3 i 2 PROCEEDINGS OF THE AMERICAN ACADEMY 



lent position of the observer, the line of siglit being always horizontal, 

 also conduces to the accuracy of the measures. Owing to the low 

 power used (about nine diameters), clockwork was not needed, the 

 stars being occasionally brought back to the centre of the field as they 

 moved away. The readings were wholly independent, as it was quite 

 impossible to distinguish the images of the two stars when brought 

 nearly to equality. 



The four positions of the Nicol, in which the images appeared to be 

 equal, were read by the graduated circle to tenths of a degree. This 

 "was repeated three times, so that twelve settings constituted a single 

 set. Successive sets were commonly taken by different observers, 

 but when one observer only was present he generally took care to 

 leave the instrument for a minute or so between the sets, so that the 

 same sources of error should not recur. Three observers took part in 

 the work, Mr. Arthur Searle, Mr. 0. C. "Wendell, and myself. They 

 will be designated by the initials S., "W., and P. 



Observations were made on thirteen nights, from Sept. 29, 1880, to 

 Jan. 1, 1881, and will be detailed in a future volume of the Annals of 

 the Observatory. The total number of sets was 230, containing 2,748 

 settings, two of the sets being incomplete. Twenty-eight sets were 

 taken when the nearest minimum was five hours or more distant. 

 They serve to determine the full brightness of the variable. Twelve 

 Bets by Mr. Searle give the excess in brightness of ft over w of 2.45 

 magnitudes ; five by Mr. Wendell give 2.68 ; and twelve by myself 

 give 2.67. As these results are confirmed by the other observations, 

 we may conclude that w appeared brighter or /3 fainter to Mr. Searle 

 than to the other observers by about 0.22 magnitudes. All of his 

 measures have been corrected by this amount to reduce all to the 

 same system. Fifty sets, or six hundred readings, were obtained on 

 Oct. 7, extending almost continuously from quarter of seven in the 

 evening to half-past three of the following morning. On Oct. 10, 

 nineteen sets were obtained from half-past six until nearly eleven, 

 when clouds prevented further measures. On Nov. 2, forty-two sets 

 were taken between six o'clock and midnight. On Nov. 1 9, fifteen 

 sets were obtained ; on Nov. 22, thirty-three ; on Dec. 9, fourteen ; 

 on Dec. 22, nine; and on Jan. 1, twenty-eight. On Nov. 19, the 

 observations of Mr. Searle appeared to differ from the results of 

 the other observers by about three tenths of a magnitude, even after 

 applying the correction of 0.22 magnitudes noted above, or without 

 this correction they differed about half a magnitude. The reduction 

 was first made retaining these, but they introduced so large a discord- 



