Appendix J. ON THE BOILING-POINT THERMOMETER. 457 



mean of the temperature at the upper station and at Calcutta, to be 

 75°- 2, and as deduced from the formula to be 73°1. Here, however, 

 the equivalent in feet for 1° temp, is in summer very high, being 

 1°= 385 feet. (See Appendix I.) The mean of all the elevations 

 worked by the boiling-point is upwards of 140 feet below those 

 worked by the barometer. 



The following observations are selected as having at the time been 

 considered trustworthy, owing to the care with which they were 

 taken, their repetition in several cases, and the presumed accuracy 

 of the barometrical or trigonometrical elevation with which they are 

 compared. A small correction for the humidity of the air might 

 have been introduced with advantage, but as in most barometrical 

 observations, the calculations proceed on the assumption that the 

 column of air is in a mean state of saturation ; as the climate of the 

 upper station was always very moist, and as most of the observations 

 were taken during the rains, this correction would be always additive, 

 and would never exceed sixty feet. 



It must be borne in mind that the comparative results given 

 below afford by no means a fair idea of the accuracy to be obtained 

 by the boiling-point. Some of the differences in elevation are 

 probably due to the barometer. In other cases I may have read 

 off the scale wrong, for however simple it seems to read off an instru- 

 ment, those practically acquainted with their use know well how 

 some errors almost become chronic, how with a certain familiar 

 instrument the chance of error is very great at one particular 

 part of the scale, and how confusing it is to read off through steam 

 alternately from several instruments whose scales are of different 

 dimensions, are differently divided, and differently lettered ; such 

 causes of error are constitutional in individual observers. Again, 

 these observations are selected without any reference to other con- 

 siderations but what I have stated above ; the worst have been put in 

 with the best. Had I been dependent on the boiling-point for 

 determining my elevations, I shoidd have observed it oftener, or at 

 stated periods whenever in camp, worked the greater elevations from 

 the intermediate ones, as well as from Calcutta, and resorted to every 

 system of interpolation. Even the following observations would be 

 amended considerably were I to have deduced the elevation by 



