92 Mr, Nixon on the Theory of the [Aug. 



Faussiffny with a view to determine the vertical difference of 

 level of the stations proved quite unsuccessful. The probable 

 reason is, that the stations differed materially in latitude, so 

 much 80, and in such a manner, that tlie correction for the latter, 

 had it been attended to, would have exceeded the other, and in 

 an opposite direction.* 



Correction for the Attraction of Mountains, 



When we consider that the surface of the earth (restricting 

 the term to that part of it coinciding in level with the sea), is 

 generally covered, especially in the scene of barometrical obser- 

 vations, with ponderous mountains, and that the correction for 

 the diminished gravity of the mercury, although reduced one- 

 half, is subtractive, and the one for the air (still additive) merely 

 hair the amount for depths ^^or differences of level), below the sur- 

 face, we must be sensible that barometrical measurements made 

 in the midst of a mountainous country must tend to err in 

 excess. 



From the attraction of the steep acclivities forming a narrow 

 valley, &c. the particles of air therein are more numerous than 

 in a similar volume of air under the same pressure, &c. taken 

 from a situation unaffected by local attraction. Now as this 

 accession of density does not sensibly extend to the strata of 

 the atmosphere incumbent on those within the valley, the den- 

 sity of the latter is too great for the pressure, and the same 

 depth of air requires a greater column of mercury to form a 

 counterpoise. The error in excess in small differences of level 

 measured in similar situations is incredibly great, as I have had 

 frequent opportunities of verifyingby a comparison of the levelled 

 ^nd barometrical altitudes. 



When the lower station is an extensive plain, and the upper 

 barometer is placed at the bottom of a deep ravine, or gorge, the 

 increase of the density of the air therein causes the mercury to 

 stand too high, and the computed altitude must fall short of the 

 truth. 



Correction for the Difference of Temperature of the Mercury of 

 the Barometers. 



Mercury expands with increase of temperature : according to 

 the approved experiments of MM. Dulong and Petit^ its volume 



at 32^; F. is altered qooT) for a variation of temperiBitiii*^ equal to 

 |o p^ The densities of equal weights of fluids being reciprocally 

 iia,t|ie Volumes, and their pressures equal to theii*'^ eights mul- 



• At the level of the sea, water boils at 2 1 2° under a pressure of 30 incbefi of mer- 

 cury, but would not the ebullition at the same temperature take place AtibQ devatkn 



«f four miles, although the barometer stood there at .'JO inches + rxr- ?^ 



