Ii6 Records of the Indian Museum. [Voi,. XVI, 



clearly specialized genus Discognathus and in what way these modifi- 

 cations are actually utilized. 



With the structure of the mouth in Discognathus I have dealt 

 to some extent. The Western Asiatic forms {D. variabilis and D. 

 laiiita var. rufits) closely resemble the Indian forms in this respect, 

 and so does the one species (D. borneensis) known from the Malay 

 Archipelago. I have been able myself to observe both the typical 

 D. lamta and its Syrian race under natural conditions in ex- 

 ceptionally favourable circumstances. Neither of these forms lives 

 habitually in mountain streams ; both affect the pools of streams 

 and rivers and even isolated masses of still water ; they are what 

 we may call normal forms of the genus with well- developed mental 

 suckers and with the pectoral fins and chest comparatively little 

 modified. My observations on the vSyrian fish were made in a 

 walled fountain at the lyake of Tiberias, those on the Indian form 

 in the Inle Lake in the Southern Shan States. Moreover, in the 

 literature on other species of the genus I can find no statement that 

 would justify, so far as m.ost species are concerned, the belief that 

 the genus is like Psilorhynchus, essentially a mountain one. Blan- 

 ford ' took the types of D. blanfordii in a stream which he describes 

 as a torrent, and Max Weber ^ states that D. borneensis lives in 

 mountain streams. All the other species are recorded from rivers 

 or lakes. Discognathus lamta, in both its races, feeds on small 

 organisms that are tightly fixed to rocks or other hard objects. 

 Its manner of feeding is this. Having fastened itself, usually in a 

 more or less vertical position with the head uppermost, to a 

 rock or post by means of its mental disk, it selects suitable food 

 with its lips, bites it off with its jaws and sucks it into its almost 

 horizontal buccal cavity. When the food in its reach is exhausted, 

 it relaxes its adhesive organ and by means of an almost impercep- 

 tible movement of its tail, thrusts itself slightly upwards. The 

 disk then takes hold again. In Palestine I experienced this process 

 by bodily sensation on placing my bare feet in the water of the 

 fountain. The fish invariably attached themselves and it was 

 possible to feel the action of the disk fixing itself, the movements of 

 the lips and the nibbliugs of the jaws, which were not sharp 

 enough to pierce the human skin. Psiloriiynchus apparently feeds 

 in the same way but clings rather by means of its pectoral fins and 

 flattened, highly muscular chest, which can probably be rendered 

 concave by muscular action. Discognathus nasutus has both means 

 of attachment strongly developed. 



It is noteworthy that in D. blanfordii, so far as can be seen 

 from Boulenger's figures, the fins and chest have a considerable 

 resemblance to those of D. nasutus, while those of the other 

 African species, which apparently live in comparatively still 

 water, resemble those of D. lamta. The Bornean species also 



^ Discofi^iiathus lamia, Blanford, Geol. Zoul. Abyssinia, pp. 460-461 (1S70). 

 "2 Indo-Australiau Fish, [II, p. 228 (1916) ; see also Vaillant, Notes Leydeii 

 Mtis., XXIV, p. 9, Hos. 25, 26 (1902). 



