( >°5 ] 



In which It was placed, the ice was not folid like that of frefh 

 water, but, on the contrary, could eafily be cut through the 

 centre of the mafs with a knife. The ice tatted equally of 

 fait through the whole mafs, in the fame manner as a like 

 quantity of fea-water. Bruifed brifkly, warned as already de- 

 fcribed, and melted, it yielded frefh water to the quantity of 

 four-fifths of the water frozen; wherefore in warning very little 

 ice was diffolved whilft the falt-water intercepted in the ice 

 was removed. 



" Mr. Barrington having obferved that an artificial freezing 

 commences from the bottom and fides of the mafs of water 

 placed as ufual in the frigorific mixture, but that natural free- 

 zing commences on the furface and proceeds downwards; and 

 it occuring to me that the fpecific gravity of incongelable brine 

 is greater than that of the congelable water ; and, confequently, 

 that this greater fpecific gravity favours the feparation of brine from 

 the ice of fea water, when the freezing commences on the furface 

 of fea-water, and may be an impediment to the feparation of the 

 incongelable brine from the ice artificially formed in the fea- 

 water, when the congelation proceeds from the bottom up- 

 wards : On thefe confederations it feemed that the foregoing 

 experiments indicate, that ice formed in fea-water cannot, when 

 melted, become frefh water, unlefs it be warned in frefh water ; 

 but do not fully prove, that ice formed on the furface only, and 

 proceeding flovvly downwards, in fea-water, may not confift of 

 frefh water, and be freed from brine, by reafon of the fpecific 

 gravity of brine and other unnoticed circumftances. There- 

 fore, on the 2 1 ft of January, at two o'clock, when the mercury 

 flood in the open air at twenty-nine, I made the following ex- 

 periment, with a view to determine whether fea-water, frozen 

 artificially from the furface downwards in the manner performed 



P by 



