S32 On the Practkalility 



perature of the atmosphere must be constantly below that, of 

 course the freezing above the surface of the sea will be as con- 

 stant, though the surface of the sea itself probably will not freeze 

 at a temperature much below 30^, even in a motionless state. 

 The same lodij of ice while freezing above water, that is in- 

 creasing in size and extent by snow and hail, and the salt water 

 freezing in washing over it, may perhaps at the same time be 

 melting under ivater ; and this process will be probably accele- 

 rated according to the magnitude of the mass, and the depth of 

 its immersion. For when the atmosphere is colder than the 

 surface of the sea, the water will (in proportion perhaps to its 

 depth) be found warmer bysome degrees, than it is at the surface; 

 and though few experiments have yet been made to establish 

 this fact, yet sufficient to warrant this conclusion. Thus, even 

 in sunnner, if the temperature of the atmosphere should bs 32% 

 and the surface of the sea, clear of land and soundings, three or 

 four degrees higher, that of the water below would probably be 

 much higher still ; so that the portion of a large mass of ice 

 above the surface of the sea would remain ice and augment ; and 

 the other portion of it helow being immersed in a temperature 

 exceeding the point of congelation, would probably be melting 

 and decreasing. 



The well attested facts, of large bodies of ice having been seen 

 to capsize or turn bottom up, prove, that their centre of gravity 

 is altered by an increase of their bulk above, or a diminution of 

 it below, according to the excess of either effect. Upon the 

 whole, however, it seems probable, that in the Arctic regions, 

 the process of freezing in the atmosphere, exceeds that of melt- 

 ing under water, particularly on those smaller masses of ice which 

 are immersed least; and therefore that there must be a general 

 increase of ice in the polar basin, from the Pole {if the ice ori- 

 ginates there) towards the lands surrounding the basin ; or from 

 those lands (if the ice first forms there) »ip towards the Pole, 

 On this question, too, opinions have been most various. Every 

 circumstance seems to weigh against the opinion of its greatest 

 formation being about the Pole, except one, and that is, because 

 the sea-water there will most probably be found to contain the 

 least salt. I am disposed to believe, that it must also in the 

 winter be colder at the surface of the sea, near the Pole, than any 

 where else. Yet, on the whole, it seems most probable, that the 

 ice is originally formed in the rivers, and along the shores of all 

 the lai'ids surrounding the Polar basin. Being afterwards de- 

 tached from those lands, and driven to sea bv the winds and cur- 

 rents in masses of more or less extent of surface, but no great 

 thickness ; it there accumulates by the falling of snow and hail, 



w4 



