NOTES. 



463 



equal to | the transmitted light ; at 26 38' it is equal to J, the variation, 

 according to Arago, being as the square of the cosine. 



NOTE 189, p. 154. Atmospheric refraction. Let a 6, a 6, &c., fig. 49, 

 be strata, or extremely thin layers, of the atmosphere, which increase in 



Fig. 49. 



density towards m n, the surface of the earth. A ray coming from a star 

 meeting the surface of the atmosphere at S would be refracted at the sur- 

 face of each layer, and would consequently move in the curved line S v v v A ; 

 and as an object is seen in the direction of the ray that meets the eye, the 

 star," which really is in the direction A S, would seem to a person at A to 

 be in s. So that refraction, which always acts in a vertical direction, raises 

 objects above their true place. For that reason, a body at S', below the 

 horizon H A 0, would be raised, and would be seen in s'. The sun is fre- 

 quently visible by refraction after he is set, or before he is risen. There 

 is no refraction in the zenith at Z. .It increases all the way to the horizon, 

 where it is greatest, the variation being proportional to the tangent of the 

 angles Z A S, Z A S r , the distances of the bodies S S' from the zenith. 

 The more obliquely the rays fall, the greater the refraction. 



NOTE 190, p. 154. Bradley's method of ascertaining the amount of 

 refraction. Let Z, fig. 50, 

 be the zenith or point im- 

 mediately above an observer 

 at A; let HO be his hori- 

 zon, and P the pole of the 

 equinoctial A Q. Hence PAQ 

 is a right angle. A star as 

 near to the pole as s would 

 appear to revolve about it, in 

 consequence of the rotation of 

 the earth. At noon, for ex- 

 ample, it would be at s above 

 the pole, and at midnight it 

 would be in s' below it. The 

 sum of the true zenith dis- 

 tances, Z A s, Z A s', is equal 



Fig. 50. 



