72 TRAVELS AMONGST THE ORE AT ANDES, chap. hi. 



feet/ or 817 feet loioer than the determination of Ilnmboldt. 

 Although I had no doubt of the accuracy of my readings/ I 

 thought it was desirable to repeat them, if possible at 11 a.m., 

 for the sake of combination with Mr. Chambers' observations at 

 that hour. 



Two aneroids (D and E) were taken to the summit, and when 

 read against the mercurial they shewed a considerable increase 

 in their errors upon those which were observed at the third 

 camp — 



Error of D at last comparison at the third camp was — 0-724 inch. 

 Do. E do. do. -0.774 ,, 



Do. D upon the summit of Chimborazo . . — 1-060 ,, 



Do. E do. do . . - 1-210 „ 



The actual readings on the summit of these two aneroids were 

 (D) 13-050 inches, and (E) 12-900 inches.' The mean of the 

 readings (12-975 inches), thus, was no less than 1*135 inches 

 lo2uer than the reading of the mercurial barometer (corrected 

 for temperature) ; and if the altitude of Chimborazo had been 

 deduced from this mean, in combination with Mr. Chambers' 

 observations at Guayaquil, the height of the mountain would 

 have come out more than a thousand feet greater than the 

 determination of Humboldt ! 



I considered that it was desirable to ascend Chimborazo 

 again, to see whether we could improve our route, to learn 

 whether our deplorable rate at the upper part was due to the 

 softness of the snow or was to be attributed to diminution in 

 atmospheric pressure ; and to remain a longer time on the summit 

 to repeat the observations of the barometer, and to obtain a 



1 Upon being re-computed by Mr. Ellis (after the Guayaquil observations were 

 known), this was reduced to 20,545 feet. 



2 The mercurial barometer was set up directly we reached the summit, but the 

 reading was not entered until it was found that the mercury of the instrument 

 had taken up the temperature of the air. 



3 These aneroids were graduated to 13 inches. 



