176 



BULLETIN 153, UNITED STATES NATIONAL MUSEUM 



the plateau between Muhavura and Mgaliinga, and between 

 Mgahinga and Sabinio. Van Someren * has found it breeding at 

 altitudes from 3,400 to 9,000 feet in Kenya Colony, The highest 

 altitude from which the species has been recorded is 13,000 feet 

 (3.900 meters) above sea level on Mount Elgon." Granvik^ writes 

 that it, " * * * is a pronounced highland bird, which even 

 occurs in the alpine region where only a few icy cold mountain 

 streams can offer the bird the required possibilities of existence." 

 While on the Smithsonian African expedition under the late Col. 

 Theodore Roosevelt, Mearns collected an immature specimen at 

 12,500 feet (3,700 meters) on Mount Kenia. The British Ruwen- 

 zori expedition did not meet with it on Ruwenzori, but at 3,500 feet 

 (1,050 meters), 80 miles west of Entebbe, and at Bosoko, Upper 

 Congo River, 1,500 feet (450 meters). This last record can hardly 

 be a breeding bird, as the altitude is too low. 



The breeding season in Ethiopia is during June, July, and 

 August. 



The light streaks on tlie upper parts vary from tawny yellowish 

 white to pale cinnamon. Two birds (unsexed) from South Africa 

 represent the two extremes in this regard. Neumann - records a 

 similar case in two males from Adis Abeba. The two South African 

 birds differ from the two Ethiopian specimens examined in that the 

 former have much broader white terminal bands on the under wing 

 coverts, giving these feathers a white appearance as they lie in 

 overlapping rows. If this character should be found to hold good 

 and not be due to freshness or wear of the plumage, the southern 

 birds may require separation. 



Younger birds have the upper parts duskier, and less brightly 

 marked, and the under parts more streaked with smaller pure white 

 areas, than older individuals. 



This species and Capella media are so similar superficially that it 

 is quite possible that some of the published records are erroneous, 

 but inasmuch as the two species occur together only during the 

 northern winter, the chances for error are considerably reduced. 



The four specimens examined measure as follows : 



2Journ. f. Ornltb., 1904, p. 332. 

 *Nov. Zool., vol. 29, 1922, p. 20. 

 ^Granvik, Jouru. f. Ornitli., 1923, Sonderheft, p. 39. 



