34 



pentainerous, yet are so closely related to Tenebrio that they 

 preserve a place near them in the Heteromera. . 



Your ficrures of the mentum and thorax of Chremastocheilus 

 were highly gratifying to me, and I shall duly acknowledge the 

 favor when I describe the Carolinian species, which certainly 

 is not the C. variolosus of Kirby, although I think C. Sayi 

 may be. 



In regard to the folding of wings in Hymenopterous in- 

 sects, all I can say is, that Leucospis^ all of the Vcspidce 

 Cexcept only Ceramius') and the Masaridce^ have them folded 

 longitudinally, and no others to my knowledge. The way then 

 to ascertain from the wings themselves whether the insect be- 

 longs to the Vespidce, is to observe whether the nervures follow 

 this type — one marginal and four submarginal cells, the fourth 

 being apical and incomplete. With Leucospis there can be no 

 doubt. 



Ceramius, which is another instance of exceptions to general 

 rules, is known by the extreme smallness of the maxillary 

 palpi, which consist of only three or four joints, while those 

 of the allied genera, Eumenes, Odynerus, etc., have six, and 

 are nearly as long as the labial palpi. According to these 

 characters, the two Hymenopterous insects which you sent me 

 are Vespidce : they are PoUstes, because the mandibles are not 

 longer than broad, are not prolonged into a beak, but have 

 their apices truncated ; because the clypeus is nearly quadrate, 

 and the middle of its anterior margin produced into a little 

 tooth, the abdomen oval (not conic,) and peduncled. The 

 maxillary retractile ap})endagcs or tongues of your Malthhii 

 are really most singular and anomalous chai'acteristics. I 

 must, however, beg of you a full description of each species 

 before I can feel qualified to determine the identity of one with 

 Cantharh hhnaculata. You may recollect that you gave me a 

 specimen which you caught in Pennsylvania on the thistle. 

 Two years ago I compared it with Olivier's description and 



