276 MEMOIRS ON THE COLEOPTERA 



produce continual exceptions in any system of tribal division that 

 can be proposed, based at least upon external characters of the 

 imago alone, so that no very satisfactory method of tribal sub- 

 division has been discovered; but, when limited to a restricted 

 fauna such as that of North America, the matter is simplified. I 

 think, for instance, that the species of the North American faunas 

 can be assigned to the four following tribal groups, without any 

 intermediates, so far as known, and therefore in a very definitive 

 manner: 



Mes-epimera distinct from a dorsal point of view 2 



Mes-epimera not visible dorsally 5 



2 Mandibles small, in large part membranous; last abdominal spiracle 

 flat and situated at or before the middle of the segment; mentum, 



tarsi and antennae normal; habits hypergeal 3 



Mandibles in great part corneous, though very small; last abdominal 



spiracle tumidulous and behind the middle of the segment 4 



3 Scutellum covered by the basal lobe of the pronotum. . GYMNETINI 

 Scutellum free, the pronotal lobe feeble and usually sinuate at the 



scutellum CETONIINI 



4 Mentum, tarsi and basal joint of the antennae abnormal; scutellum 

 exposed; sexes almost completely similar; habits often inquilinous. 



CREMASTOCHEILINI 



5 Scutellum exposed; elytra not sinuate laterally; prothorax more hex- 

 agonal; habits floricolous TRICHIINI 



These groups are not considered of equal value by Lacordaire, 

 but I can see no reason for suggesting only two principal groups 

 based upon the extent of the mes-epimeron, as the form of this 

 part is intermediate in the Cremastocheilini, which type is fully as 

 isolated in every way as the Trichiini, if not more.* 



Tribe GYMNETINI. 



The species of this tribe are recognizable easily by the elongate 

 basal lobe of the pronotum, which hides all or nearly all of the 



* I greatly regret having to use the spelling Cremastocheilus, for Cremastochilus is 

 rather better and less inconvenient, but if the laws of nomenclature are to be of any 

 permanent value they must be consistent. One of these laws, now generally recog- 

 nized, is that the spelling of generic words cannot be changed, except they be mis- 

 printed in such a way as to render them unpronounceable, and then an entirely new 

 name must be proposed, with new authorship. This objection cannot be urged in 

 such cases as Cremastocheilus, Amblycheila or Leiopus, and the spelling of these and 

 similar words, after publication in a legitimate way, should therefore not be altered. 

 Specific words can be altered, however, because these are supposed to have a definite 

 meaning in a definite language. Generic words have more the nature of arbitrary 

 symbols. 



