[ 168 ] 
confequently the diminution of the volume of mer- 
cury muft be very great indeed. This is demonftrated 
by the great defcent of the mercury in the thermo- 
meter, while it is freezing. But how great this di- 
minution of the volume of the mercury is, cannot 
exa&ly be determined ; and hence arifes no fmall dif- 
ficulty in determining its fpecific gravity, as this laft 
muft increafe, as the bulk of the mercury leftens. 
Hence as mercury, even in its fluid ftate, comes of 
all bodies, platina excepted, the neareft to gold j in 
its folid ftate, it muft ftill approach much nearer. 
Our author had three thermometers filled with the 
moft highly rectified fpirit of wine. Thefe not only 
correfponded exadtly with one another, but, in lefs 
fevere trials, correfponded reafonably well with thofe 
filled with mercury. But by the mixture of fnow 
and fpirit of nitre, which froze the mercury, he never 
was able to bring the fpirit thermometers lower than 
300. From hence it appears, that the heat, which 
will freeze mercury, will not freeze fpirit of wine ; 
and that therefore fpirit thermometers are the moft 
fit to determine the degree of coldnefs in frigorific 
mixtures, until we are in a fituation to conftrudt folid 
metallic thermometers with fufflcient accuracy. 
Our author made many experiments, to try the 
effedts of different fluids, in his frigorific mixtures. 
He invariably found, that Glauber’s fpirit of nitre 
and double aquafortis were the moft powerful. With 
oil of vitriol, the moft ponderous of all acids, he 
was never able to congeal mercury. He likewife 
tried a great number of other fluids, both acid and 
fpirituous, which though, when mixed with fnow, 
produced cold, it was in very different degrees. He 
tried 
