with Observations on its Chemical Constitution. 109 
temperature in which the experiment is performed: when weighed 
immediately on its formation, in an exhausted vessel, it gains 
no weight from exposure, but remaius the same after a number 
of hours; and when exposed to the air in the freest manner, it 
remains, after many days, perfectly dry. It was further shown, 
that when the other circumstances of the experiment are the 
same, it yields no larger portion of water when it has been ex- 
posed to the air, than it does without this previous exposure. 
And, lastly, it was proved, that when the salt has been formed, 
and is heated without the air having been admitted, water is ob- 
tained from it. This last result was even at length admitted by 
those who had advanced the opposite assertion, in an experi- 
ment performed witha view to determine the fact. The quan- 
tity of water was indeed less than what is procured in the other 
mode; ; but this was obviously owing to the circumstances of the 
experiment being uniav ourable to its expulsion,—more particu- 
larly to the difficulty of applying a regulated temperature to a 
thin crust of salt, so as to separate the water without volatilizing 
the salt itself,—and to the effect arising from the whole internal 
surface of a large vessel being encrusted with the salt; so that, if 
the heat is locally applied, the aqueous vapour expelled from 
_one part is in a great measure condensed and absorbed at an- 
other; or, if the heat is applied equally, is retained in the elastic 
form, and, as it is cooled, is equally condensed. Accordingly, 
when the experiment was repeated, obviating these sources of 
error as far as possible, the water obtained was in larger quan- 
tity. And as no fallacy belongs to the conducting the experi- 
ment in the more favourable mode in which it was first per- 
formed, (the assertion of the absorption of water from the air 
being altogether unfounded,) the quantity procured in that mode 
is to be regarded as the real result *. 
The argument was maintained, that the water might be de- 
rived from hygrometric vapour in "the gases submitted to experi- 
ment. This it was easy to refute. Dr. Henry had shown that 
aminonia after exposure to potash, and muriatic acid after ex- 
posure to muriate of lime, retain no trace of vapour whatever. 
And these precautions had been very carefully observed. The 
assertion was brought forward, too, only to account far the mi- 
nute quantity of water obtained in that mode of conducting the 
experiment which affords the least favourable result ; and, were 
it even admitted to all the extent to which it can be supposed 
to exist, is inadequate to account for the larger quantity ob- 
tained in the other. 
* Nicholson's Journal, vol. xxaii. p. 186 &e.; vol. xxxiv. p. 271. 
That 
