1901.] EVOLUTION OF PATTERN IN FEATHEKS. 323 



being that all the bars are disconnected. The Woodpeckers are from 

 this point of view a comparatively low form : in the young of Pious 

 viridis very incomplete bars are seen, while in Dendrojiicits minor 

 the longitudinal streak of an early type persists in the adolt. I am 

 farther inclined to believe that the dull dirty white of the chin 

 and throat represents the primitive patternlcss feather, but am not 

 in a position to express a definite opinion on the point at present. 



Having already dealt with the Owls and Hawks, the next group 

 to claim our attention is that of the Steganopodes. In the 

 Cormorants we find a type of pattern with which we hitherto 

 have not had to deal, namely, a feather with a darker margin. At 

 first I was inclined to regard this as merely an exaggerated form of 

 the crescent, but an examination of the young bird speedily 

 dismissed that idea, as in it the pattern is identical with the adult; 

 whereas if it were, as I imagined, an exaggerated crescent, one ought 

 undoubtedly to see traces of a light margin. An examination 

 of a young Cormorant, in which the feathers are whitish with 

 a longitudinal stripe, showed that the darker marginal border 

 could be distinguished even in the case of the white feather. 

 Green and metallic colours are almost always due, not to a 

 difference of pigment, but to a condition known as " surface 

 structure " due to ridges on the surface of a feather breaking up 

 the colours into their component parts, and acting by interference 

 so that only rays of a certain colour reach the eye 1 . Here, then, 

 is the explanation of this marking. The feather itself is a purely 

 self-coloured one, but round the margin, where the feather is moi*e 

 broken, the true colour of the pigment, which has been obstructed 

 by the process of surface-structure, is able to show itself and forms 

 the darker margin. 



While the genus Phalacrocorax has been evolving in this 

 manner, the G-annet (Sulci) has taken another course, and by 

 proceeding along the line we are more particularly studying, has 

 so far evolved as to have reached a secondary white stage. It is not 

 to see the stages between the dark brown semi-adult birds 

 and the pure white adult : great irregularity is noticeable in birds 

 assuming the white plumage, as regards the tracts in which it 

 is first assumed (although the back, wings, and tail are invariably 

 tin: last); but what is perhaps of most importance to us is the 

 hat in the parti-coloured feathers, of which there are many, 

 we lose all trace of the patterns which we have hitherto been able 

 to refer to a common origin, however different the general appear- 

 ance may have been. 



In the immature bird, however, matters follow what we may now 

 perhaps call the norma] course; and in its first plumage it starts 

 :it a high level, by the Eeathers being apparently self-coloured with 

 A-shaped white tips as iii the Nutcracker or young Thrush. 



Of the Ardeidae I have only had opportunity of studying a lew 

 commoner types, and they all seem to have very similarly 

 marked feathers, which, when not self-coloured, consist of a 



G a<wr, r Z 8. 1889, p. 240. 



_' I ' 



