HYMENOPTERA. — TENTHREDINID®. 113 
separated the species with clavate antennz (or Cimbices), under the 
generic name of Crabro, which has since been applied to very ditfer- 
ent insects. Fabricius added five other genera, and Jurine four. 
Leach divided the family into nine stirpes, founded chiefly upon the 
form of the antenne, and the number of cells in the wings ; and added 
several other genera. St. Fargeau proposed another artificial dis- 
tribution, founded entirely upon these two characters. The arrange- 
ment proposed by Latreille in the second edition of the Régne Ani- 
mal appears far more natural as regards the affinities of the genera. 
The arrangement proposed by Dahlbom is confined to the species of 
Scandinavia, and that by Hartig to the German species ; so that we 
cannot gain a knowledge of their views as to the arrangement of the 
exotic groups, as Pterygophorus, &c.* By combining the different 
most nearly allied genera together, it appears to me that the family is 
divisible into the following subfamilies : — 
1. Cimbicides. — Antenne short, clavate, with not more than eight 
joints ; larvee 22-footed, emitting drops of viscid matter from the 
pores of the body. (Cimbex, Perga, &c.) 
2. Hylotomides. —Antenne 3-jointed, terminal joint greatly elongated ; 
labrum apparent; larvae 18- to 20-footed, not emitting drops 
of viscid matter. (Hylotoma.) 
The genus Athalia seems to be the connecting link between this 
subfamily and the next. 
3. Tenthredinides. — Antenne 9- to 14-jointed, simple, filiform to the 
tip; labrum apparent; saws with parallel sides. (Tenthredo, 
Nematus, Dolerus, Selandria, &c.) 
4. Lydides. — Antenne multi-articulate, sometimes strongly pecti- 
nated in the males; posterior tibize often spined in the centre ; 
labrum minute ; saws but slightly serrated at the tip, strongly 
dilated and elbowed at the base (fig. 71. 10. saw of Lyda); larvee 
various. 
The anomalous genera Lyda, Tarpa and Lophyrus agree in these last 
characters ; and I cannot but think that those afforded by the form of 
the ovipositor (which have not been previously employed in the dis- 
tribution of this family) are of primary importance ; in which respect 
* These two authors have adopted a plan which appears sto me to be likely to 
lead to much confusion ; the genera, subgenera, sections, and tribes being all named 
as genera: thus the Tenthredo melanocephala Fab. is named Tenthredo, Selandria, 
Blennocampa, Monophadmus, melanocephala, by Hartig. 
VOL. II. 1 
