400 Mr. Walker’s Experiments on the 
cold by folution in water; it will, by the following manage- 
ment, freeze water or creams at Midfummer. 
June 1 2th, 1787, a very hot day, I poured four ounces, 
wine meafure, of pump-water, at the temperature of 50° (it 
is well known that water at fprings retains nearly the fame tem- 
perature winter and fummer, viz. about 50°, to which tempera- 
ture the water may be reduced during the warmeft weather, 
by pumping off fome firff) upon three ounces, Avoirdupois 
weight, of the above powder (previoufly cooled by immerfing 
the veffel containing it in other water at 50°), and after ftirring 
the mixture its temperature was 14 0 ; fome water contained in 
a fmall phial, immerfed in this mixture, was confequently 
foon frozen. This folution was afterwards evaporated to dry- 
nefs, in an earthen veffel, reduced to powder, and added to the 
fame quantity of water, under the fame circumftances as be- 
fore, when it again funk the thermometer to 14 0 . Since that 
time I have repeatedly ufed a compofition of this kind for the 
purpofe of producing cold, without obferving any diminution 
in its effcdt after many evaporations. The cold may be cecono- 
mically kept up and regulated any length of time, by occa- 
fionally pouring off the clear faturated liquor, and adding frefh 
water, obferving to fupply it conffantly with as much of the 
powder as it will diffolve. 
The degree of cold at which water begins to freeze has been 
obferved to vary much ; but that it might be cooled twenty- 
two degrees below its freezing point was perfectly unknown to 
me until lately. I filled the bulb of two thermometers, one 
with the pureff rain-water I could procure, the other with 
pump water ; the water was then made to boil in each, until 
one-third only remained ; thefe were kept in a frigorific mix- 
ture, at the temperature of +io°, for a much longer time 
than 
