524 george shuckburgh’s Obfervations 
Having thus the perpendicular height, as I think, 
very accurately afcertained, it remained for me to take 
the altitude of the barometer at each ftation a and c, and 
if poffible with equal precilion. Thefe obfervations it 
would be too tedious to fet down at length. I lhall, how- 
ever, premife the following particulars. Every obferva- 
tion of the barometer was triple ; that is, the height was 
read off three different times, and the mean taken ; but 
from once, reading only I could. be fure of the height to. 
of an inch, excluftve of the error of the divifions* 
which in fome places might amount to that quantity;, 
this the nonius would itfelf difcover and even correct by 
eftimation. At every feries of obfervations the float at the? 
bottom was readjufted, fo that I could conftantly be fure. 
of an alteration of the weight of the. atmofphere ex-- 
prefled by 0.00a inch of quickfilver, if not of half that 
quantity. Finally, the difference of the two barome- 
ters (>1 was conftantly taken, after being left three-quar- 
ters 
( 1 ) It may be concluded, that-this difference fhould be. conHant, and always the 
fame ; but, from what caufe I know not, it did not appear fo to me. In my journal 
Jcor the. weather for 1.775, ^ find the following note: from a mean of feventeen 
obfervations between AuguH 12th and Sept, before, at, and after, my> 
expedition to Mont Saleve and the Mole, I find the difference between my two 
barometers —,0042 inch, N° 1. Handing the higheft; in thefe comparifons, 
however, the extremes fometimes differed from the mean ~,oo6. And in my 
paffage over Mont C'enis* Dec. barometer N? i. Hood lower than N° 2. 
