TERTIARY MAMMAL HORIZONS. 55 



above alluded to, apparently introduced stem forms of Eden- 

 tates into the Ethiopian region from which were derived the 

 pangolins and aard varks ; these peculiar edentates together 

 with armadillos all occur in southern France in the lower Oligo- 

 cene (Filhol, '94); this land bridge also distributed the Cape 

 golden moles, CJirysocJiloridce ; these facts and others too nu- 

 merous to mention serve to show the vast importance of the 

 explorations in Patagonia and make us impatient for the exact 

 conclusions which are forthcoming from the materials brought 

 together by Ameghino and Hatcher. 



The third migration into Neogaea established its links with 

 Australia, bringing in Marsupials, both polyprotodont and dipro- 

 todont. The fourth was from the north, Arctoga^a, and is 

 positively known ; it occurred at the end of the Miocene, and 

 brought in the northern Carnivora, bears, wolves, cats, and 

 sabre-tooth tigers, raccoons and mustelines, the Artiodactyla, 

 deer and camels, the Perissodactyla, horses and tapirs, three 

 types of rodents, the squirrels, mice and hares or rabbits and the 

 mastodon. The Notogaeic types, as well as the animals of 

 the first invasion, in the meantime had largely died out, and the 

 introduction of more vigorous Arctogaeic types, especially the 

 carnivores, together with a change of climate, exterminated a 

 further portion of the autochthonous Neogaeic fauna. At the 

 same time, that is of this second invasion, many of the South 

 American forms entered North America ; they seemed to have 

 reached this continent in the upper Pliocene. 



We now turn to Arctog.ea. In the Eocene period we find 

 in Europe and North America what may be considered the pure 

 or autochthonous fauna of the Holarctic region, in the absence 

 of all knowledge of Asia. Southern Asia is an absolute terra 

 incognita the earliest known deposits in this region being in the 

 Upper Oligocene in which the fauna is remarkably similar to 

 that of Europe. Northern Asia is unknown palaeontologically 

 until the Pleistocene — here is a region for explorers. However, 

 we may consider it as part of a broad Eurasiatic land area — ex- 

 tending from the Rocky Mountain region to Great Britain. 

 The faunal relations are astonishingly close, between the new and 



