310 SYNOPSIS OF THE VERTEBRATE FAUNA OF THE PUERCO SERIES. 



braces Deltatheiium, Mioclfenns, and probably Triisodon and Onychodectes, of the 

 Puerco genera. Dissacus is on the bonndaries of the Mesonychidjie, but the nngues 

 are unknown, and the astragalus is not trochlear. That Hemiganus can be referred 

 to the Proviverridse is improbable. The characters of a family Hemiganida? would be : 

 Sujjerior molars with confiuent roots, and trituhercular crowns; inferior molars quadri- 

 tubercular ; astragalus not trocJilear ; ungues much compressed. • The sole character 

 separating this flimily from the Proviverridjie is the confluence of the roots of the 

 superior true molars. It is approached in this respect by the genus Conoryctes, 

 which will, however, remain in the Pl•oviverrid^e as an aberrant form, connected with 

 Miocla^nus by Onychodectes. The tendency of this group is to resemble the Tseni- 

 odonta, but whether there is any affinity involved in this resemblance does not yet 

 appear. 



The line of the Mesonychidaj shows definite changes in the following respects : 

 First, loss of the internal cusps of the inferior molars. The diminution of these 

 cusps is seen most prominently at first in the genus Triisodon, and more distinctly in 

 the later T. quivirensis than in the earlier T. hiculminatus. It has progressed far in 

 Dissacus. Second, in the reduction of the number of the digits ; the hallux has 

 become very small in Dissacus and is gone in Mesonyx. Third, in the production of 

 a distinct facet for the cuboid bone on the extremity of the astragalus. The contact 

 of the astragalus and cuboid is seen in many if not all Creodonta, but the navic- 

 ular and cuboid surfaces are not distinguished from each other. In Mesonyx and 

 Pachysena it is well known that they form distinct facets separated by an angle, as in 

 Perissodactyla. This structure appears in a primitive condition in Dissacus, the 

 angle being obtuse and rounded. It is an excellent illusti'ation of the origin of a 

 zoulogical character. The thoroughly diplarthrous type of the astragalus in Mesonyx 

 is associated with an ungulate form of ungual phalanges, and I suspect that the 

 resemblance is not accidental. The form of the distal end of the diplarthrous astra- 

 galus is probably due to impacts combined with energetic flexion and extension. 

 But why the structure should appear in Mesonyx and not in Phenacodus is, as yet, 

 an unanswered question. 



Although it is now evident that the Pinniped Carnivora cannot be derived from 

 the genus Mesonyx, as is truly maintained by Scott, it appears to me that they have 

 been derived, like other Carnivora, from some form of Creodonta. The retrograde 

 development of the molar teeth probably passed through stages like those of Dissacus 

 and Hemiganus, and the ungual phalanges might have been easily derived from such 

 as are possessed by Piichyjena. 



