258 The Edinburgh Neiv Philosophical Journal. 



scales preserve the same form, not only in the same individuals but in 

 all the individuals of the same species ; they are essentially different in 

 individuals of another family; they constitute different series of form, 

 series which are very distinct in themselves, but the members of which 

 offer all the degrees of transition which unite them to each other. 

 Scales may therefore serve as a natural character in the description and 

 classification of fishes. 



"The merit of having first called the attention of naturalists to scales 

 is due to M. Agassiz; but being deprived of the aid of the microscope, 

 not only the organization of the scale remained concealed from him, 

 but he inevitably also confounded scales of the most different forms. 

 It is thus that M. Agassiz states,* that, in the family of the Cyprini, 

 ' all the body is covered with scales formed by a pretty considerable 

 number of plates with the edges smooth and entire ; grooves or furrows, 

 more or less numerous, extend from the centre of growth to the edge 

 of the scales,' &c, and he arranges the Cyprini in the family of the 

 Cycloides ; but these furrows are canals — these entire and smooth edges 

 of the plates are nothing else than hues resulting from the fusion of the 

 cells, lines which are repeated on all the scales. We thus see that, 

 from the organization not being well known, a characteristic description 

 of scales could not be given, and some were confounded with others 

 which yet offered true marks distinctive of families. We find an exam- 

 ple of this in the memoir of M. Agassiz above cited. 



" This distinguished savant supposes (1. c. p. 48,) that the relations 

 which connect the Muges and Atherines with the Cyprini have entirely 

 escaped Cuvier, on account of the too great importance which this 

 celebrated naturalist attached to the presence or absence of spiny 

 rays on the back. M. Agassiz had therefore to seek some character 

 common to all these fishes, that he might be able to bring them 

 together, and he says that he discovered this character in their scales, 

 which are all composed of plates of growth with entire edges, and 

 which he calls Cycloides. 



" Now nothing strictly can be more different than the scales of the 

 Muges, Atherines, and Cyprini. The difference is so great, that it 

 alone determined us to place the Atherines in a separate family, 

 a measure which had previously been hinted at by Cuvier, and which 

 he would not have hesitated definitively to adopt, had he known the 

 particular characters presented by their scales. We have seen with 

 pleasure that our opinion is supported by Professor Nordmann, who 

 makes an entirely distinct family of the Atherines, as well as of Muttus. 



* Memoirs de la Societe des Sciences Naturelles de Neufchatel, t. i. p. 34. 



