1902.] AND ANCIENT GEOGRAPHY. 399 



Mexican, Orinoco and Amazonas interoceanic connections existed. 

 But this would not influence our general result that the communica- 

 tion of the oceans was interrupted definitively in the Miocene. 



We have, beginning in Mesozoic times, a differentiation of two 

 types of marine faunas, a Mediterranean and a Pacific, but these 

 faunas communicated with each other at certain points and were 

 completely separated for a comparatively short period in the 

 Upper Cretaceous by the continent of Mesozonia (if this separation 

 was at all complete at any time). 1 Generally the Mediterranean 

 fauna belongs originally to the northern hemisphere, the Pacific to 

 the southern, except that the latter largely encroached upon the 

 former in the region of the Northern Pacific. This arrangement 

 was completely upset during the Tertiary, so that at present the 

 Atlantic fauna (containing chiefly the descendants of the old Medi- 

 terranean types) and the Pacific fauna are divided, not by a line 

 running east and west, but by two lines running generally north 

 and south (in America and in the Old World). Besides, the Arctic 

 and Antarctic types have been added, the former being an offshoot 

 of the Mediterranean, the latter of the Pacific type. 2 



While in former times, in the Mesozoic and Lower Tertiary, a 

 decided tendency prevailed to mix the marine faunas of the world 

 and make them more or less uniform, which tendency was checked 

 only temporarily, we have, from the Miocene upward, a complete 

 separation of two marine types of fauna, 3 which, however, still 

 possess certain features in common that are due to conditions pre- 

 vailing in earlier times, and with respect to Central America these 

 conditions (interoceanic connections) were present for the last 

 time in the Older Tertiary (Eocene and Oligocene). 



1 We possess evidence that Mesozonia was interrupted for shorter periods within 

 the Upper Cretaceous, for instance, in the region of British Colombia (see Koss- 

 mat) and in Western Venezuela (see above, p. 343). 



2 In opposition to the belief of some authors (Pfeffer, Murray) that both Polar 

 faunas are strikingly similar, I have always held the opinion (see review of the 

 literature in Americ. Natural., Vol. 35, 1901, p. 139 ff.) that this is not so. We 

 see here also that the origin of these faunas is different, the one being derived 

 from the old Mediterranean, the other from the old Pacific fauna, the differences 

 of which, although obscured during the earth's history by frequent interchanges, 

 go back to Mesozoic times. 



3 The complete separation was brought about not only by topographical 

 factors, but chiefly by the additional action of climatic differentiation. See Ort- 

 mann, Grundzuege der marin, Thiergeograph., 1896, p. 40. 



