2 



COLEÜPTERA 



at the veiy end of the Rhynchophora (i), but the subfamily is clearly allied to the Eurynchides (2), 

 although the genera of that subfamily have clavate antenna', and the third and fourth abdominal 

 segments small. 



General characters. — ■ Head rather elongate, free. E3es large, lateral, prominent, very finel}^ 

 faceted, briefl}^ ovate. Rostrum ot variable length. Scrobes varying from foveiform to shallow and 

 oblique. Antenna^ thin to moderately stout, neither geniculate nor clavate, second joint short, eleventh 

 longer than tenth, its apical half cone-shaped, all (except some of the joints in Rhinotia) circular in 

 transverse section, usually inserted nearer base than apex of rostrum. Prothorax transverse, dilated 

 towards base, which varies irom truncate to strongh' bisinuate, sides feebly to strongly rounded, ocular 

 lobes absent (3). Scutellum of variable shape. Elytra varying from moilerateh' narrow to very narrow, 

 more or less parallel-sided from shoulders to be3'ond the middle, shoulders rounded, apex variable, 

 margins at junction of metasternum and abdomen very feebly curved. Prosternum with part in front of 

 coxae narrower or wider than part behind coxas. Mesosterninn with intercoxal process \'erv narrow; 

 iront side piece larger than the other and rounded on its front edge, so as to encroach on side of 

 prosternum; hind side piece usually interrupted (4) by a short triangular extension of each el3'tron at 

 the jtmction of its base and side. Metasternum elongate, side piece wide, obtusel}' produced at inner 

 apex and narrowed posteriorly. Abdomen elongate, convex, segments regularly decreasing in length, 

 sutin'es of second, third and fourth straight. Legs usually rather short. Four front coxa> snbglobose, free, 

 hind pair transverse, almost or quite touching elytra; front pair contiguous, middle pair narrowly 

 separated, hind pair separated by an equilaterally triangular process of abdomen. Femora dentate or 

 not (5). Tibi;c straight (except for basal curvature) or slightly curved and usual!}- denticulate below, 

 terminated b\" two short spines, i'aisi denselv setose on lower surface, first joint moderateh' hjng, 

 second shorter, third still shorter and deeply bilobed, fourth elongate; claws simple. Body winged, 

 more or less fusif(jrm or subcylindrical, clothed in })laces with short pubescence (6), usualh' with dense 

 granules in places, but seldom tuberculate. 



Of the genera, Wiinotia (although many of its characters are exacth' as in Belus) is distinct. 



Agiiesiotis is also distinct. 



Pachyiira, although regarded hy Lacordaire as a s\'nonym of Belus. is quite distinct from that 

 genus by the short stout rostrum and wide shoulders. The species standmg in it at present, however, are 

 not verv homogeneous in appearance. 



Belus is remarkable for the gieat variation that occurs at the apex of the elytra, which var\' trom 

 strongh- rounded (as in its subgenus Isacantlia) through almost imperceptible stages to strongly and 

 acutely jnoduced as in bideiilatus and others. 



Cyrotyphus. Agatliinus and Lehus are closely allied and possibly in time it may be considered 

 advisable to unite them. 



The genera may be tabulated as follows : 



Rost) urn more or less loiii; and thin. 



Antennae thin " 2. Genus Bei.us. Schoenherr. 



(1) The American Naturalist, p. 470, 1874. 



(2) Before which subfamily Lacordaire placed tlicm . 



(3) \n Pacliyiiravesiilii. Cyrolyp/ii/s fascicu/an's, J.cliiis Jiuriis, Ag:athiiiiis sexhihi li ulatus, and doubtless in others, there appear tobe 

 ocular lobes, but this appearance is entirely due to a small quantity of seta: or pubescence at the places w liore ocular lobes are usually pi esent. 



(4) Except in Agiiesiofis, in which the triangular extension of each elytron is practically absent. 



(5) This character is of specific importance only in the subfamily, as it is variable in Kliijiotin, /lei 11 (and its subjfenus/ir!Crt«///<!>and Hacliynra . 

 1 be teeth, when present, do not appear as angular extensions of a ridge (as in most weevils) but rise suddenly from the surface : there are trequently two 

 on each femur, placed (not one behind the other, but — ) transversely; with, frequently, numerous smaller ones placed behind them, sometimes in one row, 

 scjmetimes in two. In Belus itself some species are dentate in both sexes, some in the male only, and some in neither sex. 



(6) True scales are not present, although many of the species appear to have spots of scales, this appearance is really due to local density ol 

 pubescence. 



