Origin of the Fauna of Celebes. 125 



eight) species recorde;! from Sumatra, Borneo, and Java, Bali 

 possesses only two — Osphromenus trichopteras and Anahas 

 scandens. The hitter occurs on Sumbawa, Celebes, liotti, 

 Sumba, and Timor *, but in such a manner as to preclude 

 the idea of introduction by man. 



The question, therefore, as to the differences existing 

 between the eastern and western portions of the archipelago 

 as regards freshwater tishes may be answered thus : — 



1. llie transition from Borneo to Celebes in respect of 

 river-fishes is very abrupt. Out of the nine families of fresh- 

 water fishes characteristic of the Oriental region only three 

 occur in Celebes, each represented by one species, while, 

 according to our present knowledge, Borneo has 182 species. 

 The Cyprinidffi and freshwater Siluridas, which are so well 

 represented in Borneo, are entirely wanting. This want is 

 not explained by the present hydrographical condition of 

 Celebes. 



2. Since Siluridje (one species) and Cyprinidge (two 

 species) are found in Bali, and not in those islands further to 

 the east (which fact would argue their non-occurrence in 

 Lombok, hitherto unexplored ichthyologically), this would 

 coincide with the original Wallace line. But it must not be 

 forgotten that already in Java there is a decrease in fresh- 

 water fishes, not only with respect to the number of species 

 in proportion to the smaller size of Java compared with 

 Borneo and Sumatra, but also qualitatively, inasmuch as two 

 families — the OsteoglossidiB and LuciocephalidfB — are wanting 

 in Java. In Bali six out of the nine families given in our 

 tables no longer occur f. 



2. What is the Okigin of the Freshwater Fish- 

 fauna IN the Eastern Portion of the Archipelago? 



Since we have shown that the islands to the east of the 

 great Sunda group are almost entirely wanting in true 

 freshwater fishes, the question arises, What is the origin 

 of the fish-fauna on these islands? An inspection of the 

 tables for Celebes and Amboina, although dealing with a 

 much richer material, gives the same result as Ed. v. Martens + 



* Accordiug to Bleeker, Amboina has a peculiar form — Anabas micro- 

 cephalus, JBlkr. 



t Probably, however, Ophiocephcdus striatus and a species of Haplo- 

 chilus occur in Bali. 



X Ed. V. Martens, ' Preusa. Expedition uach Ost-A^ien, Zoolog. Theil, 

 Bd. i. 1874, p. 313. 



