those of the underlying beds except that some of them have well-developed lachrymal vacuities while others have none. Another new race also makes its appearance suddenly, and in great abundance, in the genus Promerycochærus—structurally derivable perhaps from some of the older oreodons, but not connected with them by intergradations. Agriochærus has disappeared. In the Upper Rosebud the Oreodon-Merychyus phylum shows a distinct and marked advance in the length of the crowns of the teeth; lachrymal vacuities are always present, the feet are decidedly more compact and elongate. Promerycochærus disappears entirely and is replaced by a very distinct and more advanced genus Merycochærus. The Leptauchenia series has disappeared temporarily, to re-appear in the Middle Miocene in a more specialized genus, Cyclopidius, the last known member of this race.
The Middle Miocene (which should follow the Upper Eosebud) is unrepresented at the locality under consideration (Pine Eidge, South Dakota), but elsewhere overlies beds with an equivalent fauna, and contains Merycochærus in one locality with Merychyus (both represented by more specialized species); in another locality it contains instead, Promerycochærus with Ticholeptus (allied to Merychyus); in a third is found the most highly specialized member of the Merycochærus line, Pronomotherium. In the Upper Miocene and Lower Pliocene the oreodonts become much scarcer, and the skulls and skeletons are known only in two or three instances. Pronomotherium certainly occurs in Montana; in Nebraska the Merychyi are more advanced in dentition, belonging to a distinct subgenus Metoreodon; but whether the skulls and skeletons are equally different we do not yet know, nor are we in a position to say whether the change is gradual or saltatory.
But the sum of results in regard to the changes from one stage to another in this best known group of fossil mammals is either that the changes are abrupt, constituting clean-cut faunal divisions marked by the sudden appearance in abundance of a more advanced stage; or else that the new form replaces the older one little by little, but on the whole can not be fairly said to be gradually converted into it by infinitesimal gradations.
This general observation applies, in my opinion, equally well to any abundant group of fossil vertebrates whose phylogeny is sufficiently known to make them worth considering.
If, therefore, we consider that the record is continuous where there is no apparent stratigraphic break, and that the known record really represents what was going on over the entire continent of North America, I do not see that we can fairly escape from the conclusion that new species, new genera and even larger groups have appeared by saltatory evolution, not by continuous development.
But—and here lies the crux of the whole question—we have no
