1910.] Cuvier and Vertebrate Palceordology; Dumeril. 67 



CUVIER AS THE FOUNDER OF VERTEBRATE PALAEONTOLOGY, 1796-1836. 



A by-product of Cuvier's early studies m comparative anatomy and 

 mammalogy was his interest in the fossilized remains of animals. As 

 early as January, 1796, he announced to the Institute (Mem. Inst. V^ 

 classe, Math, et Phys., tome II, pp. 20-21) that the fossil elephants which 

 had been known from the earliest times and had formed the subject of a 

 great number of writings were of a species different from the Indian Elephant. 

 This was his first important palseontological discovery. 



Cuvier's first independent contribution to mammalogy was his descrip- 

 tion of the MegatJicrium in the 'Magasin Encyclopedique,' Vol. Ill, An 

 IV (1796). (Palmer 1904, p. 406.) The year 1796 may consequently be 

 regarded as the date of the founding of Vertebrate Palaeontology. This 

 was followed by the first 'Memoir sur les especes d'Elephants vivants et 

 fossiles' (1799), by the memoirs 'Sur le Megalonix . . . . ' (1804), 'Sur le 

 Megatherium....' (1804), 'Sur les elephans vivants et fossiles' (1806), 

 'Sur le grand mastodonte' (1806); and "Sur differentes dents du genre des 

 mastodontes' (1806); the })alseontological researches finally culminating 

 in the famous 'Ossemens fossiles' in 1812, 1821, 182.5, and 18.34-18.36 

 (Hay, 1902, pp. 72, 73). 



DUMERIL, 1806. 



'Zoologie x\nalytique ou Methode Naturelle de Classification des Animaux'. 



Paris, 8vo. 



Dumeril examines in detail the principles of a natural classification. 

 He rejects life-habits as criteria for major classification (contrast Linnaeus) ; 

 he rejects also criteria based solely on the general or external appearance 

 ("superficie"); and finally he rejects classifications based on the variations 

 however slight of a single organ. He declares further that of late it had 

 come to be recognized (cf. Daubenton, p. 41 above), that the principal 

 end of natural history being the study of species, instead of building arti- 

 ficial classifications and then proceeding a priori to study species (cf. Brisson), 

 we should study the latter directly, constantly comparing each with each, 

 and thus gradually recognizing the larger assemblages. From this, he says, 

 springs the natural method of classification, which although "still very 

 imperfect, corrects its own errors each day and tries to fill up the gaps which 

 it sees indicated in advance." Botany (cf. Ray, Linnaeus) may furnish 

 examples and ideals of method, but these are not to be followed slavishly. 

 Dumeril protests against the straining of characters in order to fit some 



