227 



Head heavily chitinized, dorsum with the usual 2 longitudinal dor- 

 sal excisions and a smaller median posterior one. Antennae short and 

 slender. Labium heavily chitinized, triangular in outline, margin den- 

 tate. Mandibles strong, curved, their inner margin dentate. Body 

 slightly flattened dorso-ventrally, the segments distinct, with decum- 

 bent pale pile, and having long bristle-like hairs on lateral margins of 

 each segment, 1 on the prothorax and 2 on each of the other seg- 

 ments. Abdominal segments 2-8 each with a narrow transverse fusi- 

 form stripe on ventral and dorsal surfaces near the anterior margins 

 which is armed with short spinules. Apical segment tapered, cleft, 

 the margins of the cleavage with 2 upper and 2 lower processes, each 

 pair margined with fine hairs. 



Pupa. — Length, 6.5 mm. Yellowish brown, the abdomen green- 

 ish white. 



Thoracic respiratory organs about as long as diameter of thorax, 

 very stout, their bases almost contiguous, tapering from base to apex, 

 and more or less resembling the pincers of a crab. Abdomen armed 

 as in larva except that the lateral hairs are wanting. Legs extending 

 to base of antepenultimate abdominal segment. Apical segment pro- 

 longed slightly in both sexes, that of the female a trifle the longer, a 

 few small processes present in both sexes at base. 



This genus agrees well in the larval and pupal stages with the cor- 

 responding stages of Dicranomyia, the distinctions between them be- 

 ing less marked than is the case with allied genera of some other sub- 

 families. 



Subfamily ERIOPTERINAE 



Helobia and Gnophomyia are the only genera of this subfamily 

 of which identified larvae and pupae are before me. I have, however, 

 an unidentified larva that quite obviously belongs here. There is a 

 great similarity in these larvae, but judging from the available descrip- 

 tions of European species of other genera a great difference exists 

 between the forms I have and those of other European genera. The 

 description of the larva of Trimicra agrees with the characters gen- 

 erally attributed to larvae of Pediciinae — a fact that to my mind 

 throws considerable doubt upon the correctness of the present sub- 

 family-grouping, which is based upon characters of the imagines. I 

 have no intention of rearranging the genera in this or any other sub- 

 family upon the basis of characters deduced from printed descrip- 

 tions, and accordingly leave the subfamilies practically as in Willis- 

 ton's "Manual", but consider it essential to indicate the probability of 

 errors in the arrangement. 



