Reaction before Complete Equilibrium. 83 



most of the reactions begin and finish in a fraction of a 

 minute, the slowest not lasting longer than a very few 

 minutes ; and as the temperature of the liquid during the 

 reaction passes from T—T^ = about 1° to T — T^ = about o, l, 

 at is evident that under these experimental arrangements the 

 heat evolved or absorbed can be safely neglected. We 

 arrive at the same conclusion by a calculation of the total 

 heat absorbed or given off by the liquid in each individual 



•case, using the equation -=- =G(T g — t), given above. 



E. Some Further Details. 



a. A large copper bath (Ba) with double walls was used 

 for the investigation of the separation of salts from super- 

 saturated solutions. The space between was filled with 

 petroleum and the outer walls were covered with sheets of 

 asbestos, a circle below being left uncovered for the burner. 

 One thermometer (1) was in the petroleum, the other in the 

 air-chamber (2). The beaker with the liquid and stirrer were 

 placed in the inner chamber on thick felt. The rod of the 

 •stirrer was then fixed to the guide (g) which was moved by 

 .an electromotor, separated from it by a ring of asbestos, in 

 order to bring to a negligible quantity the heat passing from 

 the stirrer to the guide during the time of the reaction. The 

 number of ups and downs of the stirrer was 36 per minute. 

 The temperature of the inner chamber being kept as far as 

 possible constant, the velocity of cooling of 4^ litres of a 

 solution, almost saturated at the temperature of the air- 

 chamber, was measured and the convergence-temperature for 

 the given experimental conditions determined. (In this case 

 the conditions were : — 36 stirrings per minute, temperature 

 of room T, temperature of the petroleum T', of the air- 

 chamber T", 4t\ litres of liquid, &c.) A solution saturated 

 above the convergence-temperature was then prepared, so 

 that the temperature of the liquid after the separation of the 

 salt should be as near to the convergence-temperature as 

 possible, 4^ litres of it were poured into the beaker, and the 

 separated crystals dissolved by stirring the liquid and warming 

 it a few degrees above the temperature of the solution. With 

 this solution the final experiment was made. The beaker 

 with the solution was replaced in the air-chamber and stirred. 

 As the temperature of the liquid approaches the point of 

 saturation stirring must cease, and the liquid be allowed to 

 cool without any disturbance. A very slow and cautious 

 movement of the stirrer may be occasionally made to equalize 



G2 



