366 



On the Plasticity of Glacier and other Tee. [June 21, 



Thus the yielding of one piece was well within the errors of obser- 

 vation, of the other two only just perceptible with the instrument 

 employed, and this small yielding may well have taken place entirely 

 in the thin layer of irregular ice with which the paper was attached . 



In the early part of the winter we made, as already mentioned, a 

 large number of experiments on obtaining ice in the mould* free 

 from air bubbles. We were ultimately successful, and, though our 

 experiments proved to be of little use for their immediate object, 

 they are of some permanent interest as tests of various methods of 

 obtaining air-free water, so we shall describe a few typical ones. 

 Main, the previous winter, boiled the water and let it freeze, then 

 melted it in the mould, boiled it, and let it freeze again. The result 

 was clear ice, except for " a small core of minute bubbles up the axis 

 of the cylinder." By Main's advice we procured an air-pump adapted 

 to exhaust the air from the mould. Between the pump and the 

 mould was a good stopcock, which would maintain the vacuum for 

 several hours. When in good order the pump would boil water at 

 40° C, or below. We found that this degree of exhaustion was far 

 from removing all the air, even when applied for five hours. Boiling 

 for half an hour, cooling sub vacuo, and freezing at atmospheric 

 pressure under oil was more successful, but not satisfactory. We 

 froze the water at atmospheric pressure to make the bubbles small, 

 having placed a layer of oil on the top to prevent air entering. The 

 next method proved much more effectual. We kept water sub vacuo 

 for twenty-four hours at about 70° C, and let it cool sub vacuo, only 

 admitting air after the freezing had begun. There were a few 

 exceedingly small bubbles visible at one end of the rod of ice. 

 Thawing this sub vacuo and keeping it again for twenty-four hours 

 sub vacuo at 70° C, we got rid of the last traces of air in the rod, 

 though there were a few in the large cone of ice. 



[We conclude that, to free water from air, it should first be boiled 

 till most of the dissolved air has escaped, and then left for a con- 

 siderable time without permitting any air to have access to its 

 surface. Boiling should be repeated at intervals to remove the air, 

 which gradually escapes from the water and mingles with the 

 aqueous vapour in the space above. It is probable that a high tem- 

 perature quickens the process. — July 6, 1888.] 



The utter irregularity of the crystalline structure of the mould ice 

 is an obvious consequence of the mode of formation. The first ice 

 formed, no doubt, is a layer on the surface, but the centre of this is 

 soon broken though by water forced up from below, owing to the 

 expansion in freezing. So what we observed in its various stages was 



# This was the iron mould used by Main to form a round column of ice 2 - 8 cm. 

 in diameter and 24 rm. in length, with a conical expansion at the lower end of 

 perhaps half the volume of the column. 



