CUTTING ICE. 



and the weight drops on to the table, the piece 

 of ice beyond the cutting-line is not severed, 

 and does not fall, but remains attached, as 

 before, to the main bar of ice unless, that is, 

 the experimenter is performing in public, and 

 has imprudently announced the result to the 

 audience beforehand. In that case the cut 

 piece of ice is only too apt to drop down at the 

 same time as the weight, when the lecturer and 

 the audience are attentively watching to see it 

 stick on. 



At his next public experiment the lecturer 

 does not explain the reason of the cut piece 

 sticking on until it has stuck on, which under 

 those conditions it will usually consent to do. 

 It can then be explained as a thing that was 

 expected with the utmost confidence. 



The explanation depends on the principle of 

 regelation, which is also seen at work in snow- 

 balls and glaciers. It has been explained that 

 pressure melts ice. When a handful of snow is 

 crushed together, some of the crystals are 

 partly melted by the pressure. When the 

 pressure is relaxed the melted ice re-freezes, 

 and so the whole mass is to some extent bound 

 together. Thus we get a snowball. It re- 

 quires great pressure, however, to melt ice, 

 and greater as the temperature is lower, every 

 degree of cold requiring an enormous addition 

 to the pressure, so that we are not strong enough 

 to make snowballs of very cold snow, and the 



125 



