1871.] SURGEON F. DAY ON INDIAN FISHES. 637 



If the last is divided to the root, some observers, as Hamilton 

 Buchanan and Bleeker, count it as two ; others, as Dr. A. Giinther, 

 consider it, and, I believe, correctly so, only one. Of these ten 

 branched rays, H. Buchanan observes, " the last of them being 

 divided to the root." Deducting one from ten, I see no other result 

 possible than nine, or nine branched rays. Turning to the original 

 drawing, in the library of the Asiatic Society of Bengal, there are 

 oidy nine branched rays, counting the last divided to its root as 

 one ; consequently my statement was perfectly accurate, whilst Dr. 

 Giinther, failing to determine Hamilton Buchanan's species, which 

 he considered to have fen branched dorsal rays, placed it as a Cir- 

 rhina*, thus affording an excellent illustration of the untrustworthi- 

 ness of arbitrarily splitting species into genera, solely because of the 

 existence of nine or ten branched rays in the dorsal fin. As regards 

 Crossochilus rostratus, Giinther, from the description as now given, 

 it appears to resemble Cypr. bata, H. B., excepting in having a pair 

 of rostral instead of a pair of maxillary barbels, the species being 

 defined from a single immature specimen 4 inches in length. 



Dr. Giinther likewise observes (/. c. p. 136) that, having found 

 Barbus sophore, H. B., in the Calcutta Museum without any label, 

 I had " nevertheless supposed it to be the type of the species, I. c. 

 p. 376." This, however, being inaccurate, may be a misprint; for I 

 do not use the term " type" at all. 



In the 'Record' (p. 127), Dr. Giinther states that Hamilton 

 Buchanan's "drawings exist in triplicate, one copy being in the 

 British Museum." At p. 136, he continues respecting my remarks, 

 P. Z. S. 1869, p. 373 : — " Barbus beavani. Mr. Day thinks that this 

 might be Cyprinus chagunio (H. B.), /. c. p. 373 [but a fish de- 

 scribed as having large scales and minute barbels is not likely to be 

 B. beavani]." 



Amongst Hamiltou Buchanan's original drawings one of C. cha- 

 gunio exists, and is labelled as such ; it is 9| inches in length, and 

 is a very fair representation of the species. The drawing gives forty- 

 one scales along the lateral line, only six less than exist in nature. 

 The rostral barbels are delineated as long as the eye, and the maxil- 

 lary slightly longer. In the ' Catalogue of Fishes in the British Mu- 

 seum,' Dr. Giinther placed " 1 Cyprinus chagunio, Ham. Buch.," as 

 a doubtful synonym of Barbus clavatus, wherein he gives forty-two 

 scales to the lateral line, and "barbels well developed." 



There are several omissions in the 'Record ;' but on them I do not 

 propose offering any remarks, as they are mostly concerning facts 



* The genus Crossochilus, as defined by Giinther, has "dorsal fin without 

 osseous ray, with not more than nine branched rays ;" the existence of ten would 

 cause the fish to become a Cirrhina, according to the Catalogue (at least, as this 

 species was erroneously considered to have such, it is under that genus), which, 

 however, is defined as ''dorsal fin without osseous ray, with from thirteen to 

 seventeen rays." Now if the C. bata had ten branched ones and two unbranched 

 ones, or a total of twelve, I cannot see how it comes to be included in either, as 

 the definition of the genera would have to be altered or a new one created for 

 its reception. 



