900 THE SHAPES OF TEETH [ch. 



interlock, rather than meet and oppose* ; in herbivorous animals the 

 molars grind one against another, and wear their crowns away. 

 The teeth of ungulates have been studied with especial care by the 

 palaeontologists on the basis of Cope's well-known tritubercular 

 theory, and one is greatly daring who ventures to deal with them 

 in a different wayj. The case is neither plain nor easy. We are 

 accustomed to speak of a "tooth" as a single unit, however com- 

 plicated it may be ; but we may err in doing so, and we encounter 

 other difficulties in studying teeth whose crowns are worn away, 

 and in interpreting the "patterns" which successive stages of wear 

 and tear expose. 



The elephant's molar is manifestly composite. We see on its 

 worn surface a long succession of "enamel ridges," each marking 

 a narrow ring or island, lying transversely, filled with dentine, sur- 

 rounded by interstitial cement, and with a root or roots of its own. 

 The molars develop one after another during the animal's lifetime; 

 and each consists, to begin with, of so many separate island-elements, 

 not yet cemented together nor worn down, each with its own roots, 

 its own covering of enamel and its own transversely cuspidate crown. 

 These are true dental units, the primitive individual "teeth", corre- 

 sponding to the still simpler teeth of the dolphin ; and they illustrate, 

 and go far to confirm the view that the molar tooth is formed, both 

 here and elsewhere, by "concrescence "J. These rudimenta dentium, 

 as old Patrick Blair called them, or denticules as Owen did, soon 

 fuse together, and begin to wear down as soon as the great composite 

 tooth rolls forward and emerges from the gum. As each denticule 

 begins to wear away, it first appears as a transverse row of separate 

 rings, the so-called column^, which represent the cusps of the original 

 crown and vary in size, number and proportion with the species. 



* This is precisely what Aristotle means when he describes the dog's teeth as 

 car char odont, br sharklike, i.e. interlocking — Kapxcpobovra yap €<ttlu 6aa eTraXXdrrei 

 TOLS 686uTai Ta<r o^eTs, H.A., ii, 501 a 18. 



t See E. D. Cope, loc. cit. and H. F. Osborn, passim. Cf. also W. K. Gregory, 

 A half century of trituberculy, Proc. Am. Phil. Soc. lxxiii, pp. 169-317, 1934, who 

 says that "even the most complex molar patterns of the Ungulates are referable 

 to the trituberculate type, in strict accord with the steps postulated by Cope and 

 Osborn." 



J A view held by Gaudry, Giebel, Kiikental and others, but stoutly opposed by 

 Cope and Osborn, who see in the molar tooth a single unit, complicated by "dif- 

 ferentiation". 



