the Laws of Affinity. 397 



quantitv, then, determine the formation of the different fahs, 

 cither by a firft cryftallization, or by eryftal ligations which, 

 by fubtracting fome of the parts, mud alter the proportion, 

 and occafion a difference in the difpofition of the remaining 

 principles to cryftallize in the fucceiiive formation of the falts. 

 It is here that the application of the theory of the quiefcent 

 and divellent affinities may produce many errors, by leading 

 us to conclude, from the commencement of the phenome- 

 non, that the fubfequent effects will be all of the fame de- 

 fcription, though in reality a fuccefiion of oppofite combina- 

 tions may be eftablifhed, according to the forces which are 

 made to act. at the moment of their refpe&ive feparations. 

 The joint confideration of the difference of folubility, and 

 of the proportions employed, or which vary at the different 

 periods of an operation, muft confequently be our fole guide 

 in the explanation of the fucceffive formation of different falts 

 which have no great difference in folubility. It is true, never- 

 thelcfc, that the mutual action of the fubftances themfelves 

 may produce fome difference in the refults, which would be 

 indicated by the preceding obfervations. 



-All fubftances in a ftate of folution exert a mutual action, 

 which increafes their folubility. Hence the reafon why it is 

 difficult to obtain, by a firft cryftallization, each fait in a 

 (late of purity; except in the cafe where it differs consider- 

 ably from the others in its force of cryftallization. Hence 

 arii'e the uncryitallizable residues which fucceed the crystal- 

 lizations wherein falts are found in the liquid, which poffefe 

 but little force of cohesion. But here alfo the confidera- 

 tion of the proportions and of the folubility will ferve to 

 predict the exiftence and composition of any uncryitallizable 

 residuum. While the fubftances are in folution, the action 

 which they mutually exert renders it eafv to expel an acid 

 from a combination, though, according to the received opi- 

 nion, it ought to aflame the place of another, fuppofed to 

 be weaker. Insolubility muft not be considered as an abfo- 

 lule property, but merely as relative to the liquid in which* 

 a precipitation is made. Thus a combination infolublc'iu 

 water, may lofe this property when the water holds alkali 111 

 folution. 



In all thefe experiments, ami in feveral others which I 

 have thought it ufelefs to defcribe, I perceived no change 

 iaturation, either after the mixture of the neutral falts, 01 

 after the feparation of the precipitates or crvitallizai' 

 which took place, except in the experiments P, Q, made 

 with a metallic iubftance. This permanent iiate of neutral- 

 ization, after the change of bafes which took place, feems to 



indicate 



