KiSHts. 389 



slight separations, but where it is impossible to trace circuniscripticns 

 perfectly defined, and which did not enter at any point into each 

 other. 



But this is not exactly the case with the lophii, the batrachus, the 

 gobies, the blennies, and especially the lalirus; their characters are 

 sufficiently precise, and though partly anatomical, can be easily 

 assigned and appreciated. The small 0])ening of the branchial in the 

 first of these groups; its pectoral fins, tlie base of Avhich is prolonged in 

 the form of an arm ; tlie pectorals alike joined to the ventrals and 

 to three rays of the second ; the flexible rays of the back of the 

 third and fourth ; the fleshy lips of the fifth ; the total al sencc of 

 csecal appendices, in almost all these genera, separate them from 

 the other acanthopterygians ; and tiiis last character will bring them 

 near even the silures and cyprins, the families of which begin the 

 order of the malacopterygians, which, on their part, as we have 

 already observed, resemble the acanthopterygians in the spinous form 

 of some of their rays. 



The families of tlie malacopterygians present more differences, and 

 traits more easy to be recognised ; there are many of them as natural 

 as they fixed in their limits, so strongly does each, in clearly separat- 

 ing from the other, preserve intcni?;lly a great resemblance of details. 

 This fixity is so remarkable, that most of the natural families which 

 we shall establish in this part of the class, have been already observed 

 by Artedi, and presented under the name of genera. His silures, 

 cyprins, salmons, eels, and pikes may remain together ; there is even 

 no inconvenience from distributing them according to the presence and 

 position of the ventrals ; for this character, though slight, does not 

 vary in any of them : I have merely remarked, that it impossible to 

 retain the distinction of jugulars, thoracics, and abdominals, in the 

 terms in Avhich it is established by Linna?us*. It indeed matters 

 little, whether the ventral presents itself a little before or behind the 

 pectoral, or exactly under it ; but the important circumstance, and 

 one which depends upon the very structure of the fish, is to know if 

 the pelvis is attached to the bones of the shoulder, or if it be but 

 merely suspended in the fiesh of the belly. I have, therefore, devised 

 the word subbrachian, to designate fishes of the first category, what- 

 ever be the point at which their ventrals ajipear, which merely depends 

 on the greater or lesser length of the bones of the pelvis ; and I have 

 left the name of abdominals to those of the second. The apodes are 

 naturally the malacopterygians without ventrals. 



We shall, therefore, commence the history of fishes with tlie 

 acanthopterygians, \\hich, in reality, constitute almost but one im- 

 mense family. After these, we shall place the dltferent families of 

 malacopterygians in the order in which they seem to us to come 

 nearest to to the acanthopterygians ; but I do not wish to have it 

 supposed that they draw nigh them, only on one line, and in onLy 

 one series. 



If the abdominal malacopterygians may be thus ranked, and even 



* Those who have united the thoracics and jugulars, have only d ue so from my 

 Animal Kingdom. 



