314 



PETTETiSSON, OK WATEIi AND ICE. 



It strikes iis at once, that the latent heat, developed by 

 the freezing of sea-water, is extraordlnarily inferior to that of 

 jmre ivater. We further observe, that the law of Person has 

 no application to sea-water. Its latent heat seems to be greater 

 at lower temperatures than nearer to its melting-point. Un- 

 happily the ratio of this alteration of the latent heat cannot 

 be determined with the same exactness as in pure water, on 

 account of the slowness of the freezing process in salt water. 

 Still I think much can be done by further experiments in this 

 way, for the present it will snffice to call attention to the fact, 

 that water of the sample Y at — 7°. 78 showed a latent heat 

 -of 72.5 cal. and at — 6°.49 only 70.o cal. etc. 



From these observations the following conclusions are 

 inevitable: 



I. Any theory, which considers the salt of the sea-ice to 

 be mechanically included as crystallized salt, must be false. 

 For in this case we ought to have found for the developed 

 lieat the values of column 7 instead of those in column 6. 



II. Any theory, which ascribes the saltness to adhering 

 dröps of liquid, concentrated brine, will prove insufficient. 



Sea-water seems to be divided by freezing into three parts : 

 ice, solid cryohydrates and liquid brine containing dissolved 

 salts. The relative . proportion of these components depends 

 upon the freezing temperature. B}^ the freezing of sea-water 

 in nature, the salt brine for the most part separates from the 

 ice and mingles with the adjacent unfrozen water. This was 

 not the case in my experiments. The brine remained in the 



1 The slight amount of heat, which ought to be developed by the sup- 

 posed crystallisation of the salt, is not taken into account here. 



