[ m 1 



and nineteen, funn fix until twelve at night. The quantity of 

 ice, formed on the preceding day, was not notably augmented 

 or diminifhed ; b/uiled, waftied, and melted, it yielded two ounces 

 of water, brackifti to thetafte, in a greater degree than any of the 

 foregoing portions which were warned. 



" On the 30th of January, hnding that the, temperature of 

 the preceding evening of the night, and of this day, which was 

 between nineteen and twenty-one, had caufed no notable con- 

 gelation in the fmall quantity of water remaining in B; finding 

 alfo that the relidue of the water in A admitted of no further 

 congelation worth notice ; and confidering that the {lender laminae 

 of ice, lately formed in thefe waters, melted to falt-water, and 

 consequently that no further congelation, capable of Separating 

 the frefti water from the brine, even with the affiftance of warn- 

 ing;, could take place; I mixed the concentrated brine in A with 

 that in B, and found both fcarcely meafured a ,wine pint ; fome 

 fmall cryftals were found in the bottom of both veflels, which 

 funk in the brine^ and were to the tafte fea fait. It is hence 

 evident that fome fea fait is formed "in cryftals by the con- 

 centration produced by cold acting gradually, and caufing con- 

 gelation only on the mrface of the water, or not affecting that 

 part of it which is contiguous to the bottom of the veffel. 



" The quantity of thefe cryftals of fea fait was about two 

 grains. I poured them together with the water into a china 

 plate, fet in a fand heat, and, by cryftalization, obtained fea fait 

 and the other faiine contents of fea-water, in a dry form, near 

 two ounces, averdupoife. 



" Now, as this quantity of fea-water (that is, two gallons), 

 taken on our coaft, generally yields about feven ounces of faiine 

 matters; it appears, that two- thirds or more of the fea fait, and 

 bitter falts of fea-water, are intercepted in the ice of the fuccemve 



congelations, 



1 



