216 On Daiion's Method of Meastiritig Water of Crystallization. 



isomorphous salt may contain a small quantity of another salt 

 of the same class combined with it; and if so, the separation 

 of these by crystallization will be exceedingly difficult, if not 

 impossible. The ordinary reagents do not show one of two 

 isomorphous salts in the same solution when the other is pre- 

 sent in great excess, though the greatest source of error ap- 

 pears to be in the quantity of water influencing the condensa- 

 tion produced by each salt which is not dissolved in atomic 

 proportion. 



On continuing this research on the increase or diminution 

 of volume, I found that a chloride, iodide, or sulphuret of a 

 metal, or by the other nomenclature, the hydrochlorate, &c. 

 of the oxide of a metal, when dissolved in water occasioned a 

 condensation, while the hydrochlorate of ammonia caused an 

 expansion ; and if this line of demarcation could have been 

 established, it would have given rise to one of the most import- 

 ant laws in chemical science. For when hydrochloric acid is 

 added to the oxide of a metal, "for instance soda," we do not 

 know whether, in the presence of water, a chloride of the metal 

 is formed or the muriate of soda; though the muriate of am- 

 monia is considered by some to be a chloride of ammonium, 

 yet the theory now generally adopted is, that when muriatic 

 acid unites with ammonia, the muriate of ammonia is formed; 

 and when the same acid unites with a metallic oxide, it forms 

 a chloride ; and as the hydrochlorates expand while the chlo- 

 rides contract, we may infer that a salt, the acid of which con- 

 tains hydrogen, expands in solution, while the haloid salts 

 contract; the combination of the hydracids with ammonia 

 thus forming muriates, &c., and of the same acids with metallic 

 oxides, forming chlorides, iodides, &c. 



If this most important law could have been established, we 

 should then have a ready and easy way of distinguishing be- 

 tween the hj'drochlorates and chlorides in solution, and not 

 only of distmguishing between these two varieties ; for if the 

 muriates expand while the chlorides contract, the various com- 

 binations of chlorine and muriatic acid following the same 

 order, we should then, perhaps, have had an opportunity of 

 distinguishing the nature of bleaching powder and the like 

 compounds, and thence inferring if chlorine can combine at 

 all with the oxide of a metal ! 



But unfortunately for this law, on proceeding with the ex- 

 periments, I found that several of the ammoniacal salts did 

 not correspond with it, so that this law, like the one devised 

 by the superior skill and ingenuity of Dr. Dalton, could not 

 be practically established. 



I think Dr. Dalton's law relating to the water of crystalli- 



