THE TEETH 451 



widely distributed in the middle and upper eocene, occurring in 

 nearly all perissodactyls and creodonts. With some special modifica- 

 tions it persists to the present time in the sectorial of the Carnivora, in 

 many Insectivora, and in some forms of Artiodactyla (e.g., Dicotyles). 



" The final step in the conversion of premolars to the molar pattern 

 is given by the addition of a fourth main element at the postero- 

 internal angle of the crown, the tetartocone, which thus corresponds 

 in position to the hypocone of the molars (Fig. 370, O, H). 



" The development of the inferior premolars appears to be some- 

 what less regular and constant than that of the superior. As in the 



upper jaw, the complication begins with - and advances anteriorly, 



but it is worthy of notice that in many forms the complication of the 

 inferior molars begins earlier and proceeds farther than in the case 

 of the upper teeth. As before, the initial point must be taken as a 

 simple conical cusp, the protoconid (Fig. 370, P). Most of the 

 existing unguiculates, as well as some recent and many extinct ungu- 

 lates, retain more or fewer teeth which depart but little from this type. 

 In many forms the only addition to the protoconid consists in a small 

 posterior basal cusp, which the analogy of such Mesozoic mammals as 

 Amphilestes and Triconodon justifies us in regarding as the equivalent 

 of the metaconid of the molars (Fig. 370, Q). Frequently also an 

 anterior basal cusp, strictly comparable to the paraconid of the molars, 

 is added (Fig. 370, R), and a stage like that of the Triconodon molar 

 is attained, consisting of elements which there is every reason to 

 regard as homologous with the three primary cusps of the molars. 

 There is, however, a great difference as to the regularity with which 

 the paraconid and metaconid are present and in the order of their 

 succession ; one or the other of them may never appear at all, and 

 while the metaconid is more frequently present and generally makes 

 its appearance first, yet this is by no means invariably the case. 

 Another difference from the molars consists in the ultimate fate of 

 the metaconid in the molariform premolar, where it becomes either 

 part or the whole of the talon (Fig. 370, T) and always remains on 

 the same antero-posterior line with the protoconid, instead of shift- 

 ing to the internal or lingual side of the latter. In the premolars, 

 therefore, when a cusp occurs occupying the position taken by the 

 metaconid in the molars, it cannot be regarded as homologous with 



