300 



DR. MEREDITH SANDERSON — NOf ES ON 



the banks of the Songwe are of la similar nature. Here the fly was found 

 in two places only, one some miles up the valley of the Makeye in a belt 

 of trees about half a mile from that river, and the other on the Songwe in 

 the long grass. The latter habitat is a very exceptional one, and only two 

 specimens were seen on this occasion. 



Beyond the Makeye the hills became bigger and the Songwe is fed by 

 torrents, while the banks of the river are impassable owing to long grass, 

 scrub, etc. (4). No fly was found in these hills, nor at the few places where 

 it was possible to approach the Songwe. Near Mugofi Hill the Songwe was 

 carefully searched for some distance. Above this the river runs through 

 mountainous country, off*erIng no shelter to the fly^ and none was found. 



Fi- 1 



Maweiela Stream, Noitli Nyasa (see Map). The course of the stream is indicated 

 by tlie narrow line oi busli along its banks, the adjoining- country being open and 

 covered with long grass. 



The higher reaches of the Makeye, Chungu, Kaporo, and Lufira rivers 

 were then searched, but without success, presumably owing to the absence 

 of trees and to the elevation. With regard to the latter, no very definite 

 conclusions can be drawn, as in one instance, on the Karambu stream (5), 

 G. fusca was found in considerable numbers near the top of a hill some 

 700 feet above Songwe level, here about 1700 ft. above sea-level, — at a height 

 therefore of about 2400 ft. ; this was about 3 miles distant from the Songwe. 



The fly is on the wing and is desirous of feeding at two periods of the day 



