central spots, which pass gradually 

 into the typical checkers as we pass towards the lower (outer) ends 

 of the bars. Transitional stages of various degrees thus connect the 

 derived with the ancestral type in one and the same individual, and 

 so demonstrate that the two specific marks are not separated by im- 

 passable mutation gaps. 



While it is not necessary to go beyond the wild rock pigeons and 

 the multitude of domestic races descended from them to learn that 

 nature has here pursued one chief direction of color variation, always 

 leaving an open door, however, to minor modifications and improve- 

 ments through natural and artificial selection, it is, nevertheless, highly 

 instructive to make a comparative study of the whole group of wild 

 pigeons, in both adult and Juvenal stages. It is in this field that we 

 find the same lessons amplified and repeated in multitudinous ways, 

 confirmation confirmed, convergence of testimony complete. 



It will be sufficient here to cite a few examples. 



In the little ground doves (Chamcepelia passerina) of Florida, 

 Arizona, California, Central and South America, and the West Indies, 

 we find the turtle-dove pattern preserved in the whole breast region 

 and in the anterior, smaller coverts of the wings, while in the posterior 

 portion of the wings we meet with lateral spots or checkers, of higher 

 finish than in the rock pigeons. In many coverts of the wing, we find 

 the dark centres more or less reduced, with the distal ends of their 

 remnants in various stages of conversion into lateral spots. Here 

 again we find most striking proof of gradual change from one specific 

 type to another. 



In the brilliant bronze-winged pigeon (Phaps chalcoptera) of Aus- 

 tralia, we have still another combination type, in which iridescent 

 checkers coexist with the original dark centres. Here the checker 

 seems to arise by direct differentiation of a lateral portion of the 

 dark centre, the latter still occupying the original field and forming 

 the ground within which the checker appears as a more highly 

 colored spot. While the dark centre does not suffer any reduction 

 in its field, it does lose considerably in intensity of color. The metallic 

 spots are therefore probably built up by concentration of pigment 

 at the expense of the dark centres. As these birds make great display 

 of their colors in the breeding season, this departure from the ortho- 

 genetic trend of development may be attributed to natural selection. 



