422 Count Paul de Saint-Robert on Atmospheric Refraction. 



M 



The coefficient — of terrestrial refraction is supposed usually 



in geodesy to have always the same numerical value at all places 

 and at all times. This is obviously inexact*. I should think it 

 preferable to determine it by the observed decrease of density in 

 each particular case. Having, by direct meteorological observa- 

 tions, the densities at two stations at different levels, we can cal- 

 culate a by the formula above-mentioned, 



2 Po \Po-P J 

 and then we shall have for the coefficient of refraction 



M _ uar 



It is not to be forgotten that the densities p and p should be 

 calculated as if the air were dry. 



I shall close this long article by recapitulating the formulae to 

 be used for the measurement of heights, and applying them to 

 a numerical example. 



Barometrical Formula. 



3 



8 



P— 77 



ct 



c =(29'172)G in metrical measure, 

 c = (53 # 173)G in English measure, 



x= r (r-r ) > 

 r 



p , p atmospheric pressures at the lower and upper stations 

 respectively. 



* It has been found that the amount of terrestrial refraction undergoes 

 an hourly variation. A series of observations for determining the hourly 

 variations was carried out by M. Hossard, at Angouleme, during the 

 months of May and June 1844. The observations were made half-hourly, 

 from daybreak till twilight ; and the object observed was at a distance of 

 about thirteen miles. The conclusions to which they led are the following: — 

 The refraction is greatest about daybreak ; from 5 or 6 a.m. until 8 a.m. it 

 diminishes very rapidly; from 8 a.m. until 10 a.m. the diminution is slow; 

 from 10 a.m. to 4 p.m. the refraction is nearly constant; from 4 p.m. the 

 refraction commences to increase. 



