[BARNES & LUCAS] GROWTH OF ICE CRYSTAL 39 



Third Experiment. 



It was decided to try one more experiment with liquid air as a 

 refrigerant and cracked ice as a packing. This was carried out with 

 almost identical results as in the first case. The curve shows the same 

 change in the rate of rise occurring after the same interval of time. 



In this case the change in volume of ice and water was .1635 c.c. 

 giving the volume of ice formed 1.81 c.c. The rate of increase is 

 identical with that in the first experiment. This is shown in Fig. 4. 



Fourth Experiment. 



The rate of increase using salt and snow at — 15° C. as a refri- 

 gerant was the same as the second case. 



Theoretically the increase in the ice mantle by the gradual growth 

 illustrated by these experiments is very interesting. If we regard 

 water in the neighbourhood of the freezing point as a solution of ice 

 particles, as it is necessary to do to account fox the variation of the spe- 

 cific heat and the occurrence of the point of maximum density then 

 it is quite in accord with our knowledge of the natural growth of crystals 

 in a saturated solution. At the freezing point we must admit that 

 tihe water is saturated with ice particles, Once having started the crys- 

 tallization of ice, the process continues very slowly as long as the water 

 is kept at the freezing point by a suitable freezing point mixture. 



Why the rate of change of reading should be different for the two 

 types of refrigerants we are not as yet in a position to explain. It may 

 possibly be due to a variation in the ice density superimposed on the 

 normal rate of growth. 



There was a decided difference in the appearance of the ice mantle 

 which was formed by the two methods. That formed by the liquid air 

 was beautifully clear at the start, but after several hours was seen 

 to be traversed by fine radial cracks. The mantle formed by the salt 

 and snow was cracked at the start and presented a very mottled appear- 

 ance. 



The change in the rate of increase in experiments one and threo 

 may also be explained by a change of density. 



Further experiments are necessary, however, before anything 

 definite can be said on these points. 



