286 PROCEEDINGS OP THE AMERICAN ACADEMY 



Investigations ox Light and Ueat, made and published wholly or in part with 

 Appeopkiation from the Rumford Fund. 



XVI. 

 ATMOSPHERIC REFRACTION. 



By Edwaud C. Pickering. 



Commmunicated December 9. 1885. 



PART II. 



The observations described in Part I. relate exclusively to the re- 

 fraction of the portion of the air between two objects on the surface 

 of the earth. In astronomical observations we have to consider the 

 effect of the entire column of air traversed by the light from an object 

 outside the earth's atmosphere until it reaches the observer. The 

 variation of this quantity, and the effect of local causes upon it, is 

 an important source of error in many astronomical observations. Por 

 instance, the systematic differences in the declinations of the southern 

 stars, as determined at different observatories, may be due to different 

 refractions near the northern and southern horizons. The study of 

 this matter has usually been left to the large alt-azimuths and transit- 

 circles to be found in an astronomical observatory. From the fixed 

 position of these instruments it is not easy to vary the conditions as 

 much as might be desired. "We are therefore ignorant of the varia- 

 tions of the refraction in different azimuths, or the effect ujDon it of 

 the proximity of large masses of water, of forests, or of snow-covered 

 mountains. Even its variations in different parts of the world are 

 but little known, and it is usual to employ the refraction tables of 

 Bessel, or those of the Pulkova Observatory, under the most varied 

 conditions of climate or local surroundings. The micrometer level 

 seems to be especially adapted to measuring the atmospheric refrac- 

 tion, and it is hoped that the observations described below will show 

 that it is quite practicable for a traveller to determine this quantity 

 at any point where the results are likely to be of interest. Not the 

 least interesting of the results which may be thus obtained is the 

 determination of the law regulating the refraction at great elevations. 



