176 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



occur together, an argument in favor of his action in calling 

 flavocincta a sjDecies. 



The material available to me for study includes only 6 of the 11 

 proposed races (one of which, neumanni, is questioned by Sclater, but 

 which I have not seen and therefore can not judge), and, as the total 

 series comprises less than 30 specimens in all, I follow Sclater's 

 arrangement.^^ 



Gyldenstolpe,^^ however, upholds Zedlitz in deciding that neii- 

 manni is a valid race, and it may be that Sclater is wrong in this 

 regard. 



In the general region involved in the present report five races (six 

 if we count neumanni) are found, as follows : 



1. A. f. golzi: The coastal districts of extreme southern Kenya 

 Colony and northern Tanganyika Territory, inland in the latter 

 country to Dodoma and the Unyamwesi district. This form enters 

 into the present report because it occurs at Mombasa, the terminus 

 of the African wanderings of the Frick expedition. 



2. A. f. aequatorialis : Southwestern Kenya Colony from near Mau 

 and the Sotik district and around Lake Victoria. 



3. A. /. -flavocincta: Central Kenya Colony from the Lekiundu 

 River and Mount Kenya to Kikuyu, Ukamba, the Athi River, the 

 Taru Desert, and the plains east of Kilimanjaro, intergrading in the 

 Sotik area with aeqiiatorialis. 



4. A. f. mulensis: Extreme southern Shoa (possibly Gallaland 

 also) south through the Rendile district to the Northern Guaso 

 Nyiro River and Lake Baringo in Kenya Colony. 



5. A. f. viridiceps : British Somaliland (none seen by me). 



If neicmatini is a synonym of viridiceps, the range of the latter 

 will have to be extended south to Afgoi, in southern Italian Somali- 

 land. Sclater suggests the identity of these two but restricts the 

 range of viridiceps to British Somaliland. Reichenow,^^ on the other 

 hand, appears to consider southern Somaliland birds as malensis. 

 If this is true, then the range of malensis will have to be extended 

 east through Gallaland to southern Somaliland. The fact that there 

 are three divergent opinions about neumamii (that it is distinct; that 

 it is a synonym of viridiceps ; and that it is the same as malensis) 

 makes me wonder whether it may not really be distinct, or if the 

 distinctness of some of the other forms is not due more largely to age 

 and seasonal differences than to real racial characters. Certainly, 

 the form aequatorialis is very close to golzi, and malensis is only 

 doubtfully distinct from flavocincta. 



» Systema avium ^thioplcarum, pt. 2, pp. 523-524, 1930. 

 s^Arkiv for Zool., vol. 19A, no. 1, p. 49, 1930. 

 »«Die VOgel Afrikas, vol. 3. p. 612, 1905. 



