OF MANDIBTTLATA. 427 



les ailes u'anroient pit, a raison de hiir mnpJeur extraor- 

 dinaire, de leiir grande ihmitc, de Vobstacle que leiirs pli- 

 catiires opposent continueUement a leur extension, taincre 

 la resistance de Vair. Elles sont annexees an mesotborax 

 on an second segment du tronc, et correspondent ainsi aux 

 ailes supirieures des autres insectes." 



Thus I consider it to be established that we are to pass 

 from the Coleoptera to the Hi/menoptera by means of the 

 Strepsiptera*'. As yet, however, I am ignorant of the 

 Coleopterous insects which we ought to quit directly 

 for them. Judging from their fades, I know no Coleo- 

 ptera which approach them closer than the genus Atrac- 

 iocerus. Were I from theory to describe the Cole- 

 opterous insect which ought to come nearest to the Hy- 

 menoptera, it would be nearly as follows : Larva apod, 

 Imago with the thorax small, and the remaining segments 

 of the trunk forming a mass not liable to be confounded 

 with the abdomen. The wings ought not to be folded 

 transversely, and the elytra should perhaps be very mi- 

 nute, since we know no instance among Coleopterous in- 

 sects where they become at their full size membranaceous. 

 It is possible, nevertheless, that the transition from a Co- 

 leopterous to a Hymenopterous form is effected on another 

 principle, namely, the affinity which some Hi/rnenoptera, 

 in their perfect state, may be found to bear to imperfect 

 Coleoptera. Thus, in the Australasian genus Myrme- , 

 codes we observe many of the distinctive characters of 

 Ilymenoptera to disappear. Until the truth be ascer- 

 tained by analysis, it will perhaps be most prudent to 



* If I express myself with more confidence on this head than in the pre- 

 ceding chapter, it is because the receipt of M. Latreille's Memoire has 

 confirmed me in the opinion which I there advanced with some hesitation. 



