496 Coleopterological Notices. 



widely different directions, still I believe this fact should be men- 

 tioned as being at least of possible significance. 



3 The men turn in Lyctus, and especially in Trogoxylon, is very 

 largo, filling the entire width of the gular opening and concealing a 

 large portion of the maxilla?. In Rhysodes the enormous mentum 

 is perhaps the most marked peculiarity of the genus, this organ 

 being developed to a degree probably unknown in any other cole- 

 opteron. It seems probable that Lyctus may be related in this way 

 to Rhysodes, and thence to the Cucujinse. 



4 All the important characters of Lyctus, other than those referred 

 to, find their very satisfactory counterparts among various repre- 

 sentatives of the Cucujidae in the broad sense in which the family 

 is here considered, and it is not at all easy to perceive any very 

 striking serricorn characteristics. In fact no systematist who has 

 placed Lyctus in the serricorn series has ever dwelt with great 

 emphasis upon any particular character as proof of the relationship. 

 DuYal, who seems to have been most candid in this respect, cites 

 the form of the anterior coxa? as a reason for refusing it a place in 

 the Cucujidae, and the divergence of the larva from that of Crypto- 

 phagus as a reason for disregarding its general clavicorn affinities. 

 The first of these reasons is of but little moment, as the anterior 

 eoxse have many parallels among normal cucujides, 1 and the second 

 is not decisive, for as shown by Lacordaire, the form of the larva 

 allies it to the Scolytidae rather than to the Bostrichinae, — assuming 

 of course that the larva has been correctly identified, — and as the 

 Colydiidae are also known to have rhynchophorous affinities, this 

 argument is rather more effective for a clavicorn than for a serricorn 

 relationship. 



Regarding the Cucujidae therefore as a family of Clavicornia, in 

 which the anterior coxas are small, rounded, deeply inserted and 

 never decidedly prominent, and the tarsi generally slender, I would 

 include within its limits the following subfamilies: — 



Genas with large porrect processes ; tarsi isonierous in both sexes, hut either 

 pentamerous, subtetramerous or tetramerous ; antennas rnoniliform or cla- 

 vate PassandrinjE 



Gense without porrect processes. 



Tarsi tetramerous ; antennas frequently with a solid club. 



Elytra entire Colydiinje 



1 I fail completely to see how the anterior coxae of Lyctus can be considered 

 "conical and prominent" as stated by DuVal. This is simply a question of 

 observation and can be easily verified. 



