on the Crystallization of Salts. 47 



As it cannot be doubted tliat tbe air acts in some manner or 

 other upon the crystallization of the sulphate of soda, sipce the 

 latter takes place almost constantly in the open air, whereas in 

 vacuo it takes place under certain circumstances onlv; I have 

 supposed, in order to account for this ])haenomenon, that water 

 having the property of dissolving air, and having lost tliat which 

 it held in solution by tlie heat which ! had employed to dissolve 

 the salt, it might have been possible, when it was cooled, for it 

 to take up that which it had lost by heat, and that tlie absorp- 

 tion of this air had precipitated part of the sul])hate of soda, for 

 the same reason that one salt precipitates another from its solu- 

 tion. 



In order to verify this conjecture, I introduced into a flask 

 filled with w"ater deprived of air by boiling it, and inverted over 

 mercury, an air-bell occupying the fourteen hundredth part of 

 the capacity of the vessel. In twelve hours the air-bell had 

 diminished but little, and in four days it was not entirely ab- 

 sorbed. An absorj)tion so slow cannot account for the crystal- 

 lization of the sulphate of soda, when to a concentrated solution 

 of this salt in vacuo we add a little air. I shall remark how- 

 ever, that since a crystal determines the crystallization, it may 

 be possible that the absorption of the air, however small at first, 

 produces the precipitation of a little salt, and that afterwards 

 the crystallization might continue the same. 



However the ca'^e may be, it seems probaI)le that we may 

 describe the property which the sulphate of soda also has of 

 not crystallizing under certain circumstances, to the figure and 

 arrangement of its molecules^ which may be such that they 

 strongly resist any change of state. 



This property does not in my opinion differ in any wav from 

 that possessed by water of maintaining its fluidity below its true 

 freezing point, under circumstances nearly similar, nor from the 

 properties of several saline solutions, of remaining sometimes 

 saturated, and of crystallizing as soon as we shake them, or in- 

 troduce a foreign body. But supposing the figure of the mo- 

 lecules to be the chief cause of the foregoing phaenoinena, it is 

 nevertheless difficult to conceive how the presence of the air acts 

 in disturbing their equilibrium and favouring their junction. 



I think 1 have demonstrated that the solvent power of water 

 is totally independent of the pressure of the atmosphere; but 

 if there remain any doubts, they would be soon removed by 

 showing that there are very few saline solutions which possess 

 the property of remaining hvpcr-saturated under certain cir- 

 cumstances. A solution of phosphate of soda saturated at the 

 tenijjerature of 70" did not crystallize upon being cooled in a 

 baronjetrical tube, even after being slightly shaken : au air-bell 



did 



