IH'J On j\I(jult mid Coloiir-clioiujc in H'lnh. 



all those variations, is no more logical than to claim that if 

 the pales at one end of a fence are painted red, and those at 

 the other end hlue, with the intervening^ ones graded in shades 

 of purple, we have proof that each one of the blue pales has 

 passed through all the intermediate shades of colour ! 



A fcAV words must he said in conclusion on the study of 

 birds in captivity. Mr. Bonhote suggests that I should 

 extend my investigations to this field. This I have done to 

 a certain extent for some years past, and in the case of 

 Dolichonyx oryzivorus, Zamelodia ludovicicma, and other 

 species in which the change is less striking, I find in every 

 individual examined that a spring moult takes place, while 

 there is no indication of a colour-change in the individual 

 feathers. 



The difficulties in the way of this method of study are 

 very great, and it is extremely hard, as Mr. Bonhote admits, 

 to study any individual feather. To my mind none of the 

 rare instances yet quoted of change of pigment in the 

 feathers of live birds are at all conclusive, since so many 

 ways suggest themselves in which observers may uninten- 

 tionally misinterpret what they see, and so many details 

 are lacking in their accounts. One argument which has 

 been quoted in support of the colour-change theory is the 

 effect of a diet of cayenne-pepper in brightening the plumage 

 of Canaries ; but the breeders with whom I have talked state 

 that no change is noticed until after the moult, and that it 

 is the new feathers which are affected. Are not the alleged 

 changed feathers in living birds merely new feathers suddenly 

 expanded from pin-feathers which had before escaped notice? 

 The presence of a few of these new feathers and a number 

 of permanently particoloured feathers would apparently 

 furnish all the proof required. Moreover, the fact which I 

 liave stated elsewhere that cage-birds moult very irregularly 

 and imperfectly, adds to the difficulty of drawing accurate 

 deductions from their study. 



It is my earnest desire that many investigators may be led 

 to pursue this branch of ornithological research, and I cannot 

 but feel that if due weight be given to the points brought 



