212 Mr. F. W. Edwards 07i 



association of tliese specimens is largely conjectural, but 

 there seems to be no other described species with which 

 they can be identified. The antennae are 8-segmented, but 

 the segments are not very well separated, the 7th in par- 

 ticular being small and not clearly marked oft' from tiie 8th, 

 tliough distinguishable b}' its hair-whorl. The second seg- 

 ment is rather large, stouter than the following, and rather 

 sharply constricted on the dorsal surface near the base. 

 (Segment 8 is nearly as long as 4-7 together. The lamellae 

 of the ovipositor (fig. 11) are very large, irregularly shaped, 

 and yellowish. The wings have a rather pronounced 

 brownish-grey tint, and as is the case in the females of 

 many species of this genus (including all those known from 

 Spitsbergen) the apical part of Ri is considerably swollen 

 and very closely ap])roximated to the costa, and the basal 

 fourth of -/^2 is darkened. 



It is probable that the ? ? from Bear Island referred to 

 D. arctica by Kieft'er in 1911 really belong here, since he 

 refers to the unusually large lamellse of the ovipositor. 



Diamesa hyperborea, Holmgren. 



Bear Island : Walrus Bay, S.E. of island, 22. vi. 1921 ; 

 20-50 ft., quarter mile inland under stones on shady 

 slope ; 1 ? . 



This specimen is much smaller than those of D. ursus 

 (wing-length 3 mm., instead of 4-6 mm.) and has the lamellae 

 of the ovipositor proportionately much smaller and quite 

 differently and more regularly shaped (fig. 12). In size and 

 most other points it agrees with Holmgren's description. 

 The antennfe are 8-segraented and not 6-segmented, but I 

 do not consider that any importance need be attached to this 

 apparent diff"erence. No species of Diamesa are definitely 

 known with the antennae 10-segmented in the cJ and 6-seg- 

 mented in the ? , and it is extremely unlikely that the only 

 two species which Holmgren described should really possess 

 tliese characters ; it is much more probable that he made an 

 excusable error in counting the segments, which are at best 

 poorly defined. 



Diamesa arctica (Bohem.). 



There is no species in the Oxford collection which can be 

 positively identified with D. arctica, but I think I can so 

 regard a male in the British Museum from King's Bay, 



