1820.] Solubiliti/ of the Salts in Water. 7 



were perfectly homogeneous. I shall add hkewise that crystals 

 of nitre suspended below the surface of a saturated solution of that 

 salt, underwent no sensible diminution, during an inter\al of 

 more than two years, that they have remained in the constant 

 temperature of the caverns of the Observatory. 



The facts which I have just stated are opposite to those of 

 M. Beudant ; and if they were produced by the same cause, 

 they could not exist at the same time. It remains for me to 

 show that in fact they are owing to different causes. 



M. Beudant does not give sufficient details respecting his 

 experiments ; but they can be perfectly well imitated by taking 

 saline solutions saturated or not, letting them cool for some 

 time, and then bringing them back to their original temperature . 

 If the cooling has been sufficiently low to supersaturate the solu- 

 tions, they will have crystallized in their lower parts. If now 

 the temperature be raised, the crystals will not be completely 

 dissolved provided we do not agitate ; because the liquid in 

 contact with them will acquire by saturating itself more density 

 than the upper portions. It will, therefore, remain at the bottom 

 of the vessel, and will prevent the crystals from dissolving. 

 The solutions which did not crystallize in M. Beudant's experi- 

 ments were not sufficiently saturated to be carried beyond the 

 limit of saturation by the cooling which they experienced. It 

 appears to me certain from these facts, that the crystallizations 

 which M. Beudant observed were entirely owing to the cooling 

 of the saline solutions. 



I shall now give an account of the experiments which I have 

 made on the solubihty of the salts. 



Hanng saturated water with a salt at a determinate tempera- 

 ture, as I have explained above, I take a matrass capable of 

 holding 150 to 200 grammes of water, and vv-liose neck is 15 to 

 18 centimetres in length. After having weighed it empty, it is 

 filled to about a fourth part with the saline solution, and weighed 

 again. To evaporate the water, the matrass is laid hold of by 

 the neck by a pair of pincers, and it is kept on a red-hot iron at 

 an angle of about 45°, taking care to move it continually, and to 

 give the liquid a rotatory motion, in order to favour the boiling, 

 and to prevent the violent bubbhng up, which is very common 

 with some saline solutions, as soon as, in consequence of evapo- 

 ration, they begin to deposit crystals. When the saline mass 

 is dry, and when no more aqueous vapours are driven off at a 

 heat nearly raised to redness, I blow into the matrass by means 

 of a glass tube fitted to the nozzle of a pair of bellows, in order 

 to drive out the aqueous vapour wliich fills it. The matrass is 

 then allowed to cool, and weighed. I now know the proportion 

 of water to the salt held in solution, and this is expressed by 

 representing the quantity of water to be 100. Each of the 

 following results is the mean of at least two experiments ; 



