150 



the penultimate lower milk molar, which in this as in most groups 

 represents but the half of a true molar, furnishes opposition to its 

 most anterior surface. Although it is not always literally true, that 

 in the Artiodactyla the premolars represent each but the half of a 

 true molar, and in the Perissodactyla an entire one, it is certain 

 that in the exceptional cases among the former group, the parts re- 

 presenting the posterior division of the tooth are small, or merely 

 rudimental ; and that in the latter group, it is only in the most an- 

 terior of the series that the posterior portion of the tooth is ever 

 altogether wanting. It is also certain, that all those genera of which 

 the milk dentition has been seen, conform in that particular to the 

 general character, the distinction being well-marked in the Artio- 

 dactyla between the two last upper milk teeth, whose characters are 

 those of true molars, and those which precede them and represent but 

 half ones, the same difference also prevailing between the last and 

 those which precede it in the lower jaw ; always necessitating the ex- 

 istence of a third pair of tubercles in the last lower milk molar to 

 work in the interval of the two pairs in the penultimate above ; while 

 in the Perissodactyla, the constant existence of a well- developed pos- 

 terior pair of lobes in the penultimate lower milk tooth abrogates the 

 necessity of a third pair in the last one, and consequently we need 

 not expect to find it, even in those genera, such as Lophiodon and 

 Palceotherium, of which the additional lobe to the last true molar is 

 characteristic. Of the first-named genus, the milk dentition, so far 

 as I am at present aware, is as yet unknown ; but among the plates 

 in the 'Ossemens Fossiles' examples may be seen of the lower jaws of 

 young Palseotheria, exhibiting the milk teeth, of which the last has but 

 two lobes*. Therefore the tripartite condition of this tooth becomes 

 a constant and important character of the Artiodactyle division. 



INIost of the characters which separate the Ruminant and Non-ru- 

 minant divisions of the Artiodactyla have been pointed out in my 

 former paper, as well as those which distinguish the two subdivisions 

 of the Hog-tribe, which by the analogy of the amount of difference in 

 those of other groups, I think must be looked upon as families, — 

 SuidcB and HippojJotamideB. The striking character derived from the 

 sudden termination of the pterygoid ridge in the Ruminant, and the 

 formation of the pteiygoid fossa in the other division, has been alluded 

 to abovef. The considerable upward extension of the masseteric ridge 

 upon the os malse beneath the orbit seems also characteristic of the 

 Ruminants, as well as the bifurcation of the orbital ala of the sphenoid, 



* PL 4. fig. 1 (alluded to by Professor Owen), and pi. 56. fig. 2. 



t In the Hippopotamus the pterygoid ridge runs inwards and even a little back- 

 wards, and then forms a slight angle at the point of junction with the pterygoid 

 process, which then runs downwards and forwards, so that the outer wall of the 

 fossa exists as in the allied forms, while, as I have before observed, it is the inner 

 one which is wanting. I must again refer to the remarliable osseous bulla within 

 the orbit of this animal, since I find that the same thing exists, though of much 

 smaller size, in most ruminants ; in many skulls it is broken away, and when re- 

 maining it so lies upon the " tuberosity" or posterior termination of the alveolar 

 process of the maxillary bone as to appear at first like a part of it. It opens into 

 the nose and anti um maxillare, and has no connection with the lacvymal apparatus. 



