368 CORALS AND CORAL ISLANDS. . 



thousands of feet, even thirty-five thousand by the close of the 

 Paleozoic along the Appalachians, and much beyond this on 

 the Pacific border; and when these thick sediments had in 

 many regions been stiffened by crystallization or metamor- 

 phism ; I say it is reasonable that, finally, changes of level, 

 through the working again of the old system of forces, should 

 affain have effected most the old nucleal Azoic area of the conti- 

 nent, where there had been no thickening except what had take.i 

 place internally ; and that, if one arm of the Y, that along the 

 Canadian watershed, were raised at this time — as the facts 

 prove — the other, northwestern in trend, should also have been 

 raised, and to a greater extent. This is at least probable enough 

 to become a question for special examination over the region. 



These northern continental upward movements which in- 

 troduced the Glacial era, carrying the Arctic far toward the 

 tropics, may have been a balance to the downward oceanic 

 movements that resulted in the formation of the Pacific atolls. 

 While the crust was arching upward over the former (not ris- 

 ing into mountains, but simply arching upward) it may have 

 been bending downward over the vast central area of the great 

 ocean. 



The changes which took place, cotemporaneously, in 

 the Atlantic tropics, are very imperfectly recorded. The 

 Bahamas show by their form and position that they cover 

 a submerged land of large area stretching over six hun- 

 dred miles from northwest to southeast. The long line 

 of reefs and the Florida Keys, trending far away from 

 the land of southern Florida, are evidence that this Flor- 

 ida region participated in the downward movement, though 

 to a less extent than the Bahamas. Again, the islands 

 of the West Indies diminish in size to the eastward, being 

 quite small in the long line that look out upon the blank 



