Brazilian Baring ioi 



to any other type of Barinae. The Madopterini are, in fact, so 

 isolated that they are placed in the present work at the end of the 

 series; they are probably closer to the Madarini than any other 

 tribal group. 



Tribe Centrinini. 



This is an enormous complex of genera and species as at present 

 understood, and in order to reduce the taxonomic difficulties to 

 some extent, I have separated a large part of them, having a nar- 

 rower and parallel body, usually with much shorter beak, under 

 the name Limnobarini; but, even after this substantial reduction, 

 the numbers met with in the neotropics are appalling to the sys- 

 tematise It is not possible to define the limits of either of these 

 tribal groups at all rigidly, and it can simply be said that in the 

 Centrinini the body is of a more or less evident rhombic shape, the 

 beak usually long and the antennae more or less slender — although 

 there are many exceptions such as the thick beak of Neplaxa, the 

 anterior coxae never in mutual contact, although virtually con- 

 tiguous in Roalius, and but seldom remotely separated as they are 

 in the Madarini. The male presternum may or may not have 

 somewhat variable ante-coxal spines, which is also the case but less 

 frequently in the Limnobarini. The tarsal claws may be free, 

 connate or single and the mandibles are straight and internally 

 smooth, dentate and more or less decussate or of peculiar form, as 

 in the genus Garnia and especially in Nedestes. The body may be 

 entirely and closely clothed or completely glabrous, with all kinds 

 of intermediate vestiture. The prosternum only very rarely has a 

 deep or sharply limited pectoral canal, nor is the upper surface of the 

 body high and vaulted as in the Diorymerini, where prosternal 

 spines are unknown. The Optatini, which also have the pronounced 

 rhombic form of body, may be distinguished generally, but not 

 always, by the contiguous or veryclose anterior coxae, andabsolutely, 

 so far as known, by the entire absence of prosternal spines in the 

 male, although this is rather a negative than positive discriminative 

 character. I think that much work will have to be done on larval 

 histories before a rational final classification can be made in the 

 Barinae. 



The genera which I have thought should be associated under the 

 name Centrinini are extremely numerous; those now in my collec- 

 tion from the regions under consideration are as follows: 



Tarsal claws connate or extremely approximate, rarely single 2 



Tarsal claws free and more or less divergent 24 



2 — Body never less than about 3 mm. in length, the upper surface entirely glabrous 



— rarely excepting the scutellum, thoracic lobe or elytral suture 3 



Body minute in size, always much under 3 mm. in length and in most cases 



conspicuously clothed throughout; femora never denticulate 17 



