Expedition to the Zambesi River. 99 



decorative appearance, and reminds one of those types of 

 birds that the Indian loves to depict upon his choicest 

 silks. 



Towards evening the bird wakes up from its lethargy and 

 takes up a post close to the riverside, \vhence it skims over 

 the smooth water after insects, often striking the surface so 

 as to send up a cloud of spray. 



On November 1, not far from Zumbo, we came across 

 a nesting-station of this species. The holes, which were 

 very numerous, were in a steep portion of the river's bank. 

 On nearing the colony, these lioles began to bristle with 

 long pointed bills, then feathered heads were poked out, 

 a movement soon followed by a general exodus of these 

 birds, which, with their peculiar crying notes, thronged 

 out to the river in hundreds and tlien up into the blue sky, 

 circling round and round like a flock of Rooks, the light, as 

 they passed and turned, playing upon their brilliant plumage. 

 It was a fine sight and not easily to be forgotten. 



These birds were breeding. The holes ran very deep into 

 the bank, about 3 feet in, and w^e had no appliances for 

 digging them out. 



The last important observation on this species from my 

 diary runs as follows : — " December 23, 60 miles from the 

 Kafue river. Country open, hills receding much from the 

 river. Large numbers of Bee-eaters [M. natalensis) flying 

 high overhead, from time to time stopping to circle after the 

 manner of Rooks. They must be on migration, since a high 

 flight is always then resorted to ■" 



136. IJPUPA AFRICA NA (Ecclist.). 



Seen occasionally in pairs, from November onwards, on 

 the higher reaches of the river. 



137. Irrisor viRiDis (Licht.). 



Generally observed in small flocks, but not common. 

 Moults in August. By the end of October breeding had 

 commenced. The feathers then had a strong and disagreeable 

 odour. 



h2 



