150 HISTORY OF THE WORKS OF CUVIER. 



or imder the pectoral ; teeth phiced by turns on the intermaxillaries, the maxiUarks, 

 the vomer, iha palatine, the tongue, the arclies of the hranchi(C, &c.; the 

 form ol" these teeth, jj?afe-Z?7je, viUiforv!,pomfed, &c. ; the opercnla or covers of the 

 branchiae, whether smooth, scaly, serrated, sharp-jmntcd, and armed icith spines, 

 or ohtttse and icithout spines, &c. ; and it is on the varied combination of 

 these differential features or characters that all the different methods which have 

 been successively imagined for the classification of fishes depend. 



It will be readily understcxl that he who should employ but one or two of 

 these characters woidd have only an artificial or incomplete method, like Linnteus ; 

 that he who should indistinctly employ them all would have only a confused 

 method, like so many ichthyologists, and that the natural method — that is to say, 

 the exact and complete one — consists at once in employing them all, and in 

 employing none but according to the relative order of its impiortance. There are 

 two points, in effect, which control the whole idea of a natural method : one to 

 emplov only true characters ; the other to accord to each of these characters only 

 tlie precise degree of its importance. But, to employ only true characters — that is 

 to say, not to attribute to such or such a species such or such a character which 

 it wants, and, reciprocally, not to suppose it destitute of such or such another 

 which it possesses — we perceive that it is necessary to know all the species. On 

 the other hand, to attribute to each character only the degree of its importance, 

 we perceive that this complete knowledge of species, in itself so vast and so diffi- 

 cult, still would not suffice, and that it is, moreover, necessary to have compared 

 these characters under all their relations ; that it is necessary to have varied, multi- 

 plied, exhausted all their combinations. 



Now, on these two points which constitute, in fact, the whole of ichthyology — 

 that is to say, both for the determination of species and the valuation (f the 

 characters according to which those species are composed or discriminated up to 

 the time of M. Cuvier — everything was yet to be done. The species of fishes 

 were not known; the proof of it is in every page of the book we are considering. 

 No just idea existed of the characters which decide their union or distril)ution, 

 the proof of which is found in those perpetual transpositions undergone by the 

 same species in the difierent classifications of authors. * * * * 



IV. — LECTUEES ON COMrAKATIVE A^^ATOilY.* — LAWS OF ANIilAL OKGAX- 



IZATION. 



Two great laws control and comprise all the others ; the first is that of organic 

 correlations ; the second tliat of the subordination of organs. 



A necessary correlation binds all the functions one to another. Respiration, 

 when executed in a circumscribed respiratory organ, cannot dispense with circu- 

 lation, for it is necessary that the blood should arrive at the respiratory organ, 

 the organ which receives the air, and it is the circulation which carries it thither; 

 circulation cannot dispense with irritability, for it is irritability which deter- 

 mines the contraction of the heart, and consetpiently the movements of the blood ; 

 muscnlar irritability, in turn, cannot dispense with the nervous action. And if 

 one of these things is changed, it is necessary that all the others should change. 

 If circulation is wanting,^ the respiration can no longer be circumscribed ; it is 

 necessar}^ that it should become general, as in insects; the blood no longer seek- 

 ing the air, it is necessary that the air should seek the blood. 



There are organic conditions, therefore, which call for one another ; there are 

 such as exclude one another. A circumscribed respiration necessarily demands 

 a pulmonary circulation ; a general respiration renders a pulmonary circulation 

 useless, and excludes it. Everytliing is thus regulated by necessary relations. 



* Two volumes of this work appeared in 1800, three more iu 1805. lo the former, M. 

 Cuvier had for his coadjutor M. Dumeril ; in the latter series he was assisted hy M. Duvernoy, 

 who is publishing at the present time a second edition of the entire work. 



