ifu; 



alcly ])('l()w llic t'rccziiin' point and at the cxti-cnic lower 

 limit of suhcoolini;-, Avliile a vigorous shaking was iiicffoc- 

 livo in the internu'diato zone of temperatures. He con- 

 tends also that spontaneous erystallization occurred most 

 of the time between -0.72° and -2.9° and l)eiween -5.01° 

 and - S.17° and less frecpiently in the intermediate region. 

 To establish the existence of such zones of stability, more 

 statistical data would be necessary. 



We shall mention, to finish this section, a curious but 

 too concisely described observation of Wartman (1860). 

 Some water that he had left during an entire night, at 

 about - 4° in a glass container 31 cm. high and 16 cm. wide 

 (diameter ?), had stayed liquid, but Avhen he took the con- 

 tainer to empty it, in the morning, three walls of ice sud- 

 denly formed, making 60° angles ( f ) with each other, 

 adhering at one side to the wall of the container and 

 oblique to the axis of the latter. 



5. Capillarity. Some of the oldest investigations on 

 subcooling were made by Gay Lussac (1836) who observed 

 that water can be subcooled to - 12° when it is enclosed in 

 small tubes. 



Later Despretz (1837) reported that the congelation of 

 a liquid is "retarded" by ten or twelve degrees if the liquid 

 is enclosed in a thermometric tube or even in a tube hav- 

 ing a diameter of one centimeter. 



Mousson (1858) exposed to temperatures from -5° to 

 - 7°, 8 tubes which varied in diameter from 0.187 to 2.503 

 mm. and which were tilled with water and sealed with seal- 

 ing wax. After a night at the temperatures indicated, 

 water was frozen in all the tubes which had a diameter 

 larger than 0.9 mm. while it was liquid in the smaller ones. 

 In other experiments he sprayed droplets of water less 

 than 0.5 mm. in diameter on a dry surface and observed 

 that the smaller the drops the longer they stayed subcooled. 

 The fact that, in clouds, water remains liquid at rather low 

 temperatures is attributed to the very small size of the 

 droplets. 



Dufoni- (1861) studied the subcooling properties of small 



