84 cypEiNiD^. 



the side of the tail. Anal scales not enlarged. Pharyngeal teeth 

 5 . 4 or 3 . 3 or 2—2 or 3 . 3 or 4 . 5. 



Snout but rarely with tubercles or pore-like grooves. 



Temperate or tropical parts of the Old World. 



Many attempts have been made to divide the Barbels into generic 

 groups, as may be seen from the synonymy. And, indeed, when we 

 consider the great number of the species, and the great apparent 

 diversity between the first and last in the series, a further division 

 must appear to be highly desirable. Yet nothing would be more 

 contrary to the idea of natural genera, the transition from one 

 extreme species to the other being perfect ; and those attempts at 

 generic subdivisions prove that the naturalists who proposed them 

 had only a partial acquaintance with the species. The size of 

 the scales, the development of the third dorsal ray, the form of 

 the snout (and consequently of the prseorbital) are perfectly use- 

 less as generic characters, in consequence of the complete series of 

 intermediate forms. The lips are subject to variation in the same 

 species ; thus, for instance, some specimens of B. hijnni would belong 

 to Barhus, others to Laheobarhus. Even the number of barbels is 

 not quite constant, and much caution is required in determining 

 whether a specimen belongs to a species normally with four or two, 

 or without barbels, as these organs are sometimes quite rudimental, 

 and, abnormally, may be entirely absent : this is principally the case 

 with the upper barbel, which I have foiind sometimes on one side 

 and not on the other. Yet on such characters have naturalists 

 ventured to found genera ! 



Synopsis of the Siyecies. 



I. Normally four barbels. 



A. Scales small or of moderate size, more than forty in the lateral 

 line. 



1. None of the anal rays prolonged beyond the margin of the 



tin ; snout without tubercles or pores *. 



a. The third dorsal ray is osseous and serrated : Europe, 



Western Asia (Africa), p. 86. 



b. The third dorsal ray is not enlarged : Europe (Africa), p. 94. 



2. Snout with tubercles or pores; third dorsal ray osseous, ser- 



rated: India. 



» I have observed in some species that certfiin individuals have semiglobular 

 or conical tubercles on the snout. These tubercles appear to be developed at 

 certain seasons of the year ; and before their development, and after they have 

 dropped off, their bases appear as hollovr impressions or as pores. Other 

 species appear to have such pores only, and tubercles seem never to be deve- 

 loped. Tubercles and pores are of but rare occurrence in the species vrhich I 

 refer to Barhus, in other genera every species appears to be provided with them. 

 Thus, in using this synopsis hi the determination of species, it must be borne in 

 mind that in every species usually provided with tubercles specimens do occur 

 without them. 



