On the Osteology of the Hyopotamida?. 179 



to the same culminating point. In reaching it the lateral digits 

 will be entirely lost, the trapezium will coalesce with the magnum, 

 and the second cuneiform with the third ; the middle metacarpals 

 and metatarsals will coalesce into a complete cannonbone ; and 

 probably the stomach will become still more complicated, and they 

 will ruminate. That this state is the goal towards which the 

 Suina tend I have little doubt ; but it is more than probable 

 that man by his influence will prevent them from ever reaching 

 it. 



Our task is more difficult when we come to inquire into the line 

 of descent which has given rise to the Ruuiinantia. As stated 

 before, I cannot put the Anoplotherium, nor the Xiphodon, in their 

 pedigree. In my opinion, the line which ends in Ruminantia 

 branched off from the small tetradactyle Hyopotamidce, which were 

 so numerous in the Eocene period. I find in the Eocene of Maure- 

 mont all stages of transition between the five-lobed upper molars 

 of these Hyopotamidae and teeth having a true ruminant four- 

 lobed pattern ; these last have belonged to some small species of 

 Dichodon. Unfortunately we have no clue to the skeleton, though, 

 seeing the tetradactyle living Hyomoschus, it may fairly be assumed 

 that these early progenitors of Ruminantia were also tetradactyle. 

 The small tetradactyle Cainotherium is a very tempting genus in 

 speculations about the descent of Ruminantia ; but I must exclude 

 it for many reasons, though I cannot here give them in fall. Some 

 of these are as follows : — the Cainotherium retained till the Middle 

 Miocene five-lobed teeth on the Dichobune pattern (with the three 

 lobes on the posterior half of the tooth), while we have truly ru- 

 minant teeth already in the Eocene ; it retained its upper incisors 

 and free metatarsals, while the much older Gelacus, Aym., which 

 is already a true ruminant, had no upper incisors and the meta- 

 podials were confluent in the adult. Cainotherium seems to be a 

 direct descendant of Dichobune, and to have become extinct, without 

 leaving any successors. 



Supposing that the Dichodon had a foot true to the tetradactyle 

 type, we do not find the earliest stages of reduction ; they were 

 passed rapidly, and in very ancient times : but there can be little 

 doubt that the Ruminantia began with a tetradactyle foot, and 

 ended by a cannonbone adapted to the whole distal surface of the 

 carpus and tarsus. Such adaptation of the two middle digits could 

 not be obtained at one leap ; and certainly all stages between a 

 tetradactyle foot (in which every digit was supported by a separate 

 bone in the carpus and tarsus) and a didactyle foot (in which 

 the two enlarged middle digits have taken the whole distal surface 

 of all the carpal and tarsal bones) were passed by this group in 

 the same manner as we have seen it in the Suina ; but only a few 

 traces of this passage remain. Erom the tetradactyle Dichodon, 

 the group of adaptive Selenodonts may be said to have split into 

 two subordinate groups. In one of these, represented by the 

 Hyomoschus, the lateral digits are retained, and only the metatarsals 

 become confluent, while the two middle metacarpals continue to be 



12* 



