HOLMES DISTRIBUTION OF THE HUMAN RACE. 25 



southeastwardly to New Zealand, or to Australia, in Tertiary 

 times, may give some countenance to the idea that some land, or 

 islands, within the human period, approached nearer to South 

 America than they now do ; and the existence of sunken stone 

 buildings on the southeastern coasts of Asia and on some Pacific 

 islands, and some resemblance of construction and style of art be- 

 tween the most ancient stone structures of Java, Ceylon and Cen- 

 tral America, have been noticed in confirmation of it. But when 

 it is considered that the recent soundings show a general depth 

 of 2,000 to 3,000 fathoms over the whole ocean between them, 

 the hypothesis becomes extremely improbable, though an eleva- 

 tion of somewhere near that amount, within the Tertiary period, 

 might not greatly surprise a geologist. But an emergence of the 

 continental shores of Asia sufliicient to join the continents across 

 Behring's Straits, inclusive of Japan and the Aleutian Islands, is 

 not only geologically probable, but a due consideration of the dis- 

 tribution of both animals and plants demonstrates that such must 

 have been the fact in Miocene times and even as late as the Plio- 

 cene, and while warm temperatures continued in that latitude. 

 A comparison of the fossil plants of the island of Sachalin and of 

 Alaska, belonging to the Miocene and Pliocene, confirms this 

 view.* And here we may find an easy passage by continuous 

 land for peoples of the brown band of races from southeastern 

 Asia into North America as early as the Pliocene period. 



The case is somewhat similar on the Atlantic side. Wide and 

 deep seas must have existed between the two continents from the 

 earliest geological periods. The recent soundings, however, show 

 that a continuous land connection between northern Europe and 

 northeastern America may possibly have existed in the Miocene, 

 or even as late as the glacial epoch ; and other evidences confirm 

 this hypothesis. The myth of a sunken Atlantis is too recent to 

 deserve attention ; but the other continent may have extended to 

 the Madeiras, the Canaries, and the Cape Verde islands as lately 

 as the glacial epoch. It appears by the soundings of the " Chal- 

 lenger" {Map in the " Nature" No. 383) that an elevation of 

 1500 fathoms would make the Madeiras the point of a peninsular 

 extension of Portugal and the Canaries a headland of Africa, with 



* Miocene Flora der Inseln Sachalin, by Prof. O. Heer, p. 6 (1S78) ; Memoires de I'Acad. 

 Imp, des Sciences de St. Petersbourg, T. txv. No. 7. 



