1910.] lAitcr Steps in the Establishment of tlie Tritubercular Mohir. 189 



V-shaped mefaconule. This cusp was at first developed to fit into the space 

 between the antero-external border of the talonid of the corresponding 

 lower molar and the anterior proto-paraconid edge of the frigonid of the 

 following lower molar (Fig. 12, C). It would also help to press the food 

 against the hypoconid when that cusp appeared. 



Mechanics of tritubercular evolution in the Cretaceous and Basal Eocene. 



The Jurassic Dryolestes represented a stage in which the upper molars 

 were more than twice as many (8) as in the higher mammals (3) but corre- 

 spondingly narrow, especially on the inner side. Amphitherium was some- 

 what more progressive both in the reduction of the number to six and 

 probably in the somewhat greater antero-posterior diameter of the individual 

 teeth (Fig. 12). The chief advance in the dentition of Cretaceous and Basal 

 Eocene mammals was the reduction in 7iumber and concomitant increase in 

 antero-posterior diameter of the upper teeth and the consequent development 

 of a more open triangle and of the cusps on the posterior half of the tooth. 

 This also required a lengthening in the lower molars, especially of the talonid 

 (Fig. 12). 



With increasing size and power of the individual teeth the jaw and muscu- 

 lar attachments became larger and the opposite rami became more strongly 

 connected at the symphysis. We may assume that the outer side of the upper 

 molars widened faster, especially the metastyle, the external cingular cusps 

 greatly developed, while the para- and metacones in some phyla became 

 crescentic. The protocone is also crescentic in all the earliest tritubercular 

 types, so that the bunodont, or purely circular cusps, formerly assumed to 

 be the starting point in many orders, may be secondary (Gregory, in ( )sborn, 

 1907, p. 173, foot-note). 



This stage of evolution is represented in Pediomys Marsh (figured by 

 Gidley, 1906, pi. v, fig. 6), by Didelphops Marsh and by other Upper Creta- 

 ceous mammals figured by Osborn (1907, p. 96, fig. 47). 



Origin of the hypocone. — The upper molar in widening (phylogenetically), 

 widened faster at the base than at the summit of the crown, consequently 

 a line of fracture took place near the base of the crown, and an internal basal 

 cingulum, resulted, which was analogous to the older external basal cingulum. 

 The develo})ment of the hypocone from this structure has been shown in 

 many families by Osborn. The antero-external edge of the trigonid, namely 

 the protoconid-paraconid ridge, sheared past the posterior edge of the pre- 

 ceding upper molar. Consequently upon the appearance of the posterior 

 internal basal cingulum of the upper molar the anterior edge of the trigonid 

 now began to overlap this posterior cingulum, a stage illustrated in the Basal 



