C S J 9 ] 
eonfifts in fitting an empty vefiel H, to the appa-- 
ratus defcribed before. See figure i. By this means 
the water is forced into H, and by the ftopper at 
L it may be emptied, and put back into the vefiel F, 
the crooked tubes D and I are fitted to H, by a 
cork with two oppofite femicircular notches as at; 
K, and then well covered with lute.. 
Experiment I. 
On the diftillation offal ammoniac with quick lime. 
12 lb * of Britifh fal ammoniac, and 26 th of 
quick lime were powdered, mixed, and put into 
the iron body A (fig. 3.) ; and when the apparatus -f- 
was luted, a gallon of water was poured on it through 
the orifice (b), which was immediately flopped ; the 
lime growing hot produced a vaft quantity of ela- 
ffic air, which though highly charged with volatile 
alkaly was condenfed by the water in F, fig. 2. the 
air only efcaping at the top of this vefiel with hardly 
any fenfible volatile alkaline fmell. Next morning, 
all being cold, another gallon of water was added 
as before, and a very flow fire made under the body 
for 14 hours, in which time there diftilled near a 
pound of volatile alkaly; the fire was then made 
itronger, and continued in that ftate for twelve hours 
more, in which time there was obtained, together 
with what was firft diftilled, 8 lb 4 of volatile alkaly, 
ftrong and fit for Eau de luce; this was taken out 
of the bottle and fet apart. The veflels being cool, 
* In all the experiments averdupois weight was made ufe of. 
f The fpout of the ftone head belonging to the body A, 
figure 3, is to be luted to the receiver B, figure 2 . 
two 
