REPORT ON FISHES. TART II: SISORIDAE AND CYPRINIDAE 327 



surface, lilack dcits are irre!L;"uIarly scattered all over the body. All the fins are marked 

 with faint black sputs and bands. 



Pistribiitioii: Glyptothora.v l<:(isliiiiirciisis is an endemic species of the Kashmir Valley 

 and is fairly common in rivers and sluggish waters. 



[Attention has already been directed to the two ecological groups into which the species of 

 the genus Glyptothora-v can be divided {Rcc. hid. Mns. XXV, pp. 4-8, 1923). The first group 

 comprises less specialized forms which live in larger streams at the bases of hills and are liable 

 to be carried into slow currents, while the members of the second group are more specialized and 

 are invarialjly found in torrential streams. In the former case, the body is cylindrical and the 

 adhesive apparatus is restricted to the chest and the belly as an elongated, coin])act structure. The 

 paired fins are not horizontally placed and play only a small part in the adhesion of the fish. The 

 species of the second group are depressed and possess a flattened ventral surface ; the paired fins 

 are horizontally placed and the function of adhesion is transferred from the chest region to the 

 outer, flattened rays of the paired fins. The thoracic, adhesive apparatus is considerably reduced 

 and is restricted to the region behind the isthmus in a semilunar patch. The probable evolution of 

 Glyptotliorax from forms like Erctliistcs and Laguvia and their ultimate development into forms 

 like Glyptostcrnuin have already l)cen explained {Phil. Tran.<;. Roy. Soc. London, B, CCXVIII, 

 p. 236,1930). 



The form of the body and the structure of the adhesive apparatus of Glyptotliorax kashinir- 

 cnsis are dififcrent from the species assigned to the above groups. The Ijody, though cylindrical, 

 is much deeper and the adhesive apparatus is restricted to the chest region only {Rcc. Ind. Mus. 

 XXV, p. 23, fig. 2 a-c, 1923) ; it is almost circular in outline and possesses a deep pit in the 

 centre. Not knowing the precise habitat of the fish, it was not possilile to explain on ecological 

 grounds the differences in the structures referred to above. Mr. Hutchinson now informs me that 

 "the s])ecies is moderately common in the Jbelum near Srinagar. and also inhabits some of the 

 sluggish canals around the city." The species would thus appear to be restricted to dee]) and 

 flowing waters. Init a rapid current is not essential for its existence. The deep form of the body 

 can be assrK'iated with lift- in sluggish waters where, it seems prohable, that tlie fish feeds l>y 

 scraping ofl' algal matter from hard objects. During the feecling process the fish pnibablv adheres 

 to the substratum before bringing into action its shar]). lower jaw. This would explain the utility 

 of the adhesive ajiparatus in this species even when living in sluggish waters. 



Attention may here be directed to the mode of life of the fishes of the genus Garra (Mukerji, 

 Acliiiryya Sir P. C. Ray Goiinii. I'ol.. Calcutta, pp. 477-482, 1932) which possess a suctorial disc 

 behind the lower lip. The disc ma\' ha\e originally develojxjd for stemming rapid currents, but 

 some of the species now live in lakes and deeper streams where the sucker is used for holding on 

 to hard f)bjects while the fish scrajies off algal matter from them. There is thus a parallelism 

 between the mode of life of Glyplolhorax kasliiiiircit.\'i.\- and the lake-inhabiting sjiecies of Garra. 



The adhesive apparatus of Glyptotliorax is not a sucker device in the same sen.se as the 

 mental disc of Garra. but is a mechanical device for increasing frictii>n {Phil. Trans. Roy. Soc. 

 London, 1'. CCXN'lll, p. 262, 1930). The ridges are providetl with sharp, curved spines {Rcc. 

 [lid. Mn.<:. XXIV, pp. 55-58, 1922) which fi.x themselves into the tmevennesses of the rocks. 

 When the fish applies its adhesive apparatus to a r(x;k, it is likely that the water flows out of the 

 grooves by adpression and a series of partial vacua is created. The central de])ression in the 

 adhesive apparatus of G. kaslunircn.Kis is probably a device for the production of a partial vacuum. 

 The circular form (tf the ajiparatus and the fact that it is incomplete posteriorly shows that the 

 species is iirobably derived from a member of the highly sjieciali/.ed .group. This again show's how 

 at high altitudes torrential forms revert to primitive mode of life on finding sluggish waters, but 

 it is significant that structures acquired for stemming currents in torrential streams are retained 



