6 NATURAL HISTORY OF BIRDS. 



we hammer carefully the deep blue feathers of a macaw, the blue color immediately dis- 

 appears, and the injured part looks gray or brownish, according to the underlying pig- 

 ment. Some green parrot feathers, when treated in a similar way, become yellow, 

 since this is the color of their pigment. Here we have the explanation of the dark 

 appearance of the abraded parts of feathers of parrots and other brightly colored 

 birds. 



The gloss of feathers, independent of the color itself, is the result of their surface 

 being smooth and polished, while the metallic lustre is due to a transparent sheath 

 which acts like a prism, a fact ascertained by Mr. Gadow. The theory of the metallic 

 lustre being due to structure of a prismatic nature originated, however, with Professor 

 B. Altum. 



We mentioned above that the seasonal shedding of feathers or of their edges 

 usually causes a change in the color of the plumage. In some birds we distinguish 

 summer and winter plumages, in others nuptial and post-nuptial garbs, and in some 

 ptarmigan may be observed even four more or less distinct attires nearly corresponding 

 to the four seasons. 



There are also some interesting relations connected with the similarity and dissimi- 

 larity in color between the two sexes, and between the adults and the young. Though 

 it might seem to be the original arrangement, or perhaps just, therefore, young birds 

 and the adults of both sexes and at all seasons are comparatively seldom quite alike. 

 The Procellaridw, or petrels, may be quoted as an example, besides several others. If 

 the adults of both sexes, for some reason or another, haA r e developed alike seasonal 

 colors, the first plumage of the young is very often like that which the parents assume 

 about the same time, that is to say, their post-nuptial or winter dress. In such a 

 case the young birds undergo a change in the spring similar to that of the old ones ; 

 many of the auks (Alcidoe) demonstrate this rule. Whenever one of the adults, no 

 matter what sex, is more richly colored than its mate, the young usually resemble tin- 

 more plainly colored of the parents ; this rule is followed by a great many, perhaps 

 the majority of birds, but exceptions and many modifications occur. We are, how- 

 ever, justified in making this generalization, that species in which both parents differ 

 materially from the plumage of the young are still more specialized as to color than 

 the foregoing categories ; for we may without hesitation take for granted that the 

 plumage of the young is the more generalized, and that the amount of specialization 

 is in proportion to the departure from the first garb. It follows that we have to go to 

 the birds in the later plumage, or in that more like it, whenever we wish to ascertain 

 the relationship of different forms. It will, therefore, be necessary to arrange the 

 species according to the characters furnished by the young, or plain-colored females, 

 and not by the secondary, often highly specialized, structure of the males, if we aim 

 at a natural classification based upon affinities. It will seem as if there may be a 

 possibility of finding out the relation between the different classes of plumages, so 

 that it might be deduced whether one kind of plumage in a given case for instance, 

 a barred or spotted one is a more specialized condition than another, say a striped or 

 plain dress ; but no investigations, covering a sufficient number of species of all orders 

 and from all parts of the world, have been made as yet, without which all generaliza- 

 tions and speculations are premature and next to valueless. 



Finally, we have to consider a color problem which has only come forward of late, 

 and which still awaits its solution. There has been invented a name for the phenom- 

 enon, and we are accustomed to call it dichromatism, but of its true nature and its 



