488 Royal Society : — 



congelation, its density being less, the lower the temperature at 

 which the process of congelation took place. Although the author's 

 observations extended from +28°*5 to —42°, he was never able to 

 obtain fresh water from sea-ice, the purest specimen being of specific 

 gravity 1*005, and affording abundant evidence of the presence of 

 salts, especially of chloride of sodium, in such quantity as to render 

 it unfit for domestic purposes. 



The efflorescence alreadj' referred to appeared sooner or later, 

 according to the temperature of the air, but generally commenced 

 when the ice was ~ of an inch thick, and continued to form till 

 the ice attained a thickness of about 9 inches, when, in consequence 

 of the compactness of the frozen mass, it ceased to appear at the 

 surface. The lower the temperature at which the ice was formed, 

 the more abundant was the efflorescence. Direct experiments made 

 by freezing sea-water in a large tub, showed that the unfrozen re- 

 siduum contained a considerable portion of salts expressed from the 

 ice. The author therefore infers, that after the efflorescence had 

 ceased to form on the surface, the saline particles were precipitated 

 into the unfrozen liquid below. On exposing the residual ; liquid 

 from which the ice had been separated to a freezing temperature, a 

 second residuum was obtained, containing more salts than the first ; 

 and by repeating the process several times, there remained finally a 

 strong solution of brine. 



The author endeavoured, by reversing this process, to procure 

 fresh water. He remelted the ice from sea-water and froze it again, 

 repeating the operation several times. Ice was thus obtained, 

 which, when melted, gave water, having a density of from r0025 

 to 1-0020. 



A "heavy nip" having occurred in the floe near the ship afforded 

 an opportunity of examining the quality of the ice at different 

 depths. The thickness of the entire mass was 54 inches ; the den- 

 sity of the solution obtained by melting successive portions varied 

 from 1*0078 to 1*0050; those near the surface giving a liquid of 

 higher density than the rest. A specimen taken from the centre of 

 the mass was reserved for analysis. 



With regard to the "efflorescence," the author states that its 

 appearance was very different according as the temperature was 

 above or below — 25°. In the former case, it exhibited a plumose 

 form, with secondary plumes branching off; in the latter, it con- 

 sisted of fibrous crystals varying from ^ to 2 inches in length. This 

 efflorescence acts an important part in the breaking up of the floe. 

 From the middle of January cracks and lanes occur in the floe, 

 which subsequently become filled with new ice covered as usual with 

 the saline efflorescence and a little snow. When the sun's rays fall 

 upon this incrustation, it melts and forms a thick liquid on the top. 

 This penetrates gradually through the ice and aids greatly in break- 

 ing it up. The author supposes that a process of endosmosis and 

 exosmosis is, in fact, established through the body of the ice. A 

 similar, but less powerful, action is produced by the same cause on 

 the mass of the floe itself. 



I 



