The Rutherfurd Photographic Measures. 157 



48. While there is some dependence of one result on another 

 in those instances where the same star of comparison was used, 

 3 T et,as the measures were made by independent observers and dis- 

 similar methods, and show in themselves no indication of a paral- 

 lax of the comparison-star, it is not improper to combine all these 

 individual results into means by weights proportional to the re- 

 ciprocal of the square of the probable error. Thus we have the 

 mean of previous determinations of the 



Parallax of 61 1 Cygni = oZ'417 + o // .o2i6 



and, in like manner excluding my own determinations : 



Parallax of 61 2 Cygni = o // .335 + o // .oo76 



which give : 



5/i _ (5/2 = -fo. // o82 ±o // .o23 [-f-o^.054 ±o".oio] 



and if the Rutherfurd results be included as given in Table 

 XXIII, this becomes : 



6i l — 6/ 2 = +o // .o48 ± o // .oi4 [+o // .035 zfco^.009] 



The numbers placed in the brackets are what these values would 

 have been had Pritchard's results not been discarded. 



4:9. Let us now tabulate these differences of parallax of the 

 two stars under consideration : 



Davis' Kutherfurd, — direct measures of distance +o // .i28 + // .o53 

 Davis' Rutherfurd, — direct measures of angles -f- .051 .032 



Adopted Mean = 



Davis' Wilsing, — direct measures of distance 

 Wilsing's (61 1 — 6) — (61 2 — 6)* 

 Mean of all determinations! previous to Rutherfurd = 

 Mean of all determinations including Rutherfurd = 



The magnitude and accordance of these values when considered 

 in connection with their respective probable errors leave little 

 room for doubt as to the reality of the difference of parallax de- 

 tected by the measures of the Rutherfurd plates, and discarding 

 Pritchard's results in no respect lessens the force of the argu- 

 ment. So that if 61 x Cygni be really a binary system of which 



* Publicationen des Astrophysikalischen Observatoriums zu Potsdam : Nr. 36. 

 page 148. 



fAs determined from Table XXIII on pages 155-6, using only the values in 

 bold type. 



+ 



0^.072 ± ".028 



+ 



.088 .012 



+ 



.048 .031 



+ 



.082 .013 



+ 



.048 .014 



