Feb 7th, 1873 J * & [Crotch. 



ON THE ARRANGEMENT OF THE FAMILIES OF COLEOPTERA. 

 By G. R. Crotch, M. A . 



(Read before the American, Philosophical Society, February 7, 1873.) 



Since the publications of Dr. Leconte and C. G. Thomson, the classifi- 

 cation of the order Coleoptera has been placed on quite a new footing, 

 and though the old divisions are in great part retained, yet they are 

 defined by entirely fresh characters and are purged of various heteroge- 

 neous elements that had been added to them. Thomson has, indeed, 

 developed a singular classification, differing very widely from any in use, 

 hence I have only been able to avail myself of the details ; and certaiuly 

 no other author has shown such minuteness of observation, though un- 

 fortunately he has confined himself entirely to the Swedish fauna. It is, 

 however, to Dr. Leconte that we owe the most brilliant and fundamental 

 generalization, namely, his isolation of the Rhynchophora by the struc- 

 ture of the prosternum ; the remaining groups can then be worked out 

 with comparative ease, though the three divisions into which they natur- 

 ally fall do not admit of being sharply and definitely characterized. 

 Hence I propose to divide the Coleoptera, as follows : 



A. Anterior coxal cavities closed by the junction of the 

 epimera, elytral epipleura? obsolete, head more or less 



rostral, tarsi 5 jointed, 3d bilobed, fourth minute Rhynchophora. 



B. Anterior coxa? open, or closed by the epimera meet- 

 ing the prosternum, epipleura? generally distinct Coleoptera genuina. 



The only apparent exceptions to this division, are the genus Cossyphus 

 pointed out by Dr. Horn, and some of the Colydiidce, as Both ride res and 

 DeretaphruK, but none of these have the characteristic tarsi ; I know of 

 nothing in the Tetramerom series approaching such a structure. The 

 Coleoptera genuina may be thus divided : 



A. Tarsi 5, 5, 4 ; antenna? moniliform, filiform, or 



serrate, not clavate Heteromera. 



B. Tarsi variable, normally 5, 5, 5, antenna? variable. Isomer*/. 



The only character to be relied on here is the tarsal joints ; and un- 

 doubted Heteromerous genera occur in the lsomera, especially in the 

 Clavicornes, as Liodes, etc. The antennae are clavate in Pentaphyllus, 

 Paratenetus, and some others ; Tetratoma also has them strongly clavate 

 and is probably a Clavicom. 



Isomera. 



A. Tarsi with the 3d joint bilobed, or bifid, 4th very 

 minute connate with the last, antenna? moniliform or 



filiform Tetramera. 



B. Tarsi variable, generally filiform and simple Pentamera. 



The characteristic Tetramerous tarsi recur in Erotylida?, but are ac- 

 companied by a well marked antennal club, and securiform palpi ; in 



