K 0. 2323. LARVAE OF THE CLERIDAE—BOVING AND CHAMPLAIN. 593 



tidae and, to a less extent, the Malachiidae have a nutant and some- 

 what invaginated head and resemble in that respect both the Parni- 

 dae and the Rhipiceridae. 



With the families Parnidae, Ehipiceridae, Throscidae, Eucne- 

 midae, Malacodermidae, and Elateridae, the Cleridae are not so 

 closely associated as with the Malachiidae, Dermestidae, and Tro- 

 gositidae. This can readily be realized from the following brief 

 morphological and phylogenetic analysis of those first mentioned 

 families. 



The Parnidae and Rhipiceridae have, as mentioned above, a nutant 

 or seminutant head. Labrum is well developed and clypeus rather 

 distinct: cardo broad and low; stipes maxillae broad. In the Rhi- 

 piceridae only a single mala is present; in the Parnidae a well- 

 developed lacinia and well-developed galea. In the Parnidae, men- 

 tum has the usual quadrangular form present in most of the Cleroi- 

 dea, but in the Rhipiceridae the mentum is posteriorly constricted 

 as in some Elaterid genera. In the Parnidae the antennae are three- 

 jointed, in the Rhipiceridae very short, retracted, one or possibly 

 two jointed. In both families the ninth segment enables a perfect 

 sealing of the anal opening; in the Parnidae this ninth segment is 

 often prolonged and conical, in the Rhipiceridae operculate, convex 

 and circular; in the Parnidae the body is terminated by three 

 bunches of gills, in the Rhipiceridae no similar structures are present. 



The Parnidae must be derived from the Byrrhids but show also re- 

 lationship to tlie Dermestidae. The Rhipiceridae are undoubtedly 

 closer connected with the Dermestidae and the Parnidae, especially 

 the cylindrical Parnidae, than with the Elateridae, from which 

 family they usually are supposed to have branched. 



The Throscidae and Eucnemidae are still more remote from the 

 original simple Cleroid type than the Parnids and Rhipicerids. The 

 genus Throscus, represented by Throscus dermestoides Linnaeus is in 

 many respects the prototype of the Throscid-Eucnemid group, the 

 form from which the Eucnemidae have metamorphosed by a retro- 

 gressive development. The legs, which have disappeared in most 

 Eucnemids, are well developed in Throscus dermestoides. With ex- 

 ception of the mandibles, which in Throscits dermestoides are fused 

 with the mouth frame, but free in some Eucnemid genera, the mouth- 

 parts as well as the antennae are less rudimentary in Throscus than 

 in the Eucnemids. Labrum is absent in both families, only repre- 

 sented by a small triangular process from the epistomal margin 

 of frons. The isolated morphological development of the group is 

 indicated by the invagination of the head, but especially apparent 

 in the extraordinary rodshape of the prothoracic presternal, pre- 

 hypopleural, and paired sternal (more correctly sternellar) plates. 



144382— 20— Proc.N.M.vol.GT 38 



