74 THE EEPORT OF THE No. 36 



of the larva in bringing the glands into prominence would suggest that a fluid or 

 disagreeable odour was poured out from them. Although larvge representing several 

 species and genera were examined, no fluid or fetid odour was noted. It may, how- 

 ever, be a fetid odour perceptible only to insects. 



The larvae collected on sedge and believed to belong to the Dolerinse have on 

 abdominal segments one to seven a pair of deep oblique folds. These folds are 

 located on the segments in line with and behind the prolegs. There has been no 

 opportunity to study sections of these structures, and the mouths of the folds do 

 not appear like the mouths of glands. It has been assumed that they were glands 

 from the general occurrence along these folds of plate-like masses, which are striated, 

 grayish in colour, and project downward for a considerable distance, and have all 

 the characteristics of a glandular secretion. 



After so full a discussion of the metamorphosis, anatomy and coloration of 

 the larvge, a word should probably be said in conclusion as to their classification. 

 The larvae are divisible into four types. The first three families, Xyelidae, Pam- 

 philiidae, and Tenthredinidae, represent three of these types. They have distinct 

 thoracic legs, which would suggest some relationship. The two generalized families, 

 Xyelidae and Pamphiliidae, are clearly the most primitive in their larval as well 

 as in their adult characters, and like the adults would suggest that they have arisen 

 along different lines. The Pamphiliidae with distinctly segmented and long 

 antennae and segmented anal cerci are the more primitive so far as their larval 

 structures are concerned. The lack of prolegs is difficult to explain. The Xyelidae 

 with inconspicuous antennae and without segmented anal cerci are readily differ- 

 entiated from the Pamphiliidae, while the presence of prolegs on each abdominal 

 segment would suggest a generalized condition entirely different from that found 

 in the Pamphiliidae. From the structure of their wings the Xyelidae are unques- 

 tionably the most generalized, but from the structure of their larvae the Pamphiliidae 

 are unquestionably the most generalized. The Tenthredinidae suggest a stage of 

 specialization from the Xyelidse. They 'have a similar type of antenna, with a reduc- 

 tion in the number of prolegs. The general similarity of the habitus of the larvae 

 of the various subfamilies would suggest, like an examination of their wings, that 

 they represenjted a single homogeneous group instead of several of higher rank. The 

 Xiphydriidge, Siricidae, Cephidae, and Oryssidae are correlated together as adults, 

 they should likewise be associated from larval characters. The most striking is 

 the loss or reduction of the thoracic legs, the lack of prolegs, and the presence of 

 a caudal cuticular horn. The antenna of Cephus is similar to that of the 

 Nemiatinae. Not enough is known regarding the larvse of these later groups to 

 offer anything as to the interrelation of the families. 



