30 Mr. J. Burgess on the Measurement of Altitudes 



of barometrical pressure is 



A'-A = Lxloggj, 



where B and B' are the heights of the barometer at the lower 

 and upper stations respectively, h' — h the difference of elevation 

 uncorrected for temperature, and L a constant. Now, if Pro- 

 fessor Forbes' s hypothesis were true, we should have 



T-T'=»log|; (1) 



T and T' being the boiling-temperatures at the lower and upper 

 stations ; or under the pressures B and B ; respectively. Hence 

 we obtain as the expression for the approximate difference of 

 elevation 



h'-h=±(T-V); (2) 



and the hypothesis is correct or otherwise, according as the quan- 

 tity n is constant or variable. 



3. In the following Table are collected a number of observa- 

 tions made by M. Izarn among the Pyrenees*, Dr. Forbes in the 

 Alpsf, Dr. Joseph Hooker on the Himalaya and Khasia 

 Mountains J, MM. Martins and Bravais on Mont Blanc §, 

 M. Marie on Mont Pila||, &c. In column (4) are tabulated 

 the values of n derived from each observation, in (5) the boil- 

 ing-points corresponding to the observed pressures in column 

 (2) computed with w=112, and in (6) the differences between 

 the observed values in (3) and the computed ones in (5). 



After making allowance for errors of observation, it is mani- 

 fest from this Table that the value of n slowly decreases with the 

 pressure. But so slow is the rate of decrease that Professor 

 Forbes' s hypothesis affords good approximate results if, for heights 

 under 10,000 feet, we employ n= 1 13*33 ; and, L being =60369 

 feet, by substitution in equation (2), we have 



#-^=532-7(T-T') (3) 



* Comptes Rendus de VAcademie, vol. xix. p. 169. 

 f Edinb. Phil. Trans, vol. xv. ; and vol. xxi. part 2. p. 237. 

 % Hooker's Himalayan Journals, vol. ii.; and Edinb. Phil. Trans, 

 vol. xxi. part 2. pp. 239, 240. 



§ Comptes Rendus, vol. xix. p. 166. || Ibid. vol. xviii. p. 252. 



