670 Col. roy’s Experiments for 
experiments gave .5258, for the total dilatation of 30 
inches of quickfilver, on 1 8 o° between freezing and boil- 
ing; that, anfwering to the firft 20% between 32 0 and 
5 2 0 , was .0688; that, for the 20° in the middle of the 
fcale, between 1 1 2 0 and 1 3 2 0 , was .058; and the rate for 
the laft 20°, between 192 0 and 212°, was only .041. 
From this firft fet of the third clafs of experiments, it ap- 
-peared evident, that the expanfion of 30 inches of quick- 
iilver in the barometer, fuffering a heat equal to 1 8o° of 
Fahrenheit, inftead of exceeding Mr. de luc’s, as ap- 
peared to be the cafe from the refults of the open tube, 
really fell ftiort of it : and inftead of being arithmetical or 
uniformly the fame, for equal changes of temperature, was 
aftually progreftive ; the expanfion anfwering to the lower 
part of the fcale, being greater than that correfponding to 
the middle; which again exceeded that for high tempe- 
ratures. In thefe experiments, when the water had ac- 
quired aheat 2 o or 3 o degrees greater than that of the open 
air, a certain duftinefs was perceived in the vacuum of 
the tube. At ioo° of Fahrenheit, or thereabout, this 
appearance had fo far increafed, as to fhew clearly, that 
it could proceed from no other caufe than a vapour ari- 
fing from the furface of the heated quickfilver, quite in- 
vifible, till, by its condenfation in the cold part of the 
tube, it was formed into balls, every where adhering to 
its 
