meafuring Heights with the Barometer. 731 
quickfilver in the tube and attached thermometer did not 
keep exactly pace with each other. 
The laid point to be mentioned is {till more remarka- 
ble tha the reft ; it is briefly this : in the obfervations at 
Sun-ri ing on Saleve, though the fuperior quickfilver is 
the cold eft; yet the top of the column of air is commonly 
five or fix, arid fometimes eight or nine degrees, warmer 
than the bottom. 
Having thus fhewn the fteps that were taken, for ob- 
taining the coldeft and hotteft barometrical obfervations 
that the climate of this ifland would afford, the mode of 
obferving, and wherein the circumftances attending them 
differed from thofe on Saleve, I fliall now point out the 
general refult. In order to avoid repetitions as much 
as poffible, it is neceffary, once for all, to remark, that 
the computations of the Britifh obfervations, by the rule 
hereafter to be given, are fubdivided into their refpefilive 
claffes. Each table contains 1 5 columns, which their" 
titles fufficiently explain, that the principles from which 
the rule was deduced, the refult and error, might all ap- 
pear in one view. The laft column towards the right- 
hand fhews the ratio of the weight of quickfilver to air, 
the columns of the firft in the barometers being feve— 
rally reduced to the mean temperature of the laft. 
2 
By 
