195 



a ])oliavior is consistent with the theory that crystalliza- 

 tion occurs as an event which follows the law of prob- 

 ability. 



As to the degree of force necessary to induce crystal- 

 lization, several observers have noted that it varies with 

 the degree of snbcooling. Fahrenheit, who is considered 

 the discoverer of the snbcooled state, remarked, already 

 in 1724, that strongly subcooled water could be made to 

 crystallize by a shock, while slightly subcooled water could 

 not. 



According to Mousson (1858), water can be maintained 

 subcooled to - 12° or - 15° in a vacuum but then the 

 slightest disturbance causes solidification. The same au- 

 thor could obtain a high degree of snbcooling in tiny water 

 droplets laid on a cold plate, but touching them with the 

 point of a pin caused them to freeze at once. 



De Coppet (1907) observed that a strong stirring could 

 not cause congelation of salol subcooled 30 to 35° below 

 the freezing point, while 10 degrees low^er, crystallization 

 took place readily, on the slightest disturbance. The same 

 author said that when he gave to sealed tubes which con- 

 tained supersaturated solutions of sodium sulfate, the 

 slight inclination necessary to bring an air bubble present 

 in them from one end to the other, he sometimes observed 

 solidification. 



A study of the minimal disturbance (friction) which is 

 effective in causing the precipitation of supersaturated 

 solutions, has been made by Young (1911). 



Sensitivity to mechanical disturbance, as well as sensi- 

 tivity to several other factors, depends on the nature of 

 the material. Wright and Taylor (1921) observed that, 

 while subcooled potatoes can be induced to freeze by a 

 slight tap, berries subcooled to the same degree did not 

 congeal under vigorous tapping. 



Mez (1905) claims to have observed a rather peculiar 

 behavior in the cell sap of Impaileus, pressed out, filtered, 

 boiled, subcooled and subjected to shaking. A slight dis- 

 turbance started crystallization at temperatures immedi- 



