186 Bulletin American Museum of Natural History. ["\'ol. XXVII, 



was struck on the posterior basal side by the protocone of the corresponding 

 upper molar (Fig. 11).^ The anterior external face of the protoconid sheared 

 past the posterior side of the upper tooth, and since both protocone and 

 protoconid were wedge-shaped, the relation of reversed cutting triangles, 

 which is so indirectly effected under the Tritubercular theory, may have been 

 established at a very early period in the evolution of the tritubercular denti- 

 tion. 



How on this hypothesis did the trigonid arise ? Possibly in response to 

 the lateral movement of the posterior part of the lower jaw (p. 119) a heavy 

 basal cingulum appeared on the outer side of the u])per, on the inner side of 

 the lower, molars. This cingulum in the lower molars rose into an anterior 

 prominence, the future paraconid, a median prominence, the metaconid, 

 and a postero-internal prominence, the entoconid." The small paraconid 

 and metaconid were lingual to the line of the upper protocones and did not 

 oppose any cusps, but merely aided in holding the food in place for the 

 piercing-crushing action of the protocone and protoconid. We now have a 

 lower molar only a little simpler than that of Paurodon, as figured by Gidley 

 (1906, pi. V, fig. 5), or of Peraspalax as figured by Osborn (1888, fig. 10, No. 



9). 



The paracone and metacone sprang uj) in situ on the anterior and poste- 

 rior external slopes, respectively, of the protocone. In Peralestes the meta- 

 cone, being relatively very far internal, may have assisted in holding the 

 food for the protoconid of the succeeding lower molar which was immediately 

 posterior and internal to it. 



Thus the trigon above and the trigonid below, , although having the 

 reversed relations postidated in the Theory of Trituberculy, have probably 

 not arisen in accordance with that theory: the paraconid and metaconid 

 below may be upgrowths of the internal cingulmn, the paracone and meta- 

 cone, above may have grown up from the slopes of the protocone, in situ, 

 and well within the line of the parastyle and external cingulum. 



The conditions in Dryolestes are foreshadowed to some extent in the 

 Upper Triassic Diadernodon (Fig. 12, .1) in the following respects: 



(1) In Diademodon the small lower molars fit hetiveen, the upper molars, 

 the anterior half of the lower molar engaging the posterior half of the pre- 

 ceding upper molar, the posterior half of the lower molar engaging the 

 anterior half of the corresponding upper molar, after the manner of a talonid 

 (Broom, 1905.4). 



1 This hypothesis does not assume that " acquired" characters are transmissible but simply 

 that changes have often originated as if in response to mechanical stimuli and requirements. 



2 The latter cusp in the writer's opinion is older than the hypoconid, which later grew up 

 externally to the tip of the protocone (in the closed position of the jaws) and fitted into ttie 

 centralhasinof the trigon of the upper molar {cf. Fig. 12, D). 



