chap, vir.] DEEP-SEA TEMPERATURES. 307 



and northwards of the western entrance of the channel. 

 We found that in these two areas, freely communi- 

 cating with one another and in immediate proximity, 

 two totally different conditions of climate existed at 

 all depths helow the immediate surface, where they 

 differed but slightly. In the Fseroe channel, at a 

 depth of 500 fathoms, the bottom temperature aver- 

 aged -l°-0 C, while at a like depth in the Atlantic 

 the minimum index stood at +6° C, a difference of 

 7 degrees Centigrade, nearly 13 degrees Fahrenheit. 



The conclusion at which we speedily arrived as 

 the only feasible explanation of these phenomena 

 was that an arctic stream of frigid water crept from 

 the north-eastward into the Fseroe channel lying in 

 the deeper part of the trough, owing to its higher 

 specific gravity ; while a body of water warmed even 

 above the normal temperature of the latitude, and 

 therefore coming from some southern source, was 

 passing northwards across its western entrance and 

 occupying the Avhole depth of that comparatively 

 shallow portion of the Atlantic from the surface to 

 the bottom. 



Several important facts of very general applica- 

 tion in Physical Geography had been placed beyond 

 doubt by these observations. It had been shown 

 that in nature, as in the experiments of M. Despretz, 

 sea-water does not share in the peculiarities of fresh 

 water, which, as has been long known, attains its 

 maximum density at 4° C. ; but, like most other 

 liquids, increases in density to its freezing-point : and 

 it had also been shown that, owing to the movement 

 of great bodies of water at different temperatures 

 in different directions, we may have in close proxi- 



x 2 



