294 MISS NEWBIGIN ON THE [Feb. 18, 



velvety black. The black region extends further up the vane on 

 one side of the rachis than on the other. In mounted specimens 

 the overlapping of the quills occurs in such a manner that the 

 successive feathers cover over the black region of the feathers in 

 front. There seems to be no difference in the amount of pigment 

 present in the two regions. Eemembering the, as yet, unexplained 

 fact that metallic colours occur only on exposed parts of feathers, 

 it seems impossible to doubt that the black region is associated 

 with the overlapping of the feathers. The tail is forlted, and in 

 consequence it is hardly probable that the reetrices can be 

 separated and spread out to form the " parachute "-like structure 

 which has been described in other forms ; the black area is thus 

 permanently covered up. 



In Sappho sjiargamtra the conditions are almost reversed, the 

 exposed parts of the tail-quills are brilliantly metallic, with a band 

 of black at the apex. It seems most probable that in this case, as 

 in Topaza pella, the black colour is produced by an excess of 

 melanin pigment. 



Summary. —From the account given above it is seen that in the 

 metallic feathers of Sun-birds the radii are without cilia and their 

 distal regions are more or less completely modified into flattened 

 club-shaped bodies containing a large amount of brownish pigment, 

 and consisting of a series of "overlapping compartments," as 

 described by Gadow. It is also seen, however, that the barbules 

 near the apices of the barbs tend throughout this family to become 

 modified in this way, so that the distinction between male and 

 female, or specialized and unspecialiiied forms, is less in the struc- 

 ture of the barbules than in the amount of brownish pigment 

 present. From the fact that similar conditions are observed in the 

 Pheasant it seems probable that this occurs not infrequently. 



In the case of the Humming-birds the metallic colour is confined, 

 like the brownish pigment, to the proximal part of the barbules, 

 and its presence is not associated with any modification of the 

 barbules which affects their efficiency in the feathers of flight. In 

 their case, therefore, metallic colours may occur in the quills of the 

 tail or (rarely) wing without interfering with the powers of flight. 

 Further, as the metallic colour is associated with a specialization of 

 the proximal region of the barbule, it cannot in contour-feathers 

 primitively form a terminal band ; the apparently terminal position 

 is produced by the gradual suppression of the apical barbules, in 

 which the proximal region is never well-developed. 



The examination of specimens upon which this paper is based 

 was chiefly carried on in the Museum of Science and Art, Edinburgh, 

 and I have to record my obligations especially to Mr. Eagle Clarke 

 for his kindness in affording me facilities for doing so, and espe- 

 cially for furnishing me with several feathers for microscopic 

 investigation. 



The systematic part of the paper is based on Shelley's 'Monograph 

 of the Sun-birds' (London, 1876-80), and Salvin's Catalogue of 

 Humming-birds (B. M. Catalogue of Birds, vol. xvi. 1892). 



