9 
combinations oj Oxy muriatic Gas and Oxygcne, &c. 
I regarded the loss of weight, as the indication of the quantity 
of moisture. 
I am acquainted with no experiment on record, in which 
water has been actually collected from the ignited fixed 
alkalies, and this appeared necessary for the complete eluci- 
dation of the subject. 
I heated together in a green glass retort, 40 grains of 
potash, (that had been ignited for several minutes), and 100 
grains of boracic acid, which had been heated to whiteness for 
nearly an hour. The retort was carefully weighed, and con- 
nected with a small receiver, which was likewise weighed ; 
the bulb of the retort was then gradually heated till it became 
of a cherry red; there was a violent effervescence in the 
retort, a fluid condensed in the neck, and passed into the re- 
ceiver. When the. process was completed, the whole of the 
retort was strongly heated ; it was found to have lost 61 - 
grains, and the receiver had gained 5.8 grains. The fluid 
that it contained was water, holding in solution a minute quan- 
tity of boracic acid, and when evaporated, it did not leave an 
appreciable quantity of residuum. 
A similar experiment made upon soda, heated to redness, 
but in which the water collected was not weighed, indicated 
22. q of water in 100 parts of soda. 
It may be asked, whether part of the water evolved in these 
processes might not have been produced from the boracic acid, 
or formed in consequence of its agency ; but the following 
experiments shew that this can not be the case in any sensible 
degree. 
I heated 8 grains of potassium, with about 50 grains of 
boracic acid, to redness in a tube of platina, connected with a 
MDCCCXI. C ' c 
