PARASITIC HYMENOPTERA. 229 



resemblance of one or other of these predominates, I find the 

 resuhing groups may be separated in this manner : — 



three angles Agathis, &c. 



Second brachial I ( determinate and alike in the 



cell of the lower ) , \ sexes IVIicrog aster, &c. 



wings bounded j if' 



externally by i "V"^ ^.^ ° < varying with sex f three . . Bracon, &c. 

 'Clnn" l«"d species ;cu-S 



upper wings (two . . . Leiophron, &c. 



These characters have the appearance of being taken arbi- 

 trarily, the table being calculated simply to distinguish the 

 groups, without exhibiting their relative position or connexion ; 

 and other schemes might be drav^^n up which would attain the 

 same object, as, indeed, any one devised from such scanty 

 materials is likely to prove imperfect in a more extended 

 application. I do not attach much importance to this, as the 

 temporary fabric is easily reconstructed when further infor- 

 mation has shown its defects. But, if the standards selected 

 possess the requisite qualities, the entireness and relative 

 proximity of the groups may remain unaffected by any addition 

 to their contents. The first three are so obviously indicated 

 by external appeai'ance, that there is little difference of opinion 

 to be apprehended as to their existence or extent; but the 

 combination of the fourth group rests in great part on the 

 similarity of the trophi, since there is less resemblance in other 

 respects between the proposed type and some of the species 

 here associated with it, than the latter bear to a section of the 

 genus Perililus. I believe, indeed, that we should attribute 

 this resemblance to a real affinity, and that the supposed 

 families meet at this point. Now as they appear to approach 

 at the other extremity also, and as the series through each, 

 from one point of contact to the other, exhibits no manifest 

 interruption, we have in this way a complete circle formed, to 

 the exclusion of ApMdius, which seems to be thrown into a 

 separate group. The different habits of that genus have 

 already prepared us for such a result; but it will demand a 

 more detailed and rigorous examination of the remainino- 

 genera than I am competent to give them, before it can be 

 conclusively admitted. For greater convenience in designating 

 species, the generic denomination of the type may be extended 



