ON THE WATER IN MURIATE OF AMMONIA. 187 



librated in Mr. Murray's experiment was not derived from 

 the mviatic gas, but from the atmosphere." 



It might have been expected, that the first step these That muriate 

 gentlemen would have taken, when they assigned this as the attractTwater 

 source of the water obtained, would have been to prove its from the air 

 reality; and to show by fxperimental evidence, that the salt ^ e ' e " pro ^.j 

 on which they operated has the power of attracting- water 

 from the atmosphere. No such evidence however is given ; 

 but the existence of this property is inferred from the result 

 of an experiment, which may have arisen from causes alto- 

 gether different. Admitting lor a moment the accuracy of The experi- 

 their experiment, the obtaining water when the salt is heat- ^"5." Davies 

 ed after exposure to the air, while none is obtained when it not conclusive, 

 is heated without this exposure, is no proof, that the water 

 in the former case has been absorbed from the atmosphere ; 

 for in making the experiment in these two modes, the sole 

 difference is not the admission or exclusion of the air, nor is 

 the sole operation of the air that of affording moisture ; there 

 are other circumstances of difference equally important, and 

 which it is easy to perceive must influence the result. 



Thus the principal difficulty in the original experiment, so Principal dif- 

 as to render it conclusive, arises from the volatility of the • -^j eX ., e _ 

 ammoniacal salt, and the inconsiderable interval of temper- riment. 

 ature between that point at which any water it may contain 

 can be separated from it by heat, and that point at which the 

 salt itself will pas9 into vapour. In consequence of this it 

 most require a nice regulation of temperature to obtain the 

 owe effect without the other; and from this very circum- 

 stance, even had water not been obtained in the experiment 

 as I first performed it, it could not have been affirmed, that 

 it did not exist in the salt. Now this difficulty it is obvious This partial- 

 is much greater, when heat is applied lo a thin layer of salt ia,1 y exists in 

 encrusted over the whole internal surface of a retort, than executing it- 

 when it is applied to the same quantity of salt collected in 

 mass at the bottom -:>f a retort; and it must indeed be nearly 

 impracticable to apply the heat in the former case with such 

 a precise adaptation to the relative volatilities of the water 

 and the salt, as to expel the former without volatilizing the 

 latter. If the heat therefore is kept sufficiently low ,uot to 

 volatilize the salt, and especially if care is taken to keep it 

 O 2 still 



