Freezing-Points and the Freezing-Point Methods. 479 



The liquid in the beaker was stirred by the porcelain stirrer 

 about 30 times per minute, and the temperature was measured 

 by the fine -Ol^thermometer. The liquid-bath was stirred 

 from time to time to get a uniform temperature. Hot water 

 was introduced into the beaker till the temperature of the 

 liquid rose about 1°"2, and the time required for the passing 

 of the Hg-thread from one point of the thermometer-scale to 

 the other was measured to £ second. 1 thus found for 1250 c.c. 

 water in a liquid-bath, C to be about 20 times greater than 

 in the air-bath. Now the velocity of cooling of a given 

 quantity of a liquid is inversely proportional to its quantity 

 or volume and directly proportional to its conducting surface. 

 So e.g. 100 c.c. cool about 2*32 times quicker than 1250 c.c. 



III. The Determination of the Temperature of the Air-bath. 



This was determined and found to be about *4° higher than 

 that of the ice-bath. The measurements have been made in 

 such a way that the yoo° thermometer was at first overcooled 

 and allowed to rise in the air-bath to its maximum, then it 

 was allowed to fall from a higher temperature to the air-bath 

 temperature ; thus the same values have been obtained for the 

 temperature of the bath. 



At the time when this paper was communicated to the Royal 

 Society, I was only able to give the minimum values for the 

 velocity of ice-melting and ice-separation. The reason was : 

 the velocity with which my ■01°-thermometer assumed the 

 temperature of the liquid was found to be 3° min. -1 when the 

 Hg-thread falls, and 5° min. -1 when it rises. Again, when 

 the thermometer was in a liquid in which ice separated or 

 melted, about the same values of 3° min. -1 and 5° min. -1 were 

 obtained. It followed from this that the velocity of ice- 

 separation or of ice-melting expressed in degrees of tempe- 

 rature could not be smaller than 3° min. -1 and 5° min. -1 , but 

 could be greater. I have since continued my investigations 

 of the velocity of ice-melting and ice-separation. 



IV. On the Velocity of Crystallization of overcooled Liquids 



and Solutions. 



A beaker with 1250 c.c. of liquid is placed in a zinc vessel 

 which serves as an air-bath. This placed in a liquid -bath, 

 the temperature of which is regulated to the given temperature 

 of the room, so that the cooling of the liquid stirred is so 

 slow that for the time of the reaction it may be neglected, or 

 only a small correction required. A very effective stirrer is 

 required which does not allow the separated solid to settle on 



2N2 



