[ 239 ] 
Thefe determinations of the effect of heat upon 
the column of quickfilver, in the Torricellian tube, 
(hew the proportional alterations of the denfity of 
that fluid, by given increments, or diminutions, of 
heat. For the perpendicular height of a column, 
of any inelaftic fluid, fufiained in the Torricel- 
lian tube, by a given compreflive force, mud, by- 
the known laws of hydroflatics, be as the den- 
lities of the fluid inverfely. And as this pro- 
portion muft obtain, whatever be the flze, or 
figure, of the tube, there feems to be no me- 
thod, by which the change of denfity, or the pro- 
portional expanfion of quickfilver by heat, can be 
meafured with more precifion. Thefe conclufions, 
therefore, may be of ufe in many phyfical en- 
quiries ; and there are many cafes, in which it 
may be neceflary, to reduce the obferved height 
of the barometer, in one temperature, to an- 
other. Thus when that height is to be made 
the meafure of the variable preffure, or of the den- 
fity of the air, in fome particular place, it will be 
neceflary to choofe fome Jlandard temperature, to 
which the obferved length of the column may 
always be reduced. And it was this confideration, as 
I have gathered from many converfations with him, 
which gave occafion to M. de luc’s miftake. 
He had fettled it with himfelf, at his very fir A en- 
trance upon thefe refearches, that the point he was 
to keep conAantly in view, as the ultimate objedt 
of his whole purfuit, was to find, in the variable 
length of the mercurial column, a meafure of 
the preflure and denfity of the air. This, he faw„ 
was only to be looked for in quickfilver of fome 
5 one 
