264 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



the anterior, shorter upper tail coverts pure white, forming a white 

 transverse band. The present series amply demonstrates that this 

 character is very inconstant and that it can not be relied upon as a 

 systematic character. 



Sassi ^* has confused the issue somewhat by supposing his birds 

 to be uropygialis when they are probably congicus. The characters 

 of congicus are those of the more sooty dorsal coloration, rather than 

 the rectrix pattern, a fact that Sassi appears to have overlooked. 



Inasmuch as not a few birds occurring in both Ethiopia and in 

 tropical East Africa differ in size from the Equator northward, it 

 may be stated that while Ethiopian birds average slightly larger than 

 those from Kenya Colony, the difference is so small and the overlap- 

 ping so extensive that it is not possible to divide the race into two 

 size groups. Thus, 11 male birds from Ethiopia and very high alti- 

 tudes in Kenya Colony (Escarpment, etc.) have wings measuring 

 from 91 to 101 mm, the average being 94.4 m ; 14 males from south- 

 ern Kenya Colony have wings of from 87 to 96 mm in length, aver- 

 aging 91.5 mm. If the two groups were separated, the average 

 specimen of the southern aggregate would be indistinguishable from 

 the northern form. It follows, then, that such splitting would be 

 impossible. The size variations of the present series are given in 

 table 52. 



The present subspecies occurs from Eritrea, Bogosland, all of 

 Ethiopia, Kenya Colony (except the northern coastal strip), eastern 

 and central Uganda, all of Tanganyika Territory (not including 

 Ruanda), Mozambique, Nyasaland, eastern Rhodesia, Swaziland, Zu- 

 luland, and Natal. It does not appear to have been recorded from 

 southern Somaliland, and but a few times from British Somaliland. 

 The only "Somali" record given in Shelley's Birds of Africa -^ is 

 not a Somali record at all, but one from Kikuyu, Kenya Colony. 



Its absence from the low, arid Somali region corroborates Blan- 

 f ord's observations ^® that it is very common in the highlands of 

 Eritrea and rare in the lowlands of the Anseba Valley. Neumann ^^ 

 found it only in the middle and high altitudes up to 10,000 feet in 

 Shoa, and noted its absence in the deep, hot valleys. He found it 

 chiefly in the bushy growth around the edges of the forests, and also 

 in more open country on the mountainsides. Erlanger ^® found it 

 abundant in Ennia and Arussi Gallaland, where it was often seen 

 in the cultivated plots of the natives. 



" Ann. naturbist. Hofmus., Wien, 1925, p. 22. 



« Vol. 5, pt. 2, p. 249, 1912. 



28 Observations on the geology and zoology of Abyssinia, etc., p. 338, 1870. 



=T Journ. fur Orn., 1905, p. 227. 



2« Ibid., p. 700. 



