Notes on Crested Birds of Prey. 205 



afford good examples. Another type is that of the loose 

 bunch of spatulate feathers (PL X., Fig. 7) that decorate Ser- 

 pentarius, tlie Secretary P>ird. This form of crest can be 

 spread out radially from the place of insertion, so as to form 

 a wide crown (PI. X., Fig. 7«). A third development is tliat 

 of tlie loose tuft of elongated feathers that decorates Loplio- 

 aetns, the P)lack Crested Eagle of South Africa (PI. X., 

 Fig. 8). In many individuals of this species the tuft in 

 question is habitually inverted, and consequently droops 

 forward so that the tips of the longest feathers nearly touch 

 the crown of the head. A fourth type of occipital crest is 

 represented by the neat, smart, pointed tuft seen in Baza, 

 the normal Spizaeti (PL X., Fig. 9), Haipylialiaetus (PL X., 

 Fig. 10), and some others. 



Eacli of these styles of ornament may pass into, or may be 

 combined with, any of the others ; so that living specimens 

 of the crested Birds of Prey present a v/ider range of variety 

 in this respect than do the others. 



So far the facts. Can we discover any reasons for them ? 

 The features in question appear to me not to be due to sexual 

 selection, because the decorations are not confined to the 

 males, but are, in general, common to both sexes. That the 

 crested condition is not an independent development in each 

 species, connected with its own special need, is, I think, 

 evident from the fact that the immature forms of many 

 species are often as much crested as the adults ; and indeed, 

 in some species, appear to be even more so. This fact is the 

 more striking, because many of these crested forms of Birds 

 of Prey (unlike the majority of the remainder) pass through 

 several changes of plumage before tliey attain to the mature 

 form. Spizaetus ornatus, for example, passes through at 

 least four successive changes of plumage before it becomes 

 adult, and yet it is crested through them all. 



It appears to me, therefore, that we are fully justified in 

 regarding the possession of a crest as a character older than 

 the species. But all the species of certain genera {Spizaetus, 

 for example) are crested in one way or another. It is 

 extremely unlikely that all the species of a genus, widely 

 dispersed over the face of the earth, living under a great 



