COLEOPTERA 



187 



of the area into islands between which there is but httle communication, it would 

 be very difficult to say. But I have an impression that this is the case. The forms 

 I have placed in Colpocaccns are well-winged, and they are more nearly allied than most 

 of the other forms I have admitted as species, so that it would perhaps have been better 

 to treat one or two of them as subspecies. I incline to think that this inferior 

 differentiation may possibly be due to the greater frequency with which individuals of 

 Colpocaccns pass from one island to another. The numerous flightless forms can in 

 most cases be rarely, if ever, transmitted from one island to another. Careful inquiry 

 may possibly show that the flying forms are represented by subspecies in different 

 islands, and the flightless forms by distinct species. But on these points there is no 

 adequate evidence. 



Distribution. All the Hawaiian Carabidae are precinctive with the exception of 

 one species that is probably not native. Three or four others may possibly be found 

 elsewhere. 



There is nothing whatever to indicate any particular region as that from which 

 they were derived. Platymis and Benibidiuni are the two genera that Hawaii has in 

 common with other parts of the world, and both of them are nearly cosmopolitan. 



The species of the Cyclothorax division (Pterostichides) are very closely allied to 

 the two or three species that constitute the genus Cyclofkorax. These are found in 

 Australia and New Zealand. But in those countries only two or three of these forms 

 are known, whereas in Hawaii we have about 80. It would therefore be quite as 

 reasonable to infer that the Antipodean forms have been derived from the Hawaiian 

 as to assume the opposite proposition to be true. 



Distribution within the islands. We may summarise this by saying that nearly the 

 whole of the species are confined to a single island. When a species occurs on more 

 than one island it is nearly always the case that the localities of its occurrence are on 

 i.slands that are adjacent, such as Maui and Molokai. 



Table. Showing the genera and their distribution in the islands, and the number 

 of species in each island. 



* Since the remark on p. 195 was printed I have ascertained that the habitat of Atrachycnemis koeheki is 

 W. Maui, not Hawaii. 



