428 ON THE TlilBES 



adopt the middle path between these two hypotheses, hold- 

 ing it for certain, nevertheless, that some of the first Hi/- 

 menoptera we can approach, -when keeping close to the 

 path of affinity, must be such as are destitute of ocelli. 



Hymenoptera. 



On looking at the Hymenoptera generally, and endea- 

 vouring to fix on the most distinct types of construction, 

 1 made choice of Formica, Chtysis, Apis, Sphex and 

 Ichneumon. In such cases, Linnaeus is a guide almost 

 infallible, from his wonderful facility in discovering the 

 minor natural groups. If he could but have combined 

 these as well as he has defined them, he would still be in 

 legitimate possession of that rank in Entomology, of which 

 be has been so long only the usurper, to the prejudice of 

 Degeer and others. Now, besides the Hymenopterous 

 gCQera before mentioned, which I have chosen as types, 

 LinncEKs has no^iorethan five, to wit, Cynips, which 

 comes so clese to Ichneumon that Latreille refers them 

 both to the same group; Tenthredo and Sir ex, which dif- 

 fer so much from true Ihpnenoptera, that it is difficult to 

 imagine them to belong exactly to the order ; MuliUa, 

 Avhicli is now with propriety referred to the Formicida; as 

 containing its type; and finally Fespa, which from its 

 form and manners seems not to be a peculiar type, but only 

 intermediate between Apjis and Sphex. M. Latreille in- 

 deed has founded his family of DipJoptera upon the Lin- 

 nean genus Fespa, and apparently M'ith much reason ; 

 but the simple circumstance of the upper wings being 

 doubled longitudinally is not sufficient, in the opinion of 

 M. Lamarck, to constitute a primary division of the 

 order. In short, if we consider those Hymenoptera of 



