124 EFFECTS of HEAT 
experiments, a confiderable additional fufibility has been com- 
municated to the carbonate. : 
THESE experiments, and many others made about the fame 
time, with the fame fuccefs, clearly prove the efficacy of wa- 
ter in aflifting the compreiflion ; and refults approaching to thefe 
in quality, obtained, in fome cafes, by means of a very {mall air- 
tube, fhew that the influence of water on this. occafion has 
been merely mechanical. ; 
Durinc the following fummer and autumn 1803, I was oc- 
_ cupied with a different branch of this fubje@, which I fhall 
foon have occafion to mention. 
In the early part of laft year, 1804, I again refumed the fort 
of experiments lately defcribed, having in view principally 
to accomplifh abfolute compreffion, in complete imitation of 
the natural procefs. In this purfuit, I did not confine myfelf 
to water, but made ufe of various other volatile fubftances, 
in order to aflift compreflion; namely, carbonate of ammo- 
nia, nitrate of ammonia, gunpowder, and paper impregna- 
ted with nitre. With thefe I obtained fome good refults, but 
none fuch as to induce me to prefer any of thefe compreffors 
to water. Indeed, I am convinced, that water is fuperior to 
them all. I found, in feveral experiments, made with a fimple 
air-tube, without any artificial compreflor, in which a very 
low red-heat had been applied, that the carbonate loft one or 
one and a half per cent. Now, as this muft have happened in 
a temperature fcarcely capable of inflaming gunpowder, it is 
clear, that fuch lofs would not have been prevented by its pre- 
fence: whereas water, beginning far below rednefs to aflume 
a gafeous form, will effectually refift any calcination, in low 
as well as in high heats. And as the quantity of water can 
very ealily be regulated by weight, its employment for this 
purpofe feems liable to no objection. 
On 
