There is a negligible error in the use of the latitude coefficients 

 as directed. It will never exceed 15" exrept when the sun is near 

 the horizon. At that time, solar observations should be 

 for any refraction correction becomes itain. With a 



rising barometer and denser atmosphere or with a depressed ther- 

 mometer, at this time of day the correction may be increased to 

 double the amount given in the tables. 



It was probably Claudius Ptolemy in the first century who 

 Discovered the effects of atmospheric refraction. It is always posi- 

 tive for the purpose of this calculation, that is, it increases numer- 

 ically the North declination and diminishes numerically the South 

 declination. This rule is the reverse of the one given under Precise 

 Leveling, p.36; but the conditions are opposed. Having made the 

 observations in leveling we are arriving at the correct position of 

 the object sighted by subtracting the refraction. In this case, 

 however, we are determining what the observation should be, i. e., 

 how high the telescope should be elevated in order to see the sun at 

 a known angular elevation above or below the equator. 



In other words, in reducing apparent to true phenomena, re- 

 fraction is to be subtracted; hut in calculating apparent positions 

 from true altitudes, refraction correction is to be added. 



The tabular corrections for refraction are exact for the middle 

 day of ihe five-day period into which the Kphemeris Tables are di- 

 vided. For extreme days of any such period, an interpolation can 

 be made if desired. 



Latitude Coefficients for Refraction Correction 





