278 



is perhaps open to objection on the ground that accidental 

 circumstances — sucii as distortion — may ai5:ect its reliability; 

 but, nevertheless, it is found on examination that the princi- 

 pal part of the propygidium is, in the case of Caulobius, sl sur- 

 face, from its M-ant of sculpture and vestiture, evidently de- 

 signed to be a covered part of the body, while in Automolus 

 the sculpture and vestiture are evidently those of an exposed 

 segment, and are more or less uniform with those of the py- 

 gidium. 



The antennae of Automolus are not easy to examine, the 

 joints between the 2nd and the club being very short, and 

 their sutures difficult to distinguish. When paucity of speci- 

 mens forbids the removal of an antenna I have been unable 

 to arrive at certainty as to the number of joints of the an- 

 tennae in the species before me. I have not, therefore, been 

 able to use this character in tabulating the Automoli, but I 

 can say that the antennae are by no means of uniform struc- 

 ture, there being in most of the species eight joints only, of 

 which three form the club, while in at least one species there 

 are certainly nine joints, of which three form the club, and 

 in two species known to me the club (of at least one sex) con- 

 sists of four joints. 



The Autowoli have a most remarkable sexual character 

 in the elytra of the female, which appears to have been over- 

 looked in the descriptions of all the hitherto described species. 

 This consists in an elevated nitid space (varying in size and 

 position with the species), which in some {e.g., poverus, 

 Blanch.) is extremely conspicuous; while in others it is small 

 enough to be easily disregarded. Other sexual characters are 

 found in the gi«eater elongation of the antennal flabellum and 

 peculiarities (very pronounced in some species) in the front 

 tarsi of the m.ale. 



According to Burmeister (who uses the name "Li'pare- 

 triclce" for the aggregate, which Lacordaire — and I in these 

 memoirs — call "Se?'ico/des" ) the gienera Automolus and 

 Caulobius belong to different sub-aggregates distinguished by 

 the comparative length of the ventral segments — the 5th seg- 

 ment in the former being longer than, and in the latter equal 

 to, the 4th. " My observations show that there is an evident 

 variation in this respect in closely-allied species, correspond- 

 ing to the variation in the size and prominence of the propy- 

 gidium on the dorsal surface — so that Burmeister's distinction 

 between Antomolus and Caulobiio^ is in reality the same that 

 I have indicated in the tabulation above. I cannot, however, 

 regard it as of sufficient importance to be used in forming 

 groups of genera — indeed, as already remarked, I doubt its 

 being even generic. 



