BIRDS OF ETHIOPIA AND KENYA COLONY 315 



of eastern and southern Africa from Kilimanjaro southward. 

 Reichenow - lists africanus from Keren, Anseba, Mahal, Uonz, Allan, 

 Dembi, and other Ethiopian localities. Even allowing for several 

 errors in identification, and assuming that some of these really refer 

 to migrant typical European rrielba^ it is not likely that all of them 

 are wrongly identified. The present specimen is undoubtedly afri- 

 canus. as it is ver}^ dark (darker than any of a small series of South 

 African birds), and has narrow dark shaft stripes on the white 

 throat feathers, thereby agreeing with afHcaniis and differing from 

 melha. However, the white gular patch is as extensive as in many 

 individuals of the typical race, more so than in typical africanus. 

 Further material may suggest that africanus is divisible into a 

 northern and a southern form, but at present this is not feasible. 



Of the locality records mentioned by Eeichenow, we may strike out 

 the Anseba Valley, as this is based on Blanford's notes,^ which read 

 as follows : 



This bird was not noticed on tlie highlands. In the Anseba Valley many 

 appeared at the end of July, none being seen before the 25th. Some of the 

 spechneus shot had pale edgings to the feathers, and were evidently young 

 birds ; others were of uniform color. 



It will be seen that these birds were found in the lowlands, whereas 

 africanus is a highland form, and they were also apparently mi- 

 grants from the north, and were probably typical Micropus iiielba 

 inelha. 



The altitude of the Gato River near Gardula is about 4,000 feet 

 (1,200 meters), and as far as the few published notes indicate, this 

 would be about the lower limit of the range of africanus. Zedlitz * 

 records it as by no means abimdant in the mountains of Arussi- 

 Gallaland. On Kilimanjaro Sjostedt ^ records it at altitudes of 

 from 5,000-10,000 feet (1,500-3,000 meters). On Ruwenzori, it is 

 replaced by a much larger form, maximus Grant, which occurs 

 from 10,000-14,000 feet (3,000-4,200 meters). It does not appear to 

 have been recorded from Mount Elgon. 



The single specimen collected agrees in size with two from South 

 Africa, having a wing length of 203 millimeters. 



Recently Hartert ^ has described a race of this swift from British 

 Somaliland, which he has named archeH. It is said to be pale in 

 color (like tmieti of Tunis) but small, especially as regards the wing 

 and tail, the former varying in length from 195-207 millimeters. It 

 may be noted that the present specimen comes within the limits of 



= VOjr. Afr., vol. 2, p. 377. 



^Geol. and Zool. Abyss., 1870, p. 3:!5. 



* .Touin. f. Oniitli., 1010, p. 782. 



" Kilimanj.irn-Mcru Exp., Vog., 1910, p. 102. 



•Nov. Zool., vol. 34, 1!)U'8, p. 363. 



