484 REVISION OF THE GENUS PAROPSIS, 



be a very striking case of mimicry, though I am unable to offer 

 any suggestion as to what is mimicked or what is the purpose of 

 the mimicry. There remains to be mentioned a third type uf 

 marking, which consists in a sharply defined discal pattern of 

 large brilliant green or golden spots and stripes which seem to be 

 very little subject to variety. The species having this type of 

 marking are among the most beautiful Coleoptera known to me, 

 and it is almost impossible to believe when looking at one of these 

 magnificent creatures alive that in a few hours all its glorious 

 colours will disappear. 



In tabulating I have separated the species having a sharply 

 defined metallic pattern as forming a main division of the sub- 

 group, but have not made separate main divisions of those 

 presenting the other two types of colouring, as I find it difficult 

 to draw a clear line founded on the possession of much or little 

 diff'used metallic colouring. It may be noted, however, that in 

 general the species ending with ohovata, Chp., in the tabulation, 

 have little or no metallic colouring, and the remainder as far as 

 iuconstans (Chp.), Blackb., have metallic coloui'ing diffused over 

 the whole disc of the elytra either in the tessellated or pseudo- 

 sulcate fashion. 



The first main aggregate of subgroup v., then consists of species 

 not having a sharply defined metallic elytral pattern, and among 

 these it will be seen that in some the seriate punctures of the 

 elytra do not run in single file and are very far from symmetrical, 

 the succession of punctures being sinuous or zigzagged, or fre- 

 quently interrupted by two or three punctui'es placed transversely, 

 or even more placed in a cluster. These I have regarded as the 

 first aggregate of species, but have been unable to find good 

 structural characters suitable for tabulation to distinguish them 

 inter se. The species (of the first main aggregate) having the 

 elytral series symmetrical, or nearly so, are divisible into two 

 sections by the presence or alisence of lateral inequalities on the 

 elytra. These consist of shallow transverse depressions by no 

 means strongly marked, but quite discernible, and which more or 

 less sti'ongly interrupt the symmetry of the 10th elj'tral series. 



