124 Prof. Dr. Max Weber on the 



From this Table it may be seen that east of Borneo and Java 

 the Mastacembelidaa, Siluridse, Cyprinidas, Nandidae, Lucio- 

 cephalidae, and Osteoglossidae are entirely wanting. In our 

 earlier list of freshwater fishes existing in Celebes, the Siluride 

 genera Plotosus and Arius were given. These were never 

 found, however, in the interior, but only in the neighbourhood 

 of river-estuaries ; and, what is more important, they are 

 equally at home in brackish water or in the sea. They do 

 not therefore come under consideration here, where we are 

 dealing exclusively with true freshwater fishes. Further, it 

 is also worthy of notice that along the southern chain of the 

 islands the transition is not so sudden as between Borneo and 

 Celebes. The above Table shows us that already in Java 

 the Luciocephalidge, Nandidae, and Osteoglossidae are wanting, 

 while Bali, so far as we know at present, only possesses two 

 of the Cyprinida? and one of the Siluridae. Were we inclined 

 to continue the boundary-line in the usual way between Bali 

 and Lombok, that boundary-line which, for freshwater fishes 

 also, almost completely separates Borneo from Celebes, we 

 should have to bear in mind our complete ignorance of the 

 freshwater fauna of Lombok. Supposing for the moment 

 that the conditions in Lombok were the same as those in the 

 more eastern islands — Sumbawa, Flores, &c. — this boundary- 

 line would have little to mark off. At the best, so far as our 

 present knowledge goes, there would be on the one side Bali, 

 with two Cypvinidse and one Siluridae, while on the other — 

 the eastern side — neither is represented. The difference, in 

 itself unimportant, becomes still smaller when we examine 

 how matters stand in Celebes as regards the Ophiocephalidge 

 and Labyrinthici, equally characteristic of the western portion 

 of the Archipelago. Sumatra has nine, Borneo eleven, Java 

 only four species of Ophiocephalus ; not one is recorded from 

 Bali ; but Celebes, Flores, even Amboina still have Ophio- 

 cephalus striatus. This startling fact was recognized by 

 Bleeker and v. Martens for Celebes and Amboina. v. Martens 

 believed it was possible that this fish, characteristic of the 

 Indian fauna, had been introduced by man ; but against this 

 is to be placed its widespread occurrence in Celebes, even in 

 places where the population has not reached such an advanced 

 state of civilization as to allow the probability of such an 

 introduction. This objection holds good in a still more 

 marked degree for Flores, where I also found this species in 

 a small stream unsuitable for the cultivation of fish. Besides 

 this, the inhabitants of Flores are of a low type and seldom 

 fish. Finally, as regards the Labyrinthici, of the nine (or 



