July 2, 1915] 



SCIENCE 



21 



In questions of this kind the only satis- 

 factory way is to judge by the results. 

 From observations made on fifty nights, 

 values of the parallax are obtained not 

 nearly so accurate as the best heliometer 

 observations, but still of considerable ac- 

 curacy. Finally, the parallaxes of four of 

 the stars which had been previously deter- 

 mined by measures with a heliometer 

 showed satisfactory agreement. 



This method has been employed by Jost 

 at Heidelberg, very extensively by Flint at 

 the "Washburn Observatory of the Univer- 

 sity of Wisconsin, and is now being tried 

 at the Cape by Voute, a pupil of Kapteyn's. 

 It appears to me that this method can never 

 give results of the highest accuracy, but 

 that it may be of use in a preliminary 

 search for stars of large parallax. The 

 argument of the facility of the method 

 compared with the heliometer has, how- 

 ever, lost much of its force ; for, as I hope 

 to show next, the highest accuracy attain- 

 able with the heliometer can be secured 

 much more easily with a photographic 

 telescope. 



The application of photography to the 

 determination of stellar parallax was first 

 made by Pritchard in Oxford between 1887 

 and 1889. He took a large number of 

 photographs and measured on them the 

 angular distance of the star which he was 

 considering from four of its neighbors. In 

 this way he determined the parallax of five 

 stars. He began this work late in life, and 

 it was left for others to develop the photo- 

 graphic method and find what accuracy 

 could be attained with it. At first sight 

 it seems very easy, but experience shows 

 that there are a number of small errors 

 which can creep in and vitiate the results, 

 unless care is taken to avoid them. 



It has gradually become clear that with 

 a few simple precautions and contrivances, 

 a greater accuracy can be reached in the 



determination of parallax by photography 

 and with much less trouble than by any 

 other method. Between 1895 and 1905, 

 several astronomers succeeded in obtaining 

 from a few plates results as accurate as 

 could be obtained from many nights' ob- 

 servation? with the heliometer by the most 

 skilled observers. In the last five years a 

 large number of determinations have been 

 made. In 1910 Schlesinger published the 

 parallaxes of twenty-five stars from photo- 

 graphs taken with the 40-inch refractor of 

 the Terkes Observatory, and in 1911 Eus- 

 sell published the parallaxes of forty stars 

 from photographs taken by Hinks and 

 himself at Cambridge. The opinion ex- 

 pressed by Gill on these observations' was 

 that but for the wonderful precision of the 

 Yerkes observations, the Cambridge results 

 would have been regarded as of the high- 

 est class. The facility with which the 

 Terkes results are obtainable is expressed 

 very tersely by Schlesinger :" 



The nuiubeT of stellar parallaxes that can be de- 

 termined per annum will in the long run be about 

 equal to the number of elear nights available for 

 the work. 



With the heliometer at least ten times as 

 much time would have been required. Dur- 

 ing the last year two further installments 

 of the results of the Terkes Observatory 

 have been published by Slocum and 

 Mitchell, giving the parallaxes of more than 

 fifty stars. It might be thought that the 

 high accuracy attained by them is largely 

 attributable to the great length of the tele- 

 scope. From experience at Greenwich, I 

 do not think this is the case, and believe 

 that similar results are obtainable with tele- 

 scopes of shorter focal length. As several 

 observatories are now occupied with this 

 work, we may expect that the number of 

 stars the distances of which are fairly weU 

 known will soon amount to thousands, as 



3 M. N., Vol. LXII., p. 325. 



