DISTRIBUTION 327 



icial means, such as with provisions, stores, building timber, ballast, or 

 growing plants ; many of these species are nearly cosmopolitan. Second, 

 species that have arrived in the islands, and have become more or less 

 completely naturalized; they are most of them known to be wood- or 

 bark-beetles, but some that are not so may have come with the earth 

 adhering to the roots of floating trees ; a few, such as the Dy tiscidae, or 

 water beetles, may possibly have been introduced by violent winds. 

 Third, after making every allowance for introduction by these artificial 

 and natural methods, there still remains a large portion standing out 

 in striking contrast with the others, which we are justified in considering 

 strictly endemic or autochthonous." Among the introduced genera are 

 Coccinella, Dermestes, Aphodius, Buprestis, Ptinus and Cerambyx. The 

 immigrant longicorns appear to have been derived "from the nearest 

 lands in various directions" the Philippine Islands, tropical America 

 and the Polynesian Islands and the same conclusion will probably be 

 found to hold for the other immigrants, when their general distribution 

 shall have been sufficiently studied. The endemic species number 214, 

 or exactly half the total number of species, and are distributed among 9 

 families, as follows: 



FAMILIES. SPECIES. GENERA. ENDEMIC 



GENERA. 



rabidae 51 7 7 



Staphylinidae 19 3 i 



Nitidulidae 38 2 i 



Elateridae 7 i i 



Ptinidae (Anobiini) 19 3 3 



Cioidae * 19 i o 



Aglycyderidae 30 i i 



Curculionidae (Cossonini) 21 3 3 



Cerambycidae 10 i i 



Sharp writes: "I think it may be looked on as certain that these 

 islands are the home of a large number of peculiar species not at present 

 existing elsewhere, and if so it follows that either they must have existed 

 formerly elsewhere and migrated to the islands, and since have become 

 extinct in their original homes, or that they must have been produced 

 within the islands. This last seems the simpler and more probable sup- 

 position, and it appears highly probable that there has been a large 

 amount of endemic evolution within the limits of these isolated islands." 



The parasitic Hymenoptera of Hawaii, according to Ashmead, num- 

 ber 14 families, 69 genera arid 128 species; only eleven genera are en- 

 demic and most of the other genera are represented in nearly all the 



