1818.] the Imperfection of the Barometer. 207 



would be as ascertained by other means, the agreement between 

 the heights, as indicated by the two methods, was found so 

 nearly to coincide as to lead M. Ramond to conclude that the 

 barometer might be safely employed as an instrument for mea- 

 suring heights : 



Names of the Stations. 



Difference. 



The barracks 

 Crete de Pradelle 

 (Without a name) . 

 Shepherd's Bridge 



Chevasson , 



Col des gourdes. . 



Metres. 

 + 0-06 or 

 -0-54 — 

 + 0-76 — 

 -0-45 — 

 -0-04 — 

 -0-09 — 



Inches. 



2-34 

 21.6 

 30-4 

 18-0 



1-6 



3-6 



An account is also given of some contemporary observations 

 that were made at Geneva and at Vincy, a place 18 miles 

 northward of this city, and so much elevated above it that the 

 mean height of the barometer at Geneva was 5*332 lines above 

 that at Vincy. The observations were made contemporarily at 

 the Observatory in Geneva and at Vincy, at sun-rise, sun-set, and 

 at two P.M. ; they were 66 in number; and, compared together, 

 gave the following results : 



14 agreed with the mean difference. 

 9 differed by -^ of a line or -005 inch nearly. 



10 tV -oi 



12 fo -0166 



8 -rV -02 



J> -. * -04 



66 



We learn from this table that, when we observe at the same 

 time two barometers at 18 miles of horizontal, and 444 feet of 

 vertical distance from each other, and without any obstructing 

 body between them, it is about one to three that the difference 

 of the heights of the mercury will be precisely what corresponds 

 to their actual elevation ; about one to two that the heights will 

 not differ more than ^ of a line, or even less ; and that it is 

 about an equal chance that the difference will not be more than 

 -tV of a line. 



It may appear a little remarkable that the barometers that are 

 employed in France and in Geneva should be more perfect than 

 those in England. But we may conceive that the instruments 

 in the above cases were made for the express purpose of the 

 comparative experiments, and would be constructed with pecu- 

 liar accuracy. 



