210 ACCOUNT OF THE ARCTIC REGIONS. 



That fresh water obtains its greatest density at a 

 temperature a few degrees above the freezing point, 

 and that it expands on any farther reduction of the 

 temperature, are facts well established ; but that 

 sea- water follows a different law, and continues to 

 contract down to the point of freezing, is a question 

 which has not been sastisfactorily decided. Count 

 Rumford, in his " Essays," (vol. ii. p. 302.) says, in- 

 deed, that "sea- water continues to be condensed, as it 

 goes on to cool, even after it has passed the point at 

 which fresh water freezes;" but from the circumstance 

 of an under stratum of water in the Spitzbergen Sea, 

 being generally warmer by some degrees than that at 

 the surface, though of similar specific gravity, it 

 would appear that the warmer water is in this case 

 the most dense, or why does it not rise and change 

 places with the colder water at the surface ? Hence 

 I think there is reason to believe, that sea-water . 

 follows the same law as fresh water, with regard to 

 the extreme of density being a few degrees above 

 the freezing temperature, and that the under-stream 

 of comparatively warm water, observed in the Spitz- 

 bergen Sea, which is of a temperature 16° to 20° 

 above the mean temperature of the climate, is an 

 under-current derived from a southern region. 



In some situations near Spitzbergen, the warm 

 water not only occupies the lower and mid regions 

 of the sea, but also appears at the surface. From 

 inspection of the preceding Table of the Specific 



