6 M. Gai/-Ltmac 071 the [Jan. 



its liquidity below the point of congelation cannot be ascribed to 

 affinity, as in the saline solutions, where it may be supposed that 

 the supersaturation is owing to the affinity of the salt for its 

 solvent ; but as the effects are the same in circumstances abso- 

 lutely similar, it is very probable that their causes are the same. 



There has been long entertained an opinion respecting the 

 permanence of the saturation of saline solutions which 1 have 

 never adopted ; because it does not appear to me sufficiently 

 demonstrated. The opinion is, that water saturated with a salt 

 is capable of depositing a portion of it when left at rest, though 

 its temperature does not alter. But M. Beudant in his memoir 

 on the causes which may produce changes in the form of the 

 crystals of the same mineral, has cited several facts in support 

 of this opinion, which he appears to adopt. I consider it, 

 therefore, as necessary to discuss it here. 



" 1 have remarked," says M. Beudant (Ann. de Chim. et de 

 Phys. viii. 16), " that very regular crystals may form without 

 any evaporation whatever in solutions otherwise very dilute ; but 

 it does not appear that all the salts are in this state. To satisfy 

 myself on the subject, I placed dilute solutions of different salts, 

 all at the same degree of density, in flaggons completely filled, 

 and well stopped, which I left in a press. On visiting them 

 long after, I obs^erved that they were all equally full, and that 

 consequently there had been no evaporation ; but in several of 

 them the salts had crystallized. I remarked that it was precisely 

 those which had the greatest degree of cohesion, as the sulphate 

 of potash, alum, borax, muriate of barytes &c. ; wiiile those 

 whose cohesion was weaker, as nitrates of potash and ammonia, 

 sulphates of ammonia and iron, muriate of soda, &c. had not 

 crystallized." 



In the Ann. de Chim. et de Phys. vii. 79, 1 have already 

 stated experiments in my opininn very decisive, showing that 

 saturated sahne solutions, whose temperature is constant, do not 

 deposit salt ; and that they remain homogeneous through their 

 whole extent ; but since that time I have made new experi- 

 ments to answer all objections that could be advanced, and the 

 result of them has been precisely the same as that of the former 

 ones. 



I took two glass tubes, two metres in length, and three centi- 

 metres in diameter. I put into one a solution of nitre saturated 

 at the temperature of the caverns under the Observatory, and 

 into the other a solution of sea salt equally saturated. Two 

 other tubes were filled with similar solutions, in which there was 

 not more than four per cent, of each salt. These tubes well 

 closed remained six months in the caverns of the Observatory in 

 a vertical situation. At the end of that time I determined by 

 evaporation the quantity of salt contained in the water of the 

 upper and lower part of each tube, and 1 found that the solutions 



