750 



SCIENCE 



[N.S. Vol. XXV. No. 645 



Motion.' The greater part of the observations 

 were made by Dr. F. L. Chase, assistant 

 astronomer, and a smaller number by Mr. 

 M. F. Smith, assistant, and Dr. W. L. Elkin, 

 the director. 



It was thought at first that a small number 

 of observations at each of two successive 

 epochs of maximum parallactic effect would 

 certainly show if the parallax amounted to as 

 much as 0".20, and give some indication of 

 a value as small as 0".10. Later the plan 

 was extended to include two more epochs 

 taken in reverse order. The results were 

 made to depend wholly on measures of dis- 

 tance, and in general two comparison stars 

 were selected on opposite sides of the star 

 whose parallax was sought. 



Systematic personal error, due to the direc- 

 tion of the stars, or to differences in color or 

 brightness, was avoided by the use of a revers- 

 ing prism eye-piece. Also, gauze screens were 

 used to equalize the brightness of the stars to 

 within half a magnitude. For each star in 

 most cases twelve complete observations were 

 made. Every precaution to eliminate known 

 sources of error was employed, so that the 

 authors have good reason to believe the re- 

 sults to be free from systematic error, except, 

 perhaps, one due to color. Such an error 

 seems to be theoretically possible. In order to 

 ascertain whether an error due to this cause 

 is appreciable in actual observations, Dr. 

 Chase made a series of observations on five 

 highly colored stars. The results appear to 

 indicate, that there is a discernible color 

 efiect, which is in accordance with theory; 

 but the errors involved are so small that even 

 in extreme cases, they come within the prob- 

 able errors, never amounting to more than 

 0".03. 



All the observations were given equal 

 weight. The effect of poor definition seems 

 to have been inappreciable. 



A reliable estimate of the systematic errors 

 of the stars in general is obtainable by a com- 

 parison of the Yale parallaxes with those 

 deduced by other able observers. A table of 

 twelve such stars is given from which the 

 average difference between the Tale values 

 and the others amounts to 0".036, fron which. 



assuming equal accuracy for both results, the 

 total probable error of each is =t 0".017. 



The authors do not claim great precision 

 for the individual results, but attach im- 

 portance to the mean values of various groups 

 which they formed. With the addition of the 

 ten stars of the first magnitude previously de- 

 tennined by Elkin, the number of stars con- 

 sidered is 173, from which five groups were 

 formed. In each group the stars are arranged 

 as indicated in the following mean results. 



TABLE I 



Results arranged according to Proper Motion 



TABLE n 



Results arranged according to Stellar Magnitude 



TABLE nl 



Results arranged according to Size of Parallaa 



TABLE IV 



Results arranged in order of Right Ascension 



