548 



THE AMERICAN NATURALIST [Vol. XL V. 



farther than the Cyprinidnp. In these families one should 

 include only those whereof all, or by far the greater 

 majority, of the species are exclusively confined to fresh 

 water. It is naturally evident that in other cases the pos- 

 sibility of their fortuitous transportation across the sea 

 is not to be excluded. The appended table gives in outline 

 the distribution of fresh-water fishes in the archipelago. 

 It is gleaned, for the most part, from Weber (1894, 

 1897A), who lias done more than any one else to add to 

 our knowledge of the dispersal of the fresh-water fishes 

 throughout the islands. For comparison a table is added 

 giving the distribution also of the families of amphibians. 



The eastern portion of the archipelago, including Cele- 

 bes, possesses a very small fresh-water fish fauna. Those 

 which are found most commonly in the rivers and lakes 

 are, as Weber has pointed out, for the greater part 

 "marine immigrants," that is to say, sea-fishes which 

 have passed over to a fresh-water life. Of the true Indian 

 fresh- water fishes which occur to the eastward of Wal- 

 lace's line and Lombok are the Siluridae; the New Guinea 

 species, however, related to those of Australia and hence 

 being in reality outside of our present range of discussion. 

 The remaining ones are almost exclusively those which are 

 able to withstand a drying-up— if this is not too long 

 continued. These are, in truth, only a few species. 

 Haplochihis crlcbcusis among the Cyprinidontida 1 does not 

 get further than Celebes, nor does Notoptems kapirat; and 

 two species of Symbranchida-— Si/mbranchus bcngalensis 

 and Monopterus jara>n(>>isis — Oph)oceph«hiS striatus, 

 Anabas sccnideus and Annhas m'nroeeplialus and possibly 

 Betta pugnax are spread further to the eastward. The 

 last-mentioned species have all a peculiar structure of the 

 gill cavity which makes it possible for them to remain 

 for a considerable time without water; and it has been 

 shown by Volz for Monopterus that it may even live over 

 through the dry monsoon without water. 



The distribution of mammals in the eastern half of the 

 Indian Archipelago is evidently not yet completely known. 

 Even for New Guinea itself, with the exception of bats, 



