and Viscosity of Supercooled Liquids. 



243 



wider glass tube forming an air-jacket was tried. This was 

 found to diminish the velocity considerably, as we should 

 expect owing to the increased difficulty with which the heat 

 developed could escape. Thus with the water-bath at 32° 

 the velocity with the additional tube was 0*88 ^^, and without 



J mm.' 



the additional tube it was M?^. 



min. 



To test the validity of the formula v=C^-, the viscosity of 



supercooled salol was measured at a number of temperatures. 

 The apparatus was of the ordinary form, consisting of two 

 small glass bulbs connected by a capillary and bent round 

 into a U-shaped tube. The time required by the meniscus 

 of the liquid to pass between two marks, one above and one 

 below the upper bulb, was determined whilst the apparatus 

 was fixed in a water-bath at a known constant temperature. 

 The capacity of the upper bulb was about 2 cub. cent., and the 

 time for water at 15° C. was 24*5 seconds. 



The open ends of the U-tube were connected by india- 

 rubber tubes to bulbs containing calcium chloride to prevent 

 moisture getting to the liquid. When the liquid began to 

 solidify, it was melted by putting the tube in hot water. The 

 following table gives the results obtained for the viscosity of 

 salol. 



Salol. 



Melting-point, 41° C. 

 Density 1-195 at 35° C. 



Temperature. 



Time. 



Viscosity. 

 (Water at 1 5° = 1.) 





sees. 





420 



149-3 



7-27 



359 



1965 



9-57 



290 



278-5 



13-57 



25-0 



350-0 



1705 



19-5 



519-2 



25-29 



15-0 



649-0 



3i -eo 



These results are shown graphically in fig. 3. 



s 

 The formula v=G^ indicates that the velocity should in- 

 crease proportionally to the supercooling when it is small. 

 Now the relation observed between the velocity and super- 



