28 Zoologia: N. Y. Zoological Society. [Vol. I 



trace of this color had disappeared. Hence data for comparison 

 is confined to the rectrices. Besides the three outer pairs of these 

 feathers, no white was left on the dove except an irregular 

 whitish-gray blotch on the longest left under tail covert. 



The most interesting fact brought out by the monthly feather 

 renewal of this black dove is the reduction of the white on the 

 rectrices, instead of an increase, as was the case with the dove 

 in the third moult.* 



The most logical explanation would appear to be that ex- 

 cessive humidity causes, not only an increase but, especially dur- 

 ing the first few moults, a concentration of the melanin — result- 

 ing in a segregation of the black and white and consequently an 

 increase in area of the latter. This condition is at least 

 paralleled in wild Arizona and Venezuelan specimens (vide 

 Table I) . If, however, the humidity continues to act, and here 

 we get beyond the stage of any wild Scar da fella, a point is reached 

 where further concentration of melanin is impossible, and the 

 overproduction of granules is forced into the surrounding epi- 

 dermal structures. (Fig. 4.) 



From now on a gradual reduction, (Table VI), of the white 

 area takes place until the dark hue is everywhere predominant. 



It may here be stated that in a number of cases measure- 

 ments were made when the feathers had just broken from their 

 sheaths, and again, eleven months later, just before the annual 

 moult, and no change was apparent. This eliminates from these 

 cxj^eriments any theory of increase of melanin ivithout moidt. 



I have elsewhere recorded the facty that the pigmentation 

 of the choroid coat of the eye is, at least in owls, correlated with 

 the pigmentation of the plumage, the choroid of a dusky horned 

 owl being very strongly pigmented as contrasted with the all but 

 absence of pigment in a snowy owl. The optical fundus of a 

 small dove such as Scardafella is very difficult to observe, and 

 thus far I have been able to make out only the grosser details. 

 But in my completely melanized specimen the great increase of 

 pigment over a normally colored Scardafella inca is very notice- 

 able, and is interesting as showing that the pigmentation of this 

 part of the eye is as plastic as that of the plumage. It is, how- 

 ever, only what we should expect when we recall their identity of 

 origin; the feathers being epidermal structures arising from an 

 induration of epidermal papillae containing a vascular core, while 



* Fig. 4 shows that this restriction of the white of the rectrices is not only from the 

 proximal part of the feathers, but at the tips where thei-e has appeared an area of encroach- 

 ing black. 



t "Owls of the Neartic Region," Eleventh Ann. Rep. N. Y. Zool. Soc. 



