276 GEO. H. HORN, M. D. 



extremely unsafe one to use ia any system of classification beyond 

 the separation of species. 



Tlie separation of the genera of the tribe as found in the books is 

 by no means happy and in some respects apt to seriously mislead^ 

 The tables for the greater part either begin with dividing the genera 

 in two series based on the similarity of the tarsi in the two sexes or 

 their dissimilarity, or else the contractile power is taken as the start- 

 ing point, these systems are sometimes varied by first excluding those 

 genera with five-jointed tarsi on all the feet. It is certainly true that, 

 taken as a whole, those genera with the tarsi similar in the two sexes 

 are not contractile, while those with dissimilar tarsi are contractile, 

 often very much so. Unfortunately we find that both Cyrtusa and 

 Colenis are more contractile than some Liodes or even Agathidium 

 and the character at once fails. The similarity or not of the tarsi 

 in the sexes is a much better means of separation than the other, un- 

 fortunately it requires both sexes for positive determination, as a female 

 Colenis is equally well placed in either series if the male is unknown. 



The presence or absence of an antenual groove which has done 

 such good service in other parts of the Clavicorn series is here a 

 character of very great value. Reaching its highest development in 

 Agathidium and Liodes, the groove becomes feebler in the other 

 genera but still well marked even in Cyrtusa where a distinct ridge 

 exists within and behind the eye beneath, and the inner limit of the 

 groove is formed by an elevation of the edge of the buccal cavity. 

 Although the existence of the groove has been casually mentioned 

 by various authors, its importance as a means of dividing the genera 

 has not been recognized. By reference to the table it will be found 

 that the genera with similar and dissimilar tarsi in the sexes are 

 rather sharply separated, Cyrtusa being the only exceptional case and 

 this seems to me no detriment to the value of the character inas- 

 much as it fixes that genus as a very natural link between the two 

 series of genera, a fact which seems admitted by the position in which 

 it is placed. 



As in Necrophorus a portion of the clypeus is membranous, so in 

 the present tribe we find a tendency to the same structure. Triarthron 

 has the clypeus entirely membranous, Hydnobius partly so, while in 

 other genera the anterior border is slightly membranous or the entire 

 clypeus may be corneous. The suture between the clypeus and front 

 is rarely well marked, usually entirely obliterated and not visible except 

 the front be slightly translucent. 



