284 PROCEEDINGS OF THE ACADEMY OF [1887. 



premolar early replaces a violariform milk tooth di and the fourth 

 true molar is very late in coming in. Trlacanthodon is thus a 

 synonym of Triconodon. (4) Phascolestes. There is no doubt that 

 this genus is distinct from Peralestes,^ but it is a question whether 

 the type mandible does not belong to a genus near Stylodon. Form, 

 i 4 c 1 pm 4 m 8. (5) Leptodadus, is widely separate from Stylodon.* 

 Its molars, Avith single recurved cusps and elevated heels are whollj 

 unique. The post-canine formula was probably pm 4 m 6. (6) 

 Peramus is also widely separated from other genera in its dental 

 formula, pm 6 m o; all other polyprotodont genera of this period 

 having four or more molars. The teeth in the type of this genu3 

 are seen upon the outer surface only. It is probable that when the 

 niner surface is known, the formula will be modified to pm 4 m 5. 

 (7) In the genus Spalacothermm the molars and premolars are well 

 differentiated; the formula is i ? 2, c 1, pm 4, m 6. (8) PeralesteS 

 probably has a closely related form in Peraspalax, in fact the generic 

 distinctness of the latter is doubtful. The post-canine formulae are, 

 Peralestes pm 5 m G; Peraspalax, pm 4, m 7. (9) The maxilla 

 which was referred to the genus Stylodon by Professor Owen, must be 

 removed to a distinct genus, Athrodon, characterized by the compact 

 position and peculiar wearing pattern of the crowns. (10) The 

 maxillary formula of Bolodon is found to be i ? 2, c 0, pm 3, m 4. 

 The characters of these teeth are clearly shown in the accompanying 

 cut. 



Leaving out of view for the present, all consideration of relar 

 tionships to recent forms, we find that the Mesozoic Mammalia di- 

 vide into two larger groups. In the first group (I) one of the incisors 

 is greatly developed at the expense of the others and of the canine; 

 there is a diastema varying in width in front of the first premolar 

 and the true molars are invariably characterized by two or more 

 antero-posterior rows of tubercles separated by longitudinal valleys 

 or grooves. For this group we may adopt the sub-order Midtituher- 

 culata, proposed by Professor Cope.^ 



The second group (II) does not show such close internal rela- 

 tionship among its members as does the foregoing, but is well separ- 

 ated from it by such characters as the following: 



The incisors are numerous and subequal in size, the canines are 



3. Professor Owen separated this <^eiuis doubtfully from Peralestes. 



4. Professor Owen placed it doubtfully near this genus. 



5. .\merican Naturalist, lSS-1, p. 687. 



