32 



THE MERYCOIDODONTID/E 



Dentition: Very brachyodont. Superior premolars simpler than those of Merycoidodon; 

 molars with small protoconule and well-developed cingula, not always internal, and with posterior 

 border of protocone of M 2 extending backward rather than inward to meet wall of hypocone. (The 

 term hypocone is used in this monograph to designate the posterointernal cusp of the molar teeth. 

 Some authors consider this cusp as an enlarged metaconule rather than a true hypocone. If this 

 proves to be correct, then Dichobune and a few of its allied forms appear to be the only selenodonts 

 with a well-developed hypocone. In this event, read metaconule for hypocone in the following 

 descriptions of the dentition of the species of Merycoidodontidse.) Inferior premolars simpler in 

 construction than those of Merycoidodon, Pj caniniform, C incisiform; molars with conical inner 

 cusps and wide, open valleys. 



Scott (1899, p. 88) described the DP 4 as follows: 



The deciduous premolar is completely molariform, but has more the molar pattern of Agriochcerus than 

 have the true molars of Protoreodon. This approximation to Agriochcerus is manifest ( 1 ) in the greater con- 

 cavity of the external crescents; (2) in the more massive and rounded shape of the anteroexternal buttress; (3) 

 in the greater breadth of the external median buttress, which is not compressed as in the true molars, but is 

 invaded by the valley, just as in Agriochcerus, though the buttress is much less prominent than in the latter. 



Skeleton: Somewhat more primitive than that of Merycoidodon but essentially like it, except 

 for longer manus and pes. P. medius: length 1 130 mm.; height at fore limb 414 mm.; height at 

 hind limb 442 mm. 



Discussion: Of the five species in this genus so far described I should align P. pumilus, P. 

 parvus, and P. paradoxic™ as nearer to the stem stock, while P. medius is farthest away from it, 

 with P. minor holding an intermediate position. All these species are distinctive, and yet certain 

 characters overlap. P. parvus and P. fumilus are the nearest, but even so the teeth differ in that the 

 posterior crescent of P 2 and of P 3 is better developed in the former, while P 1 is isolated only in the 

 latter. In P. paradoxicus the premolars are simpler in that P 2 and P 3 have no anterior crescent and 

 the posterior one is very poorly developed. P 1 is partially isolated. The parastyles and mesostyles 

 are more prominent on the molars than in either P. parvus or P. fumilus. The P 2 of P. minor has 

 a faint posterointermediate crest and small anterior and posterior crescents. The median crest is 

 very weak; P 3 has the very small anterior and posterior crescents about equal, while P 4 has no 

 internal cingulum. The styles on the molars are small, and the tooth row is closed, as it is in 

 P. parvus. P. medius has P 1 isolated, as in P. fumilus; P 3 has two small anterointermediate crests, 

 which are reflected in P. fumilus by the tiniest of ridges, not more than mere striations, and these 

 two crests are present in the anteroexternal corner of P 4 in P. medius but not in P. fumilus. In the 

 former, P 4 has no cingulum and is more subtriangular, both characters that are unlike those in 

 P. pumilus. The premolars of P. medius are more complicated and appear more like an aberrant 

 branch from the main stock. From the evidence of the teeth, either P. parvus or P. pumilus is 

 nearest to what we should perhaps expect the ancestor of Merycoidodon to be, so far as now known in 



this country. 



A consideration of the molars shows that those of P. parvus are more nearly square and those 

 of P. paradoxicus are the widest in transverse diameter in relation to length, while those of P. parvus 

 and P. minor are intermediate. The molars of Agriochcerus are nearly square, while those of 

 Merycoidodon are longer than wide. The molars of Protoreodon, therefore, have to lose the proto- 

 conule and change their shape from wider than long to longer than wide before they attain the 

 condition of those in Merycoidodon, which may not have taken place. In other words, it is very pos- 

 sible that Protoreodon is not directly ancestral to Merycoidodon but is an aberrant branch from the 

 stem stock. 



