236 ACCOUNT OF THE ARCTIC REGIONS. 



10°, and the ebullition continued brisk whenever 

 the phial was plunged into cold water or snow, 

 for a period of 15 or 20 minutes. This showed 

 that there could be little or no air in the phial. No 

 crystals of ice were observed, until an hour and a 

 half after the ebullition ceased, and then the process 

 of freezing went on briskly. In two or three hours 

 afterwards, the whole of the water was consolidated. 

 Air-bubbles were observed moving towards the sur- 

 face, as the process advanced, and, when completed, 

 the ice had a milky appearance throughout, and 

 was found crowded with minute globules of air. 

 Hence it is probable, either that water is not en- 

 tirely freed from air by boiling, or that some of 

 the water is decomposed during the progress of the 

 freezing process. 



In consequence of this disengagement of air, ice 

 formed in small vessels, or in confined situations, 

 cannot be altogether transparent ; for whenever a 

 pellicle of ice covers the surface, the air as it is 

 dislodged, is prevented from escaping, but rising 

 as high as the ice will permit, there gets inclosed 

 by the formation of new crystals, and renders the 

 ice obscure. But where, from accidental circum- 

 stances, the air, as it is disengaged, can make its 

 escape, the ice which is formed may present consi- 

 derable masses totally free from any visible pore. 

 Thus, when water contained in a large cylindrical 

 vessel, is exposed to a low temperature, the first 



