New Tables for Finding Heights hy the Barometer. 71 



Method of Measuring Heights b}'- means of the Barometer," 

 by G. K. Gilbert, the following table is given, from the 

 investigation of actual measurements. The assumptions are 

 that one station is 5000 feet higher than the other, that 

 the two places are fifty miles apart, and that they are 

 situated in the temperate zone, remote from the ocean. The 

 observations are supposed to be made near the middle of a 

 fine summer's day, with a light wind blowing : — 



It will be noted how lai-gely the temperature error exceeds 

 all the others ; and from the second column it will be seen 

 that, excluding the error of observation (which, owing to 

 mistakes, may be unlimited) the total error may amount 

 to 420 feet, or more than one-twelfth of the whole amount 

 measured. The probable error of 10 feet assigned for im- 

 perfection of observation is meant for a mercurial barometer ; 

 for an aneroid it would be much larger, even when used 

 with the greatest care, and frequently compared with the 

 mercurial column for finding its errors. At our observatory, 

 we have an apparatus for testing mountain and other aneroids. 

 I append the results of some of the late comparisons : — No. 

 1 has an error of "09 in a range of 26 to 80 inches. No. 2, 

 •30 in a range from 23 to 30 inches. No. 3, "13, range 24 to 

 30 ; and •] 8, range 23 to 30. No. 4, •] 3, range 24 to 30 ; 

 •23, range 23 to 30. No. 5 (which is a French barometer 

 metrically divided), -40, range 25 to 30 ; and -68, 24 to 80. 

 In the last three instruments it will be remarked, how the 

 error increases as the extremity of its scale is reached. I 

 have on several occasions, when travelling on our railways, 

 carried an aneroid with me, and recorded its readings, as 

 well as the temperature of the air at stations where we have 

 stopped, and have compared the heights thus deduced with 

 the accurate ones of the railway levels. At first I treated 

 the results as matters of curiosity only, and did not preserve 



