I 95 3 



the firft experiment which I made with regard to freezing fea- 

 water, I deduced the fame inference that he hath done, having 

 warned it in frefh water for the fame reafon that he did, viz. 

 tQ get rid of the brine which might adhere to the furface of 

 tiie ice. 



"To determine, therefore, whence this frefhnefs in the thawed 

 ice might arife, I placed a large piece of what remained frozen 

 {without being warned at all in pump-water) to be diflblved be- 

 fore the fire, which tailed very fait as one might naturally fuppofe. 



The weather continuing to be very fevere, I froze more .fea- 

 water, repeating the experiment of frelhening it or not, by 

 leaving, or not leaving it, in pump-water, which always 

 turned out uniformly to be the fame ; and the reafon of which 

 is the following. 



When fea-water is frozen, it does not form ice fimilar to that 

 from frefh water, being by no means fo foiid or tranfparent, 

 as it confifts of thin laminae or plates, between which the brine 

 is depofited, and if the ice is accurately examined, the fmall 



the ice which floats in the Northern Seas is formed from the falt-water 

 or not, he therefore fhould have thawed the ice precifely under the fame 

 circumftances with the fea-water adhering, as -the navigators take it up. 

 The truth is, that, if the piece of ice tormed from fea-water is at all 

 large, the adhering falt-water can fcarcely affedf. the tafte at all; and I 

 have melted the central parts of a pretty large mafs, which became very 

 fait after diffolution, though entirely detached from the fea-water in 

 which it had been frozen. " In the fevere froft laft January (viz. 1775), 

 *' fome falt-water, being fet abroad, froze into an ice, which was not 

 " folid but porous, the hollows being filled with the falteft part of the 

 <e water, for the ice when drained was quite frefh. The falt-water being 

 * again fet abroad, froze as before, what remained ftill unfrozen was 

 " now become exceeding fait, but t'he ice drained and diffolved was 

 " little if at all brackifh; by this experiment, if another time more fully 

 iC repeated, it may be found to what degree the faltnefs of water may 

 tc be increafed, by continuing to freeze away the frefh-water." Mr, 

 Barker in Phil. Tranf. Vol. LXVI. p. ii. 1776. p. 373. 



portions 



