Changes of Plumage. 559 



which the feathers are moderately lengthened, and with the 

 web narrowed in the middle ; the third stage is shown by a 

 specimen which has part of the midrib bare, and terminated 

 by a spatulate web ; in another the bare midrib is a little dila- 

 ted and semi-cylindrical, and the terminal web very small; in 

 a fifth, the perfect black horny riband is formed, but it bears 

 at its extremity a brown spatulate web, while in another a por- 

 tion of the black riband itself bears, for a portion of its length, 

 a narrow brown web. It is only after these changes are fully 

 completed that the red side-plumes begin to apjjear. 



The successive stages of development of the colors and 

 plumage of the birds of paradise are very interesting, from the 

 striking manner in which they accord with the theory of their 

 having been produced by the simple action of variation, and 

 the cumulative power of selection, by the females, of those male 

 birds which were more than usually ornamental. Variations 

 of color are of all others the most frequent and the most 

 striking, and are most easily modified and accumulated by 

 man's selection of them. We should expect, therefore, that the 

 sexual differences of color would be those most early accumu- 

 lated and fixed, and would therefore appear soonest in the 

 young birds ; and this is exactly what occurs in the paradise 

 birds. Of all variations in the form of birds' feathers, none 

 are so fi'equent as those in the head and tail. These occur 

 more or less in every family of birds, and are easily produced 

 in many domesticated varieties, while unusual developments 

 of the feathers of the body are rare in the whole class of 

 birds, and have seldom or never occurred in domesticated 

 species. In accordance with these facts, we find the scale- 

 formed plumes of the throat, the crests of the head, and the 

 long cirrhi of the tail, all fully developed before the plumes 

 which spring from the side of the body begin to make their 

 appearance. If, on the other hand, the male paradise birds 

 have not acquired their distinctive plumage by successive va- 

 riations, but have been as they are now from the moment they 

 first appeared upon the earth, this succession becomes at the 

 least unintelligible to us, for we can see no reason why the 

 changes should not take place simultaneously, or in a reverse 

 order to that in which they actually occur. 



What is known of the habits of this bird, and the way in 



