1858.] On Ilypsometrical Measurements. 345 



as nearly as possible with the experiments of Ure, Southern and 

 Dalton.* From these results, he calculated a table of heights cor- 

 responding to the boiling point for each degree of temperature from 

 17G° to 211°. From observations made simultaneously with the 

 barometer and boiling point thermometer, Colonel Sykes inferred 

 that Prinsep's table gave altitudes generally less than those indi- 

 cated by his barometer. Sykes made no observations at greater 

 altitudes than 4500 feet, but at that height, the average error in 

 Prinsep's table, seems to be over 100 feet. 



This subject has since been discussed in a paper by Professor 

 J. D. Porbes, published in the transactions of the Royal Society of 

 Edinburgh for 1842-3, founded on observations made among the 

 Alps. By projecting the elevations as derived from barometrical 

 observations, but uncorrected for temperature, in terms of the observ- 

 ed boiling points, he found that a straight line passed almost quite 

 through the whole of the projected points. Hence he inferred 

 " that the temperature of the boiling point varies in a simple arith- 

 metical ratio ivith the height." 



2. This hypothesis seems to require examination. The general 

 form of the formula for heights as measured by barometrical observ- 

 ations is, — 



B 

 h = L x log — ; 

 B' 



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



lower stations respectively, h the difference of elevation, and L the 



constant determined in the preceding section. But were Professor 



Forbes's hypothesis true, we should have — 



B 



T — T' = n log — , (I) 



B' 



T and T' being the boiling temperatures at the two stations or 



under the pressures B and B' respectively. And combining these 



formula, we obtain — 



L 

 h = — (T — T') (2) 



n 

 as the expression for the approximate height. 



* Jour. Asiat. Soc. of Bengal, fcpril 1833, pp. 194 2()0. 



