ARTIFICIAL COLOURING OF BIRDS 269 



all traces of the white vanish. Coincident with this intense 

 blackening of the plumage there occurs an increased develop- 

 ment of black pigment in the choroid coat of the eye. Finally, 

 when the concentration of the dark pigment has attained a 

 certain stage, a change occurs in its colour, which is transformed 

 from dull dark brown or black to brilliant iridescent bronze 

 or green, which reaches its maximum development on the 

 wing-coverts and inner secondary quills, or just in those regions 

 where iridescence of a similar type is met with in many genera 

 of tropical and subtropical doves. All these remarkable changes, 

 it is important to mention, take place by continuous variation, 

 there being no colour-change unaccompanied by a moult. 



Some similar changes have been observed to take place in 

 the case of the yellow-rumped weaver-finch (M it nia flaviprymmis) 

 of Australia, when kept in confinement, the throat and upper 

 part of the breast developing dark mottlings similar to those 

 which occur in a more marked degree in the chestnut-breasted 

 weaver-finch {M. castaneithorax), accompanied by a darkening 

 of the sides of the head. In this instance it has been suggested 

 that the yellow-rumped bird is merely the desert phase of the 

 chestnut-breasted species ; and that confinement in a relatively 

 more humid atmosphere induces reversion towards the original 

 type. 



As regards the conclusions to be drawn from these experi- 

 ments, Mr. Beebe points out in the first place that the 

 distribution of the pigeons in America indicates the group to 

 be essentially a tropical one in that continent. Hence, in the 

 case of the Inca dove the alteration in colour induced by 

 humidity is a reversion to the ancestral type. Further, in the 

 case of the individual Inca doves, it is clear that the apparently 

 atavistic characters are acquired. This leads to the question 

 whether, if these melanistic doves were to breed, would the 

 acquired characters be inherited ? 



Another point which may be raised is that the experiments, 

 so far as they go, indicate that natural selection has nothing to 

 do with such phenomena as protective coloration. This, however, 

 the author thinks, is not borne out by the facts ; for even 

 among the four or five individuals upon which the experiments 

 were made, there was a noticeable difference in the response 

 to the abnormal conditions. 



The next question is to what extent, if any, these experi- 



