1893.] DB. C. J. FORSYTH MAJOK OK MIOCENE SQI^IEEELS. 201 



form by the addition of the so-called lieel or talon to the original 

 trigon. 



The name of lieel or talon is borrowed from the teeth of Car- 

 nivora, where this part, as the name implies, generally appears 

 in a reduced form compared with the rest of the tooth. In 

 other orders the so-called talon is, as a rule, a well-developed 

 part of the inferior molar, so that it seems a priori highly improb- 

 able to consider as a later development that part which, in the 

 majority of Mammalia, constitutes the whole posterior moiety 

 of the lower molars. What we at present know of the oldest 

 forms of Perissodactyla, Condylarthra, Eodeutia, and even of some 

 forms of Creodonta, as revealed especially by the Cernaysian fauna 

 of Eeims\ does not in the least justify a similar assumption. On 

 the contrary, the " talon," far from showing a tendency to dis- 

 appear, is in several of these archaic Eidheria very prominent, and 

 even more distinctly developed than in any later form, not only 

 in longitudinal extension, but partly even in the elevation of the 

 cusps, as compared with those of the anterior moiety. 



The question of the heel leads us to an objection made by 

 Fleischmann, who on the whole accepts the tritubercular theory, 

 but maintains that the cusps of upper molars are not directly 

 homologous to those of tlie lower molars ; or, in other woi'ds, that 

 not only does the intei'ual side of upper molars correspond to the 

 external side of lower molars, as admitted by Cope and Osborn, 

 but that, besides, the anterior part of upper corresponds to the 

 posterior part of lo«er molars '". 



In reply to Fleischmann, Osborn stati-s that "this objection 

 w ould be fatal to a uniform system of nomenclature for the upper 

 aud low er cusps if it could be sustained," but that it is disproved by 

 a comprehensive survey of the Mesozoic trituberculates, especially 

 of the Amblotheriidfe and Spalacotheriidse \ ISTeitlier Osborn nor 

 Fleischmann seem to be aware that, if the latter is right, his 

 objection will be fatal not only to the homology of upper and low er 

 cusps, but also to the theory, for the primitive trigon which in- 

 cludes the protocouid, the supposed homologue of the reptilian cone, 

 would in that case be found to be formed for the greater part by 

 the very heel which the theory considers to be a late addition. 



There can be no doubt as to the correctness of Fleischmann's 

 statement, which is easy to verify. A left upper anterior milk- 

 tooth of Dvlelphis, for instance, is at first sight very difficult to 

 distinguish from one of the right lower series. Even in such 

 specialized molars as those of modern Eumiuants, in holding side 

 by side a right upper and a left lower molar, or vice versa, \^■hat 

 appear to be the mutual homologies are to be traced out even to 



' Lemoine, •'Etude d'ensemble sur les dents des Mam uii feres fossiles des 

 environs de Keims," Bull. Soc. Geol. de France, iii'' strie, t. xix. 1891, pp. 

 263-289, pis. X., xi. 



^ A. Fleischmann, " Die Grundform dei- Backziibne bei Saugethieren und die 

 Homologie der einzelnen Hooker " (Sitzungsber. d. k. Preuss. Akad. d. Wiss. 

 Berlin, 1891). 



^ Osborn and Wortm.in, /. c, pp. 84, 8.5. 



