BIKDS OF ETHIOPIA AND KENYA COLONY 245 



Occasionally specimens best identified with ferreti are found as far 

 south as the Athi River. This race is like viridis, but, in the great 

 nvajority of cases assumes the white-backed, white-tailed plumage 

 in the second (not the third) year, so that long-tailed (i. e., adult) 

 brown birds are scarce, 



4. T. V. harterti: Southwestern Arabia (Yemen district). Simi- 

 lar to ferreti but wdth noticeably larger bills (in males only), 

 measuring 20 mm as against 18 mm in the latter. 



5. T. V. suahelica: Southeastern Uganda, the Sotik and southern 

 Kikuyu and Ukamba districts east to the Taveta Forest in Kenya 

 Colony, south through Tanganyika Territory intergrading with 

 perspicillata in the valley of the Rovuma River. This race never 

 assumes the white plumage found in the above three, but has white 

 edges on the secondaries in adult birds. 



6. T. V. perspiciUata: South Africa from Cape Town east to Pondo- 

 land, north to Natal, the Transvaal, Zululand, Swaziland, Nyasaland, 

 and Mozambique, merging with suahelica along the Mozambique- 

 Tanganyikan boundary. Occasionally specimens of perspicillata are 

 found in central Tanganyika Territory, the northernmost locality 

 Iniown to me being Bagilo in the Uluguru Mountains, but such cases 

 are uncommon. This race never gets any white, even on the edges 

 of the remiges. I think Sclater is wrong in considering this form 

 specifically distinct from viridis,' and plurribeiceps is clearly a dis- 

 tinct species from both. 



Of the long-tailed birds in the present series three are brown- 

 backed and brown-tailed, five are brown-backed and either wholly 

 white-tailed or with brown outer and white inner rectrices, and only 

 two are white-backed and white-tailed. The five with white tails 

 and brown backs have much white on the wings and are freshly 

 feathered except on the brown backs. It therefore appears that in 

 molting into the white plumage the tail is affected before the back. 

 I have seen no specimen of this or any other race in which the reverse 

 is true, i. e., with a white back and brown tail. The white rectrices 

 vary in that some have black shafts (and even in one case broad 

 shaft streaks) while others are wholly white. 



That the species breeds in its second year is shown by the fact that 

 one of the brown-backed, long white-and-brown-tailed males was 

 observed with its mate and nest and young at Gato River. 



Males have wings of from 78 to 88 mm in length, females from 

 74.5 to 85 mm. The tails in adult males (with fully developed 

 elongated middle rectrices) vary from 320 to 400 mm, the extent of 

 the central pair of rectrices beyond the ends of the lateral ones 

 measuring from 220 to 294 mm. 



A nestling male, taken from the nest on May 1 at Gato River, is 

 partly clothed in the juvenal plumage — black crown and occiput, 



106220—37 17 



