532 Sir george shuckburgh's Obfervdtiotu 
Thefe obfervations then feem to prove that the baro- 
metrical rules were a little defective as to the true ratio be- 
tween the gravities of air and quickfilver, •viz. in the 
value of an inch of quickfilver in the torricellian tube, 
exprefled in inches of the atmofphere with a given tem- 
perature. The firft comparifon gives for this error in 
Refect -19.8 feet in every 1000 feet; the fecond, 24.0 
feet; and the laft, 25.4 feet: the mean of the three is 
23.1 feet; and by fo much we may conclude that thefe 
rules, in greater heights alfo, will give the difference of 
elevation too little, viz. by ~ nearly (m> . But it will be 
fair to make the experiment. 
(m) Left any fufpicion fhould arife of a difagreement between the a£lual 
meafures taken by Mr. be luc and myfelf, I may obferve, that the mean refult 
of three obfervations, which I made independently of each other on the height of 
the Pitton or point c above the lake of Geneva, agree with the mean refult of 
Mr. de luc’s operation from the levelling and the quadrant, to lefs than twelve 
inches ; a greater correfpondency than which cannot be expelled : and this was 
the true reafon why I chofe the fame fpot he had pitched upon. <c Le rocher 
* c ifole , qui domine toute la montagne As a further confirmation, I compared his 
ftandard fteel rod of twelve Paris inches, which his brother obligingly furnifhed 
me with, with my brafs one, and found twelve inches on Mr. de luc’s rule 
was on my rule, with 71 0 of heat, — — 12.784 Eng. inches. 
Correflion for the difference of expanfion between! , 
brafs and fteel with 16 3 of heat, — — J + °7 
Length of Mr. de luc’s French foot with 55°, — 12.7847 
True length of the French foot ( vide Phil. Tranf.) 12.7890 
Error or difference from the true Paris foot — — ,0043 =: jdxrs nearly. 
I 
The 
