in Fluids. 327 



vantage which sea water possesses over fresh water, con- 

 sidered as an equalizer of the temperature of the atmos- 

 phere, which arises from the comparative lowness of the 

 point of its congelation ; supposing even sea water to 

 freeze at as high a temperature as fresh water, namely, 

 at 32 ; and supposing (what is strictly true) that as soon 

 as either sea water or fresh water is frozen at its surface, 

 and this ice covered with snow, the communication of 

 Heat from the water to the atmosphere ceases almost 

 entirely, we will endeavour to determine how much 

 more Heat would, even on this supposition, be commu- 

 nicated to the air by salt water than by fresh water, after 

 both have arrived at the temperature of 40. 



When fresh water, in cooling, has arrived at this tem- 

 perature, it ceases to be farther condensed with cold, and 

 its internal motions (which, as we have already more than 

 once observed, are caused solely by the changes produced 

 in the specific gravity of its particles) cease, of course, 

 and ice immediately begins to be formed on its surface; 

 but as the condensation of salt water goes on as its Heat 

 goes on to be diminished, its internal motions will con- 

 tinue ; and it is evidently impossible for ice to be formed 

 at its surface till the whole mass of the water has become 

 ice-cold, or till its temperature is brought down to 32. 

 It would therefore give off a quantity of Heat equal to 

 8 degrees, at least, of Fahrenheit's thermometer, more 

 than the fresh water would part with before ice could be 

 formed on its surface. 



To be able to form an idea of this enormous quantity 

 of Heat, we have only to recollect what has already been 

 said, and we shall find reason to conclude that it would be 

 sufficient to melt a covering of ice equal in thickness to 

 S 2 g of the depth of the sea. It would therefore be suffi- 



