292 



PROCEEDINGS OF THE AMERICAN ACADEMY 



TABLE XIII. 



The fact noticed by Argelander, that the refraction derived from the 

 setting sun is less than that of a star is well shown in this table. The 

 difference amounts to one or two minutes of arc. 



In this investigation the value of one division of the screw in seconds 

 must be known with accuracy. It was therefore redetermined August 

 6, 1885, with the same result, 13".95, as that originally found. 



This paper is intended to show that the micrometer level is capa- 

 ble of giving useful results where a larger instrument has generally 

 been considered necessary. Its portability, and the rapidity with 

 which observations may be made by it, adapt it especially to the wants 

 of travellers, and would permit the accumulation of valuable informa- 

 tion regarding the atmospheric conditions of comparatively inaccessi- 

 ble points. If required, much greater accuracy could doubtless be 

 attained than is indicated by the stellar observations described above. 

 The instrument was mounted on a wooden balcony, and the times were 

 only taken to whole seconds. Instead of moving the telescope each 



