165 



coefficient of Ramond, there is only the second v^hich does 

 not give a greater height than the geometrical, operations- 

 If Deluq's or Tremblay's formula be substituted for that of 

 Laplace, the heights, instead of being too great, will be too 

 little. Supposing the Peak to be really 1905 toises high, 

 Laplace's formula, applied to Messrs. Lamanoin and Gordier's 

 observations, would be erroneous only 5| toises, or a three 

 hundred and forty sixth ; an extremely sm all quantity, and 

 which would be the half or the third only of that, to which 

 excellent observers may be often exposed 



The first coefficient f of the barometric formula of Mr. de 

 Laplace, published in 1798, was founded on the comparison 

 of the barometrical and geometrical meas urement of the vol- 

 cano of TenerifFe, made by Mr. de Borda . The illustrious au- 

 thor of La M&anique celeste having after tvard found, that this 

 coefficient did not give exact heights, substituted another, 

 furnished by the excellent observations of Mr. Ramond. On 

 examining the manuscript narrative of Bc»rda*s voyage, we can- 

 not guess at the source of an error, which seems considerably 

 to surpass that of the barometric measurement of Mount 

 Blanc by Saussure. The correspondent barometer was ob- 



* Mr. d'Aubuisson concluded, aft,er having discussed a 

 great number of observations calculated after the formula of 

 Laplace, and compared with exact, geodesical measures, 

 " that in avoiding the manifest cauejes of inexactness, such 

 as the morning hours, the considerable changes of weather 

 from one day to another, storms, a nd the influence of local- 

 ities, we may consider a hundredtbi as the limit of the mis- 

 takes." He adds, that " most commonly, by fortunate 

 compensations, the error will be only some thousandths." 

 Journal de Physique, t. lxxi, p. 3- f >. 



+ The coefficient, 17972 metres. Exposition du Systeme 

 du Monde, ed. 1, p. 82, Ramond, Mem. sur la Formule 

 barometrique, p. 2. 



