[ 32 ]■ 



Don Ulloa tells us, p. 34 of the above-mentioned work, and 

 Gentil repeats after him, that, at the village of Guancavelica, near, 

 but fomewhat below this fpot, the mercury flood at the height 

 of 18 inches, one line and ^. Now Bouguer, Figure de la Terre, 

 p. xxxvi, exprefsly lays that Qjiito is elevated from 14 to 1500 

 toifes only over the level of the fea, and is the higheft inhabited 

 part of the globe; Guancavelica, therefore, which is inhabited, and 

 muft have Deen well known to Bouguer, cannot be fo high as 

 Quito. Quito was geometrically meafured, Guancavelica was 

 not. 



Bouguer alfo tells us, Hid. in note, that the mercury at Quito 

 flands at the height of twenty inches and one line ; how then is 

 it poffible that it fliould ftand at eighteen inches and one line at 

 Guancavelica which muft be much lower, if the barometer were 

 not ill conftruded ? No difference of temperature between Quito 

 and Guancavelica could caufe fucH an enormous difference as two 

 inches in the mercurial height. The art and neceffity of freeing 

 barometers fromair were not generally known * before Mr. De Luc's 

 immortal work on the modifications of the atmofphere, which was 



not 



* I fay not generally known, becaufe though Mr. Du Fay had fhewn the advan- 

 tage of expelling air from the mercury in the tube, in Mem. Par. 1723, and 

 CafTinl had followed that method in meafuring the heights of Puy de Dome and' 

 Mont D'Or in 1740, yet it was not generally adopted until Mr. De Luc had proved, 

 its neceffity and perfedlcd the method of performing it in 1772. See i De Luc,^ 



