OF WASHINGTON. 219 



only peculiar form is the large, furcate spine of some of the Blen- 

 nocampids. These are never true warts. Occasionally there may 

 be other fine hairs scattered over the skin. The setae are distinctly 

 variable in number, as just stated, even in closely allied forms, 

 and it is somewhat surprising to find them not possessed of any 

 thing like the fixity or constancy of the primary setae of the Lep- 

 idopterous larva. They correspond in their modifications rather 

 to secondary hairs, and it may be stated that primary setae, as 

 that term is understood in Lepidoptera, are not present in the 

 larvae of Tenthredinidae. Consequently they are of much less 

 value in the. broad lines of classification, and it is interesting to 

 see what, at first sight, appears the same character of such differ 

 ent value in the two groups of insects. As compared with the 

 Lepidopterous larva, the saw-fly is much generalized. In the 

 shape of the clypeus and position of the antennas it shows less 

 divergence from the mature insect. The thoracic feet are pre 

 dominant, and the abdominal ones are simply membranous pads. 

 The hair structures are very simple. The smooth skin but rarely 

 bears processes, which do not exceed the simple papilla in com 

 plexity. Eversible glands, however, may be present, either ven- 

 trally or laterally, and the skin may secrete a pruinose or long 

 woolly coating. The coloration is of the simplest. The skin is 

 usually transparent, and the larva receives its color from the green 

 blood. The lines consist of the outlines of the dorsal vessel and 

 tracheal line, supplemented by fat granules. Spots are formed 

 by black pigment, often at the setae, but the deposition of pig 

 ment is seldom extensive, never complicated. Mimicry of special 

 objects other than the leaves it feeds on is rarely attempted by the 

 larva. I have in mind two cases, one from the Nematinae, the 

 other from the Tenthredininae (neither larva bred), where the ex 

 crement of birds is mimicked. One or two others are somewhat 

 brightly colored. 



I should like to present a classification of saw-fly larvae, based 

 on the characters just reviewed, but my material is too scanty. 

 When our species are better known it may be possible to do this. 

 In the meantime I will indicate some of the differences between 

 the different groups as seen in a few examples. 



HYLOTOMIN^E. 



Hylotoma macleayi. No feet on segment 8 ; segments divided 

 into three annulets, each with three tubercles above the tracheal 

 line ; a large oblique subventral ridge with many setae. 



NEMATIN^E. 



Ncmatus ribesii. No feet on segment 8 ; segments divided 

 into three annulets, each with three tubercles, the anterior ones 

 sometimes partly wanting; a small cluster of hairs on subventral 

 ridge. 



