20 THE ROYAL SOCIETY OF CANADA 



One large piece 6" X 28" X 40", furnished the largest number of 

 individual blocks. The direction of the crystalline axis was deter- 

 mined by knowing the plane of freezing, and the various blocks were 

 cut by a hand-saw accordingly. The ends of the blocks were 

 melted smooth on a hot plate to fit the plates of the press. When the 

 pressure was applied, considerable melting resulted on the ends. 

 The blocks were kept outside at a temperature ranging from -8°C. 

 to 0°C. and were brought into the laboratory for the tests. One 

 interesting fact was noted, that the blocks were heard to crack at a 

 pressure, approximately one half the ultimate crushing force. It was 

 repeatedly observed that as soon as the pressure was increased suffi- 

 ciently to cause the first audible crack, the block appeared to stiffen, 

 and the pressure ran up much quicker, with much less melting. In 

 some cases the melting appeared to cease altogether. I. concluded 

 that the giving way of the ice under pressure allowed the melted ice to 

 run into the cracks, where it must have frozen and cemented the block 

 more firmly. I was unable to see any of the cracks which we could 

 distinctly hear at half the crushing pressure. 



The only effect of varying the position of the axis of the ice with 

 respect to the direction of the pressure, appeared to be the way the 

 block burst. When the axis was parallel to the applied pressure, 

 the ice burst sideways into innumerable long needles, resembling a 

 cake of ice which has all but fallen to pieces in the sun. The cake 

 fell to pieces on being removed from the press. When the axis was at 

 right angles to the applied force, the block cracked lengthwise and 

 transversely without shattering. 



Fig. 1 shows a cake crushed in the direction of the axis, while 

 Fig. 2 shows the reverse effect. 



In Table II I give the results of the tests. 



It will be seen that the mean value for all the tests for parallel 

 axis is a little higher than the mean for perpendicular axis, but the 

 difference is too small to make it possible to draw any definite 

 conclusions. 



The results show considerable variation, which may be purely 

 accidental or may have some bearing on the character of the initial 

 distribution of the pressure. 



The relation between the first cracking of the ice and the final 

 crushing force is one which must be further investigated. 



The question of the relation of temperature to the crushing 

 strength is one of importance. It has been assumed that ice becomes 

 stronger at low temperature. The hardness of ice increases consider- 

 ably as the temperature falls to 0°F. but I am inclined to think that 

 the ice also becomes more brittle. In the neighbourhood of the freez- 



