SEAS AND LAKES. 



the surface to the bottom. This -will continue so long as the 

 superficial stratum is rendered heavier, volume for volume, than 

 those below it, by being lowered in temperature. But the super- 

 ficial stratum, and all the inferior strata, will at length be reduced 

 to the uniform temperature of 38'8. After this the system of 

 currents upwards and downwards will cease. The several strata 

 will assume a state of repose. When the superficial stratum is 

 reduced to a temperature lower than 38 '8 (which is that of the 

 maximum density of water), it will become lighter, volume for 

 volume, instead of being heavier than the inferior strata. It will 

 therefore float upon them. The stratum immediately below it, 

 and in contact with it, will be reduced in temperature, but in a 

 less degree ; and in like manner a succession of strata, one below 

 the other, to a certain depth, will be lowered in temperature by 

 the cold of those above them, but each stratum being lighter than 

 those below, will remain at rest, and no interchange by currents 

 will take place between stratum and stratum. If water were a 

 good conductor of heat, the cooling effect of the surface would 

 extend downwards to a considerable depth. But water being, on 

 the contrary, an extremely imperfect conductor, the effect of the 

 superficial temperature will extend only to a very limited dep{h ; 

 and at and below that limit, the uniform temperature of 38 '8, 

 that of the greatest density, will be maintained. 



This state of repose will continue until the superficial stratum 

 falls to 32 *, after which it will be congealed. When its surface 

 is solidified, if it be still exposed to a cold lower than 32, the 

 temperature of the surface of the ice will continue to fall, and this 

 reduced temperature will be propagated downward, diminishing, 

 however, in degree, so as to reduce the temperature of the stratum 

 on which the ice rests to 32, and therefore to continue the process 

 of congelation, and to thicken the ice. 



If ice were a good conductor of heat, this downward process of 

 congelation would be continued indefinitely, and it would not be 

 impossible that the entire mass of water from the surface to the 

 bottom, whatever be the depth, might be solidified. Ice, however, 

 is nearly as bad a conductor of heat as water, so that the super- 

 ficial temperature can be propagated only to a very inconsiderable 

 depth ; and it is found accordingly, that the crust of ice formed 

 even on the surface of the polar seas, does not exceed the average 

 thickness of twenty feet. 



32. The thermal condition, therefore, of a frozen sea, is a state 

 of molecular repose, as absolute as if the whole mass of liquid were 

 solid. The temperature at the surface of the ice being below the 



* For sea water the freezing point is 284. 



75 



