170 BULLETIN 15 3, UNITED STATES NATIONAL MUSEUM 



example has the abnormal number of 11 * * *. It is thus evident 

 that the birds from Ruwenzori, though differing in no \\'ay in 

 plumage from specimens from Shoa, show symptoms of becoming 

 gradually differentiated into a distinct form." I have examined a 

 large series from Kenya Colony and find that while the majority of 

 birds have 12 rectrices, occasional individuals have only 10 (i. e., 

 naturally, not due to two feathers being shot off). 



This bird inhabits swampy ground covered with thick, luxuriant 

 vegetation and occurs only in the highlands, the altitudinal range 

 being approximately 5,000 to 13,500 feet. Von Heuglin ^^ writes that 

 it occurs on the mountains of Simien, and the high plateaus of Bege- 

 meder, Lasta, and Gallaland at altitudes of from 9,000 to 12,000 feet. 

 On Mount Elgon, Granvik found it on the fringe of the bamboo zone 

 at 9,500 to 10,000 feet, but also met Avith it at 8,000 feet near Londi- 

 ani. On Ruwenzori it has been taken from 6,000 to 10,000 feet; on 

 Kilimanjaro from 6,000 to 12,000 feet. The highest point at which 

 it has been found is 13,500 feet on Mount Karissimbi, in the Kivu 

 district of the Belgian Congo, where Gyldenstolpe obtained it. 

 Nakuru, Naivasha, Kericho, Fort Hall, and Lekiundu River appear 

 to be the lowest localities at which it occurs. 



Not much is known of the breeding season, but nestlings have 

 been taken on Mount Kenya in the middle of October. 



BRADYPTERUS ALFREDI FRATERCULUS Mearns 



Bradypterus hahacculus fratcrciilus Meiabns, Smithsonian Misc. Coll., vol. 61, 



no. 20, p. 3, 1913: Escarpment, Kenya Colony. 

 Specimens collected : 1 male, Escarpment, Kenya Colony, September 10, 1912. 



This specimen is the type of this race. 



This bird is not a race of hrachyptcrus, as considered by Sclater 

 and others, or of harratti^ to which (under the name habaeculus) 

 Mearns thought it was most closely allied, but appears to be a form 

 of alfredi nearest to the Kilimanjaro form sjostedfi, but darker. 

 Similarly, it seems that usamharae is another race of this species. 

 Furthermore it appears not impossible that nyassae may be still an- 

 other race, although I can not say without seeing material. Like- 

 wise, I can not repress a wonder whether roehli and usamharae may 

 not be the same, but again lack of material of both sexes of each of 

 the forms prevents me from deciding (types of both examined). 



The forms of B. alfredi are as follows : 



1. B. a. alfredi: Only known from the type locality (and the type 

 specimen) — Njangalo, west of Lake Albert. 



2. B. a. alhicrissalis : Only known from the type locality (and the 

 type specimen) — Mubuku Valley, eastern Ruwenzori. 



« Ornlthologie Nordost-Afrikas. etc., vol. 1, p. 275, 1869. 



