Brazilian Barin^e 3 



It has been a rather difficult matter to invent names for the very 

 large number of new genera which seem to be necessary. It may 

 possibly be thought by some systematists that the number here 

 proposed is excessive, and, havuig this point of view constantly in 

 mind, I have tried to propose new names only when it seemed 

 irrational to pursue a contrary course. The ratio of species to 

 genera in the present work is about 3.4. Taking the old Henshaw 

 list of nearctic Coleoptera, I find that the ratio of species to genera 

 in the Otiorhynchidae is 1.6, and in the Curculionidae 5.0, or, in a 

 combination of these two families, 3.7. In the Cerambycidae this 

 ratio is 3.0. So it is not probable that the proportion of genera to 

 species here proposed is materially excessive, when we consider the 

 necessity of increasingly discriminative taxonomy, if we desire to 

 render classification amid the almost endless diversity of structural 

 types forming the tropical Coleoptera, in any way clear through 

 descriptive writing alone. If some of the earlier authors, like 

 Schonherr for instance, had given more attention to what they 

 considered unimportant structural details, it would have been much 

 easier to identify material from their published monographs; but, 

 as it is, these are practically useless, even measurement of linear 

 dimensions being omitted in the great work of the author mentioned. 



I have found it convenient to divide the Brazilian Banna; into 

 thirteen tribal groups. As before mentioned, the characters used 

 to define these groups, which are in the main structural but also in 

 some degree habital, are however subject to so many exceptions, 

 that any dichotomous table drawn up to define them would be 

 more or less unsatisfactory, and a paragraph is therefore given at 

 the end of this work, which is designed as a very general index for 

 the purpose of directing the attention of the student, with material 

 for identification, to the tribal group most likely to include his 

 species. 



It is almost unnecessary to add that measurements of length in 

 this memoir, in no case include the beak, and also that only the 

 original spelling of generic names is adopted, even though philolog- 

 ically erroneous. 



Tribe Ambatini. 



With the Ambatides I here unite the Peredinetides of Lacordaire, 

 to form the tribe Ambatini, the only important structural difference 

 being the moderately separated anterior coxae of the latter, as 

 compared with the closely contiguous coxae in Ambates and allied 

 genera. This incongruity may be observed in nearly all the tribal 

 groups of the subfamily; in the Optatini, for example, the con- 

 tiguous coxae of Lydamis and the distinctly separated coxae of 

 Eur y pages, Syprestia and other related genera, can be noted. The 



