xx Contents 



PAGE 



north-westerly direction, and away from the western 

 coast of the continent which is situated to windward of 

 it. 



The translation of water which takes place under the 

 influence of the trade-wind is the mechanical equivalent 

 of the translation of air which has been extinguished. 

 The mass of water moved is very great, but its superficial 

 extension is also very great, consequently the local 

 velocity produced in any part of it is small, but it is 

 continuous. 



It has been shown that the water on these coasts 

 has a temperature lower than that corresponding to 

 the local climate, and it was this fact which arrested 

 Humboldt's attention in his travels in Peru; therefore 

 it must be drawn from a source which furnishes water 

 colder than that corresponding to the local climate. 

 Of these there are two, namely, the surface water in 

 high latitudes situated at a great distance, and the sub- 

 surface or abyssmal water contiguous with the coast 

 and with the water actually being removed. 



The steamer in which I travelled from Valparaiso as 

 far as Panama, anchored at sixteen different places on 

 the South American Coast between Valparaiso and 

 Cape Blanco and communicated by means of boats with 

 the shore. She always anchored' in the cold fringe of 

 water. With the ship thus lying at anchor for an hour 

 or two at each place, I could not detect any current 

 whatever. 



If the water were supplied from the immediately 

 adjacent and enormous store of abyssmal water of low 

 temperature it could be furnished by an upward 

 current of practically insensible velocity . . .107 



As a matter of mere probability the likelihood of 

 the cold water having this abyssmal source is so great 

 as to amount to a certainty, and all the experimental 

 evidence confirms it. 



The cold inshore water found on the Atlantic Coast 

 of North America is a similar phenomenon in a homo- 

 logous region. Like that on the Pacific coast of South 

 America, it was attributed to a surface current from 

 Arctic seas and was called the Labrador Current . . 108 



Had there been an in-shore surface current along 

 the coast from Newfoundland strong enough to place 



