SCIENTIFIC ItESULTS 



9 



whirli are caught a smaller luimber of salt crystals, to«;etlicr with an 

 aj)preciable quantity of concentrated brine. The relative amount 

 of the latter is dependent upon the temperature; the colder, the more 

 brine is imprisoned in the cellular structure. 



It is an interestino- question whether all the salts contained in 

 >cii water are to be found in sea ice; also if the salts in the ice are 

 combined in the same relative proportions as they are in the sea 

 water from which the ice was frozen. Laboratory experiments to 

 tost the solubility of its various salts recpire a supercooling; and 

 aptation of the liquid, seldom, if ever, experienced under natural 

 conditions, but such investigations furnish interesting information 

 reo-arding the eutectic point of the salts and also regarding the salin- 

 ity of the resulting solid. According to Johnstone (1923, p. 186) 

 every sea-water salt has been found to possess its own individual 

 temperature of solubility which if depressed, causes that particular 

 salt to precipitate from the mother liquid. First to crystallize, as 

 i)reviously stated, are the water molecules wdiich for a sample of 33 

 per mille salinity takes place at -1.8° C. (28.8° F.). When 

 -8.2° C. (17.3° F.) is reached NagSO^ begins to precipitate, and at 

 -23° C. ( — 9.4° F.) solid XaCl separates from the remaining brine. 



The behavior of sea water when supercooled has led to the estab- 

 lishment of a theory that yearly great quantities of SO., are removed 

 from polar regions and deposited in the Atlantic as ice floes saturated 

 with that substance move out of the north to melt in lower latitudes. 

 Malmgren (1928, p. 7) had an excellent opportunity to test this 

 theory by determining the ratio of SO3 and CI, for samples of sea 

 water and of sea ice from several of the freezing zones of the polar 

 sea. If a selective process of salt crystallization prevails, SO3 and 

 CI would be the substances first to exhibit variations. In all cases, 

 however, the sample pairs of ice and mother liquid showed almost 

 the same ratio of SO3 and CI. 



Wiese (1930) reports Liakionoff's investigations of this phe- 

 nomenon in the Barents Sea the summer of 1930. Liakionoff found 

 contrary to Malmgren that a deficit of CI and a surplus of SO3 

 actually prevailed both in the melting water and in the sea ice 

 itself. The ratio ranged from Krummel's value of 0.1150 for ordi- 

 nary sea water to a maximum of 0.1700 in old sea ice. These results 

 agree moreover w^ith those of Ringer and Pettersson, and therefore 

 it appears that a selection of the salts in sea water occurs upon 

 freezing. Future investigations will determine this with more cer- 

 tainty. The freezing point of sea water is defined as that temperature 

 at which the first ice crystals appear in the liquid. 



The freezing points of samples of water of varying salinity are 

 given herewith : 



