104 BIRDS OF SOUTH AFRICA. 
noise, rapidly uttering harsh and discordant sounds: its notes are at 
times not unlike the sound produced by a broad-bladed knife passing 
through a tough piece of cork, but in a louder key. This species 
seeks much of its food on the ground ; but sometimes it watches from 
some elevated position, and, the moment its prey comes within sight, 
darts upon it with unerring certainty, its habit in this respect being 
very much like that of the Butcher-bird Shrikes. It is a most useful 
bird, feeding largely on centipedes, scorpions, tarantulas, and other 
insects, as well as on small snakes and lizards,” 
“This Roller is one of the earliest breeders in Damara Land, and 
makes its nest in the hollows of trees, usually such as have been pre- 
viously occupied by some Woodpecker ; the stems of these trees are 
mostly very tall and straight ; and in consequence of this and of the 
smallness of the apertures, the nests are very inaccessible. I have 
frequently seen such breeding-places without being able to reach 
them ; but I believe that the eggs are white and two in number, 
and that both parents assist in their incubation.’’_| 
General colour reddish-brown glossed with olive-green; under 
surface dull reddish-brown, streaked with white; wings dull rufous 
green, with purple reflexions; the primaries dark blue; tail ultra- 
marine, the middle feathers olive-green ; forehead and chin white, as 
well as a distinct patch on the nape; cheeks rusty red with white 
streaks to many of the feathers ; bill black; legs and tail greenish 
or yellowish brown ; iris varying from dark or hazel brown to oliva- 
ceous brown. ‘Total length, 13 inches; culmen, 1:5; wing, 7°4; 
tail, 5°6; tarsus, 0°8. 
Fig. Wevaillant, Rolliers, pl. 29. 
100. CoractAs CAUDATA. Lilac-breasted Roller. 
This species inhabits the same localities as C. garrula, but like 
that species, it has not fallen under our personal observation. Mr. 
R. Moffat, jun. has forwarded several specimens from Kuruman, Ver- 
reaux from Kurrichaine, and Mr. Ayres has found it in Natal. The 
same gentleman has met with it on the River Limpopo. Mr. T. E. 
Buckley says that it was first observed by him a few days north of 
Pretoria and found far up into the Matabili country. “ This species,” 
he remarks, “was extremely shy and it was difficult to procure 
specimens. Just before the breeding-season they fly high up in the 
air, rolling about from side to side, and uttering a harsh note all the 
