364 Mr. Nixoii on the Accuracy of the [May, 



lial to complete success. The observations include, on the con- 

 trary, not only the extremes of pressure, but also most rapid rates 

 of its variation, and comprehend temperatures ranging, in calms 

 and storms, from 26° to 78° F. Add to which the operations at 

 the reference stations commenced in general much too early in 

 the morning, and were frequently protracted, especially in 

 winter, to a late hour after sunset. 



The following methods were resorted to, in order to ascertain 

 the rate of the fall or rise per hour of the barometer, whence to 

 obtain by interpolation the pressure at the base at the instant of 

 the observation on the summit of the mountain. Firstly, by 

 noting the height of the barometer at the lower station before 

 as well as after the ascent ; secondly, by repeating the observa- 

 tion of the pressure after a sufficient lapse of time either at the 

 base, or in some eligible situation in its vicinity ; and, lastly, by 

 obtaining the pressure at some other low station of which the 

 altitude, as well as that of the one visited before the ascent, had 

 been determined trigonometrically. In one or two instances, the 

 pressure at the summit, when the stay there was considerably 

 protracted, was taken twice ; not with the erroneous idea of 

 ascertaining the rate of the variation, but in order to furnish two 

 independent measurements, of which the mean would probably 

 be found more accurate than either separately. 



The calculations were made exclusively from the set of tables 

 given in the present volume of the Amials. Twice the observed 

 temperature of the air at the upper station being considered as 

 the sum of the detached thermometers, the deduced altitude was 

 finally augmented, as pointed out in Table VI. 



With regard to the trigonometrical operations, as I shall pro- 

 bably have occasion to recur to the subject, I restrict myself for 

 the present to the following brief particulars. The horizontal 

 distances were, with few exceptions, derived from the accurate 

 bases given in the third volume of the Ordnance survey; the 

 requisite angles being measured on a good theodolite reading off 

 to 30". The zenith distances were taken by a most excellent 

 instrument, expressly designed for the measurement of small 

 vertical arcs to the accuracy of 2" or 3". On a comparison of the 

 results derived from different bases, &c. it would appear that the 

 error in altitude could never amount to ten feet, and would most 

 probably rarely exceed as many inches. 



As the table does not comprise heights surpassing 1700 feet, 

 we may inquire whether the errors, in the case of the measure- 

 ment by the same method of still loftier mountains, will continue 

 to increase in proportion to the elevation ? In the first place, 

 those inaccuracies arising from the impossibility of ascertaining 

 the exact temperature of the mercurial column, which give the 

 appearance of such m.arked irregularity in the barometrical 

 measurement of small differences of level will be nearly constant 



