638 Records of the Indian Museum. [Vol. XXII» 



deal of importance is generally attached to the number of fin rays 

 and scales, but in Garra these characters are variable and it is 

 impossible to use them in separating one species from another. 

 Jenkins (1909) relied on these very characters and came to the 

 conclusion that^ "there are no specimens of Discoonaihus in the 

 Indian ;\Iuseum which justify me in considering that there is more 

 than one Indian species of this genus." In both the Indian and 

 African species the general rule is that there are seven to eight 

 branched and two to three unbranched rays in the dorsal, while 

 in the anal there are five branched besides one or two that are 

 unbranched. The number of scales along the lateral I'ne varies 

 from 33 to 44. ;\Iarked deviations from these numbers occur only 

 as abnormalities and minor difl'erences are alwa3's bridged over 

 when a large series of specimens are examined from the same 

 locality. 



In the absence of any well-marked characters, in the number 

 of scales and fin-rays, Giinther (1868) attributed to G. lamia a 

 very wide range extending from "Syria to Assam," and ever since 

 this statement was made, authors in general have attributed an\' 

 species of this genus from any part of this region to G. lamta. 

 In particular Day, who had previously recognised several species 

 from South India, subsequently (1871) referred them all to G. 

 lamta, but later on (1878) insisted on the specific validity of G. 

 jerdoni. 



No less confusion has been caused by a black spot that is 

 present in many species behind the angle of the operculum. Also 

 there is often a series of black spots at the base of the fin-rays of 

 the dorsal fin. 



In certain species of Garra a proboscis is present on the snout 

 and this has been regarded as a secondary sexual character res- 

 tricted to males ; in G. stenorhynchus and G. bicorniita, however, the 

 proboscis is known to be common to both sexes. Very little is 

 known about the variation of the proboscis or the conditions which 

 influence its formation. I have found after examination and 

 dissection of a large series of specimens from all parts of India 

 that wherever a well-developed proboscis is present it is always 

 common to both sexes except in the classical species G. lamta, in 

 which a peculiar proboscis is present in the male sex only. 



The credit of stimulating research in this genus belongs to 

 Boulenger, who recognised many species of Garra from Oriental 

 Africa and pointed out in 1907 that the Asiatic species of the genus 

 were much in need of revision. Annandale (1913, 1919) attempted 

 to revise the Indian species in a series of valuable papers, but 

 unfortunately the old collection of the Indian Museum containing 

 Day's types was at that time interned in Austria, having been 

 sent to Dr. V. Pietschniann before the outbreak of war. 



