CLASSIFICATION. 61 



objective is the best ; if lower powers are used some of 

 the details are apt to be overlooked. The mouth 

 organs and other portions of the body can be mounted 

 in the same way. 



The larvae may be reared in the customary methods 

 followed by lepidopterists. Owing to so many of them 

 remaining over the winter as unchanged larvae, they 

 are not always easily reared, but no special difficulties 

 are met with in dealing with the summer broods, 

 which pass rapidly through the larval and pupal stages. 



They may be preserved for the cabinet by holding 

 them, after being placed in a pill box which is enclosed 

 in a tin canister, over the flame of a paraffine lamp or 

 over the gas for a minute or two, when they will 

 become perfectly hard. If proper care be taken, fairly 

 satisfactory specimens may, by this plan, be obtained, 

 provided that only fully grown specimens, with empty 

 food canals, are operated upon. The preserved larvae 

 are perhaps kept best on pins stuck through cardboard, 

 another and stronger pin being stuck in this, and by it 

 kept in position in the cabinet. 



5. CLASSIFICATION. 



The first who attempted the classification of the 

 Tenthredinidce on an extensive scale was the English 

 naturalist, W. E. Leach (Zoological Miscellany, vol. 

 iii). He divided the family into nine " stirpes," 

 two of which were grounded on Australian forms. 

 Stirpe 1 contained C-imbex, Trichiosoma, Clavellaria, 

 Zarcea, and Abia. Stirpes 2 and 3 were formed for 

 the Australian genera Perga and Pterygophorv* re- 

 spectively. Stirpe 4 had one genus, Lophyrus^ 5 two, 

 Hylotoma and Cryptus ; 6 had four, Messa, Athalia, 

 Selandria,aiidFemi,8a; 7 included Allantus, Tenthredo, 

 Dosytheus, Dolems, and Emphytus ; 8 Coeesus and 

 Neinatus; and 9 Tarpa and Lyda. Saint Fargeau 

 (Mon. Tenth.) had an arrangement of his own, but, 



