Experiments on freezing Mixtures . 253 
12. The freezing point differs remarkably, according to the 
ftrength of the acid. In the diluted dephlogifticated and com- 
mon fpirit of Art. 7. and 8. the freezing point was — 5 0 and 
— i°f. In the dephlogifticated and common fpirit of 
Art. 5. the decanted parts of which were ftronger than the 
foregoing in fcarcely fo great a proportion as that of four to 
three, it fee med to be - 30° and — 31°! , It may indeed be fuf- 
pedfted, that as this point was determined only by pouring a 
fmall quantity of the acid into a glafs, at a time when the air 
and glafs were much colder than the acid, thefe decanted 5 
liquors might be cooled by the air and glafs, and thereby make 
the freezing point appear lower than it really was : but I do 
not think this could be the cafe ; for as the decanted liquors 
were full of fmall filaments of ice, they could hardly be cooled 
fenfibly below their freezing points without freezing ; ana any 
cold, communicated to them by the air or glafs, would ferve 
only to convert more of them into ice, without fenfibly in- 
creafing their cold : fo that I think this experiment determines 
the true freezing point of their decanted part ; but it rnuft be 
obferved, that as the decanted part was rather ftronger than 
the reft, it is very poflible that the freezing point of the unde- 
canted part might be confiderably lefs cold. 
A circumftance which might incline one to think, that the 
way by which the freezing point was determined in this expe- 
riment is defective is, that the freezing point of the dephlo- 
gifticated acid N° 27. though nearly of the fame ftrength as 
that laft mentioned, but rather ftronger, was much lefs low, 
being only - 19 0 . But I have little doubt that the true reafon 
of this is, that in the former acid the ftrength of the decanted 
part, which is the part whofe freezing point was tried, was found 
to be at leaft ^ greater than that of the whole mafs ; whereas 
in 
