19 6 THE EAKL OF BERKELEY ON SOME 



end of this period the density was found to be 1*2539 (temperature 61'00 C.). 

 Having reason to believe, from previous work, that saturation had been reached, the 

 temperature of the thermostat was lowered 1 C. and the solution stirred for another 

 3 hours; its density was then l - 2546 (temperature 59*85 C.). 



The next day it was 1*2555 (temperature 59*90 C.). 

 1-2572 ( 60-OOC.). 



1-2591 ( G0*00 C.). 



Between each observation about 10 hours' continuous stirring was given to the 

 solution, and all the time there had been about 5 cub. centims. of solid salt in contact 

 with the solution ; another 20 cub. centims. of salt was then added, and the stirring 

 continued for 12 hours, with a resulting density 1*2810 (temperature 60'00 C.). 

 And a further 12 hours gave T2813 (temperature GO'OO C,). 



On the other hand, the following shows the reverse phenomenon, i.e., that a 

 considerable time must elapse before a supersaturated solution attains its true point 

 of saturation. A solution of Na e SO 4 , saturated at GO C.. was heated to the constant 

 temperature of 75 C. (it must be remembered that Na^SO^ is more soluble at GO C. 

 than at 75 C.) and stirred at the rate of 13 revolutions per second in contact with 

 the anhydrous salt for 3^ hours; the density of the solution was found to be 1*2738 

 (temperature 7o"00 C.). The next day, after 12 hours' stirring, the density was 

 1-2729 (temperature 75'00 C.); 20 cub. centims. of boiling water was then added 

 (if cold water had been added, as before explained, the solution would have become 

 supersaturated), and after 12 hours' stirring its density was 1*2727 (temperature 

 75-00 C.). 



Where it was suspected that the solutions, when at the higher temperatures, might 

 decompose non-reversibly, the observations for the lower temperatures were first 

 recorded, in the manner already outlined, and those for the higher temperatures were 

 obtained by heating to the constant temperature required and stirring the solution 

 until the density was constant ; the temperature was then lowered by 1 or 2 degrees, 

 and after a sufficient length of time the density again determined ; the mean of this 

 last observation, and of the constant density first obtained, was considered to be the 

 density of a solution saturated at the mean of the corresponding temperatures. 



The following are the important points to be observed in obtaining a saturated 

 solution : 



(1) A sufficiency of solid salt should always be in contact with the solution ; 



(2) A thorough stirring should be continuously kept up ; 



(3) A sufficient length of time should be allowed to elapse before taking the 



required observation. This last condition seems to depend on the nature 

 of the salt, the speed of stirring, and on the temperature. 



I had hoped to have been able to determine both the rate of attainment ot 

 saturation, and the time at which it is attained, by observing the change in the 



