894- 



PROCEEDINGS OF THE AMEEICAN ACADEMY 



observed at the Harvard College Observatory by tlie same instrument 

 which was used in determining the light of the comparison stars. The 

 image of a Ursae Minoris was rendered equal to the variable and to 

 DM. 81°30 alternately. Seven settings were made in each set, begin- 

 ning and ending with the variable. Systematic errors were thus greatly 

 reduced, since those only would enter which affected one star and not 

 the other. Although a large number of readings were taken, the 

 results were not satisfactory and probably have but little value. The 

 result for September 23 was 8"^ 57™, Cambridge mean time, and for 

 September 28, 8*^ 7™, the difference between the two being five days 

 less fifty minutes instead of five days less twenty minutes. On neither 

 evening were observations obtained before the minimum, owing to twi- 

 light. 



TABLE XIV. — Comparison of Observed Minima. 



