368 ZOOLOGY. 



it is this difference of plumage in the two sexes and the 

 colouring of the female which tend to unite this bird so 

 closely with Grandala ccelicolor. 



From December up to the end of February this bird 

 was entirely absent from the passes below Senaf^. About 

 the commencement of March flocks began to make their 

 appearance, and the numbers gradually increased. The 

 flocks consisted of males and females. In May, when I 

 returned from the highlands, I found these lovely birds 

 most abundant, and almost always in pairs. They 

 occurred from about 2,000 feet elevation up to 5,000 or 

 6,000 feet. I again met with them in theLebka valley 

 at very nearly the same elevations, one bird being seen 

 rather lower down at Ain ; but individuals were scarce 

 below about 2,000 feet. On the Anseba it was occasion- 

 ally seen, all the birds noticed being males. The females 

 were probably in July engaged in incubation. I could 

 never find the nesffc. 



Grandala leucogaster keeps much to trees and high 

 bushes, but is occasionally seen on the ground. Its prin- 

 cipal food appears to consist of seeds. Dimensions : — 



Wing. Tail. Tarsus, 



in. in. in. 



Male 4-2' 275 0-9 



Female 3-9 2-4 0-9 



The male is perhaps the handsomest of all the birds 

 met with in Abyssinia. The play of colours in the sun- 

 light from pure azure to deep crimson, and again to 

 copper colour, is most exquisite. Brehm is incorrect in 

 sajdng that the iris is brown. 



