98 DR. R. BROOM OX THE [Feb. 7, 



The outer cusp of ihe lower molaT would fit into the hollow on the 

 inner side of the outer cusps of two upper molars. 



The molars of Diademodon are ^compared by Seeley with 

 the molars of Oi'nithorhynchtLS, Ctenacodon, Plagiaulax, and 

 Tritylodon, and by Osborn with the molars of Microlestes. 

 Although there is some superficial resemblance between the 

 tuberculated molars of the multituberculate mammals and those 

 of Diademodon, it seems to me more probable that there is no 

 close affinity between the teeth, and that those of Diademodon 

 have originated in quite a different manner from those of the 

 Multituberculata. The structure of the premolars in Diademodon 

 gives us a clue to the way in which the molars have been formed. 

 There we find an outer and an inner cusp, which we may perhaps 

 call the "protocone" and the " deuterocone." In the molar we 

 find evidences of the same two cusps, but instead of being long and 

 sharp they are here obtuse, and by the sides of the molar we have 

 a number of small other cusps. Like the premolars, the molars 

 are single-rooted, and they bear almost the same relations to each 

 other as do the molars and premolars in Cynognatlms. Diade- 

 modon and the allied GomjjhognatJnis aiid Trirachodon are so 

 closely allied in the structure of the skeleton to Cynognathiis as 

 to suggest that the foims with broad molars are descended with 

 only slight modifications from carnivorous types. Among mammals 

 we occasionally find flat-ci'owned teeth in types closely allied to 

 others with sharp teeth — as, e. g., in the Sea-Otter {Enhydra) 

 and the Common Otter {Lutra^, or in the Bear (Ursus) and the 

 Dog (Canis); and though in Diademodon and Cynognathus the 

 difference probably is greater in- degree, it does not seem to be 

 different in kind. 



GomphognaiJiKs and its near allies are regaided by Seeley as 

 herbivoi'oifs forms, and I aufcuot aware that this view has been 

 questioned by any later worker. There is something, however, to 

 be said against it. Gomphognathths, Trirachodon, a,nd Diademodon 

 have all powerful canine teeth, and in Gemj^hogvathtis and 

 Trirachodon, at least, these are sepai'ated by small incisors. 

 The condyle of the jaw is in a, line with the molar teeth. There 

 is a very large -coronoid process and the temporal fossa is of laxge 

 size. These prove conchisiA-ely that the forms with flat molars had 

 at least powerful temporal muscles, such as are rarely or never 

 found in herbivorous mammals. We have also evidence that 

 Gomphognaihus was able to open its jaws very widely, as in the 

 type specimen the mandible is found open at about 90° without 

 much dislocation of the joint. When we look at the teeth we find 

 that they do not seem suited for a vegetable diet. The second 

 last molar in the above-mentioned specimen of Diademodon must 

 have been for a considerable time in use, but the enamel has not 

 yet begun to wear off; ai:id as the layer of enamel is not thicker 

 than a sheet of notepaper, it will be manifest that whatever it 

 was used to crush it is not likely to have been vegetable fibre. It 

 seems to me pi-obable that Diademodon and Gomphognaihus fed 



