April 7, 1916] 



SCIENCE 



481 



sible parallax. This has happened ia Italy to as- 

 tronomers of the very first reputation. The Dub- 

 lin instrument is superior to any of a similar con- 

 strnetion on the continent; and accordingly it 

 shows a much less parallax than the Italian as- 

 tronomers imagined they had detected. Conceiv- 

 ing that I have established, beyond a doubt, that 

 the Greenwich instrument approaches still nearer 

 to perfection, I can come to no other conclusion 

 than that this is the reason why it discovers no 

 parallax at all. 



Within fifteen years after this statement 

 by Pond, observations had been obtained 

 which showed a measurable parallax of 

 three different stars. The announcements 

 of these results, each by a different astron- 

 omer, were practically simultaneous. 



W. Struve, using a filar micrometer, 

 determined the distance of a Lyrte from a 

 small star about 40" distant on 60 differ- 

 ent days over a period of nearly three 

 years. He obtained a parallax of 0".262 ± 

 0."025. Bessel, using his heliometer, deter- 

 mined the distance of 61 Cygni from two 

 small stars distant about 500" and 700", 

 respectively. He .obtained for this star a 

 parallax of 0".3M ± 0".020. Henderson, 

 using determinations of the position of a 

 Centauri by meridian instruments, deduced 

 a parallax of 1".16 ± 0".ll. All three of 

 these results were announced in the winter 

 of 1838-39, and indicate that the three 

 stars are distant from the earth about 750,- 

 000, 650,000 and 200,000 times the distance 

 of the sun from the earth. 



The accompanying table exhibits the ob- 

 served displacement of 61 Cygni by 

 monthly means as given by Main from 

 Bessel 's observations. The last column 

 gives the computed displacement on the 

 assumption of a parallax of 0".314. The 

 reality of the parallax is seen at a glance. 



In 1888, fifty years after the first deter- 

 mination of what we now kaow to be a true 

 stellar parallax, Young, in his General 

 Astronomy, gives, in a list of known stellar 



parallaxes, 28 stars and 55 separate deter- 

 minations. Within the next ten years the 

 number of stars whose parallaxes had been 

 determined about doubled, due principally 

 to the work of Gill and BUdn. 



PARALLAX OF 61 CYGNI 



Observed Computed from 



Displacement 0".514 



Mean Date " " 



1837 August 23 + 0.20 -)- 0.18 



September 14 . . -|- 0.10 -f 0.08 



October 12 -1- 0.04 — 0.05 



November 22... — 0.21 — 0.22 



December 21... — 0.32 — 0.27 



1838 January 14.... — 0.38 — 0.27 



February 5 — 0.22 — 0.23 



May 14 -f 0.24 -|- 0.20 



June 19 -f- 0.36 -|- 0.28 



July 13 -1-0.22 -fO.28 



August 19 -f 0.15 -F 0.19 



'September 19.. . 4- 0.04 -|- 0.06 



Probably the most extensive piece of 

 stellar parallax work in existence is that 

 with the Yale heliometer. The results to 

 date were published in 1912 and contained 

 the parallaxes of 245 stars, the observa- 

 tions extending over a quarter of a cen- 

 tury, the entire work having been done by 

 three men, Elkin, Chase and Smith. In 

 selecting a list of stars for parallax work an 

 effort is made to obtain stars which give 

 promise of being nearer than the mass of 

 stars. At first the brighter stars were 

 selected, and then those with large proper 

 motions. The Yale list of 245 stars eon- 

 tains all stars in the northern heavens 

 whose annual proper motion is known to be 

 as much as 0".5. Of these 245 stars, 54 are 

 given a negative parallax. A negative par- 

 allax does not mean, as some one has ex- 

 pressed it, that the star is "somewhere on 

 the other side of nowhere," but such a re- 

 sult may be attributed to the errors of ob- 

 servation or to the fact that the comparison 

 stars are nearer than the one under inves- 

 tigation. It is safe to say, however, that 



