TERTIARY LACUSTRI^'E FORMATIONS OF THE WEST. 595 



corresponds very closely with that of the Palo Duro horizon, and represents a forma- 

 tion nearly or quite contemporaneous with it. These Archer fossils have some- 

 times been regarded as the equivalents of those in the Peace Creek beds, but they 

 are much older and represent a very different fauna. 



The two older Loup Fork horizons, the Deep River and Nebraska, belong un- 

 questionably to the Miocene, the reference of them to the Pliocene, which has been 

 made by some authorities, is due partly to incorrect identifications of fragmentary 

 fossils and partly to a mingling of fossils from different horizons, such as very 

 frequently occurs in localities where the Kquus beds overlie the Loup Fork. For 

 stratigraphic reasons Dall regards the Archer deposits as Pliocene : and if this con- 

 tention proves to be well founded the line between Miocene and Pliocene should 

 be drawn between the Nebraska and Palo Dnro horizons of the Loup Fork, The 

 following table will display these relations : 



Pleistocene Eqiais beds. 



{Peace creek. 

 Blanco. 

 C Palo Duro and Archer = Hippidium beds. 



Upper Miocene Loup Fork. ■ • ■ \ Nebraska = Cosonjx beds. 



( Deep river = Cyclopidius beds. 



A much more difficult task is to correlate these various horizons with their 

 equivalents in Euroi)e. ]\Iany of our leading geologists deprecate all such attempts 

 in the present state of knowledge, and from the stratigraphic point of view this is 

 very natural. From the standpoint, however, of morphologic paleontology, which 

 deals with questions of origins and migrations, correlations are simply indispensable 

 and must l^e attempted again and again until a satisfactory result has been reached. 

 A false correlation will sooner or later refute itself by the . contradictions and 

 absurdities to which it leads, as is abundantly shcnvn by the inferences which 

 follow from regarding the White River deposits as Miocene rather than 01igO(;ene. 

 For the purpose of correlating the lacustrine deposits of Europe and North America, 

 the principal stress must be laid U])()n the mammals, and from the comparative 

 rapidity of change which this group displays notiiing could be better adapted to 

 the purpose, provided the proper limitations of this method be observed. The 

 comparison of mammalian faunas in widel v separated and isolated regions can lead 

 to no satisfactory result, and it is this which renders the determination of the South 

 American fresh- water Tertiaries so very difficult. Throughout Tertiary time South 

 America was isolated from the northern continents and developed a fauna entirely 

 peculiar to itself. The effects of this long isolation are strongly marked, even at 

 the present time. On the other hand, North America was repeatedly and for long 

 periods connected with the Old World, and frequent migrations of mammals 

 occurred in both directions. During periods when this connection appears to have 

 been interrui)ted very peculiar faunas develc^ped on this continent in isolation, a 

 striking example of which is the Bridger Eocene, but then subsequent removal of 

 barriers reestablished similarity. 



Time forbids entering into these correlations in detail. LTere it will suffice to 

 indicate three points which appear to be well fi.x;ed as equivalents on the two sides 

 of the ocean : 



1. The Wasatch Eocene represents the Suessonian of France (Cope). 



2. The White River beds are the equivalent of the Oligocene of Ronzon. 



3. The Deep Riyer horizon of the Loup Fork corresponds to the Upper Miocene 

 of Sansan. 



