COLEOPTERA 185 



inconsistently. I have placed in Atrachycncmis A. sharpi with setae, though the other 

 two species of the genus have none. But in the case of the Mecyclothorax I have 

 separated specimens generically, solely on account of a difference in the setae. My 

 reasons for this discrepancy are as follows. In Atrachycncmis the setae are extremely 

 reduced in size, all of the setae of the body being very small and feeble : moreover in 

 this division of the Hawaiian Carabidae the setae appear to be of less importance than 

 they are elsewhere, they being, I suggest, rendered functionally useless by the exudation 

 with which such genera as Blackbttrnia, Atrachycneviis and Anchotejflus are so frequently 

 covered. On the other hand the Cyclothorax forms have the setae of the body well 

 developed, and their presence or absence appears not to be affected by any considerations 

 such as those I have alluded to as existing in the case of Atrachycneniis. The problem 

 as to whether Mecyclothorax robtisttts and Thriscothorax rolmstiis (differing almost solely 

 by their setae) may be really only dimorphic forms of one species, is not the least 

 interesting of the questions raised by my slight study of the Hawaiian Carabidae, and as 

 the insects are apparently not rare on Haleakala it may be possible to decide it by obser- 

 vation of the forms in their haunts. I have alluded to it when speaking of those species. 



A most interesting condition is present in Atelotlirus transiens. In that species the 

 pit is present but the seta that should grow out of it is usually absent. We have 

 received fifty examples of this species, and in only one of them is the seta present ; it is 

 then on one side only, but it is of normal size. This is a highly interesting variation. 

 Taken in connection with the fact that the setae in Hawaiian Carabidae are so much 

 below the average it induces one to suggest that this is a case in which the structure is 

 at present actually in the process of specific atrophy. 



The case of Metronietms pavidus and Mesotliriscus truncatus already alluded to is 

 much more obscure. Among 240 examples of the former species — which is one having 

 no thoracic setae — I found three examples that have a seta. It is medianly placed, and 

 I have described them as a separate form, Mcsothrisciis triiucatiis. If this should prove 

 to be a sport of Metromeiius pai'idns it may be considered a case of atavistic reversion ; 

 the species formerly had setae ; has lost them ; and yet examples are occasionalh' 

 produced having the setae. 



One more point may be mentioned ; there are in the Hawaiian fauna a certain 

 number of Anchomenids that depart considerably from the average of their fellows. 

 They are more elongate in shape, with longer legs, and are also of a paler colour than is 

 usual in other forms, Plate VI. fig. 10. In accordance with my system these forms come 

 into three genera, Atelotlirzts, Mesothrisczis and Mctronicuits. I was at first inclined 

 to the opinion that the separation of these similar forms was a result that invalidated 

 the system. On comparing these forms--of which there are some eight or ten — it 

 appeared, however, that none of those that I considered different on account of the 

 setae could be correctly associated as one species even if the setae were altogether left 

 out of consideration ; and I therefore look on these long-legged forms as instances ot 

 convergence. They are all very rare, and are, as it were, being brought together 



24 — 2 



