638 REVISION OF THE GENUS PAROPSIS, 



great majority of the species to form one vast group. I think, 

 however, that there is one character founded on form that ma}' 

 be profitably employed in constituting primary groups, viz., the 

 shape of the prothorax, as there is a large number of species 

 obviously allied inter se, the sides of whose prothorax are mucro- 

 nate in front (in many instances bisinuate) and very few indeed 

 possessing this character which there can be any hesitation in 

 regai'ding as naturally allied to them. In following Dr. Chapuis' 

 system of groups I have, however, found it desirable to modify 

 it by somewhat increasing the number of primary groups, and 

 also transposing the position of some of his groups, as I feel con- 

 fident that the natural place of his fourth group is immediately 

 after his first group. I propose, therefore, the following division 

 of the genus into primary groups : — 



A. Sides of the prothorax mucronate in front (in many species 



bisinuate) Group i. 



AA. Sides of the prothorax evenly arched. 



B. Punctnration of the elytra without any linear arrange- 

 ment Group ii. 



BB. Puncturation of the elytra more or less linear iu 

 arrangement. 

 C. About 20 more or less regular rows of punctures 

 on each elytron. 



D. Elytra verrucose Group iii. 



DD. Elytra devoid of verruca? Group iv. 



CC. The linear arrangement is very partial and merely 

 the result of several longitudinal unpunctured 



spaces (4roup v. 



CCC. About 10 defined rows of punctures on each 



elytron (iroup vi. 



In the above scheme Groups i. and ii. together include almost 

 exactly the species of Dr. Chapuis' Group i. ; Groups iii. and iv. 

 together equal (again almost exactly) Dr. Chapuis' Group iv. ; 

 Groups V. and vi. equal Dr. Chapuis' Groups ii. and iii. respec- 

 tively. 



