best Mefljods of producing artificial Cold. 277 
the dotted line at b, fig. 4.) being cemented into the tin, and 
thereby forming that part in which the nitrous acid was first 
cooled, and the mixture afterwards made in which the quick- 
silver was frozen : but from the trouble and impediments aris- 
ing from letting out the mixture, and clearing the bottom 
from the lumps of ice, &c. adhering to it, I was led to the ad- 
dition of the other part (fig. 5.) by which all these difficulties 
are got rid of, and it is besides a much more comfortable and 
neat way of conducting it ; the upper part which contains the 
nitrous acid being lifted off and placed on the table, imme- 
diately before the powdered ice is added. 
The whole of this apparatus may be of tin, that part only 
(when the cooling mixtures are made without using any corro- 
sive acid) in which the acid mixture is to be made, being pre- 
viously coated in the manner above mentioned ; or a thin glass 
tumbler of a proper size may be cemented in. 
I have occasionally used a thin glass tumbler for the mix- 
ture in which the quicksilver is to be frozen, immersing it with 
the acid in a frigorific mixture till the acid is sufficiently 
cooled, then adding the ground ice to it, previously removing 
the tumbler out of the frigorific mixture, as in the experiment 
above mentioned ; this simplifies the apparatus, but is less con* 
venient on many accounts. 
The scale of this apparatus may be diminished or increased 
at the will of the operator ; for there is no doubt that a small 
quantity of quicksilver may be frozen at any time with one- 
fourth of this quantity, with an apparatus of this kind, by any 
one conversant in such experiments. 
I have frequently frozen quicksilver, by mixing together, at 
o°, three drams of ground ice with two drams of nitrous acid. 
mdccxcv. O o 
