BARBUS (PSEUDOBARBUS) PALLIDUS. 
groupes must be acknowledged to be but very imperfectly known, and badly established. Of the various 
families into which fishes have been distributed, there does not appear to me a second family in which the 
characters assumed as those of gToupes are more vague, or even so vague and imperfect as those which have 
been selected to characterize the lower divisions of the Cyprinidas ; and I cannot, without hesitation, 
say that I believe I have not myself, even in this Number of my Illustrations, contributed to increase the 
obscurity and confusion. Be that, however, as it may, I feel convinced that it is not to the individual 
w-ho possesses only a very limited number of species that the Ichthyologist is to look for improvement and 
reform . his expectations must be turned towards men who have the opportunity of examining and 
comparing large collections of species ; and it is from them that he must expect light as to the proper 
classification. 
With respect to the three subordinate forms which I have ventured to characterize, and of which figures 
are given in Plates X. XI. and XII., I feel fully satisfied only of the legitimate claims of one Abrosto- 
mus ; the other two may eventually be referred to some other of the groupes already instituted, though I 
confess my owii inability to discover in the combined characters of any of such groupes the essentials which 
would justify me in selecting for the South African species positions different from that in which I have placed 
them. When some hundreds of species shall have been got together, the difficulties which are now expe- 
rienced will doubtless be readily surmounted, and it will then be seen whether Cheilobarlm and Pseado- 
lurlus are to be retained as component forms of the genus Barbus , or are to be translated to some other 
genera. At present I regard both of them as forming minor divisions of the genus Barbus , and in my 
opinion are to be regarded as two of its subgenera. 
