RECENT ADVANCES IN SCIENCE 121 



cautions must therefore be taken to obviate all possible sources 

 of error. The problem of the " Measurement of the Distances 

 of the Stars" was chosen by Sir F. W. Dyson, Astronomer 

 Royal, as the subject for the Halley Lecture delivered by him 

 in the University of Oxford on May 20 last (reprinted in the 

 Observatory, vol. xxxviii. No. 488) ; this forms a valuable account 

 of the difficulties attendant upon this problem, and of the manner 

 in which they have gradually been overcome. It was not until 

 1838 that the parallax of any star was determined with any 

 certainty, but in that year the parallaxes of three stars were 

 obtained by three different astronomers : that of a Centauri 

 by Henderson, from meridian observations made at the Cape; 

 that of 61 Cygni by Bessel, using a heliometer ; and of a 

 Lyrae by Struve, using a filar micrometer. The introduction 

 of the heliometer by Bessel was a considerable advance, and 

 until the application of photography this remained the best 

 method for the determination of parallaxes. Its disadvantage 

 was that the observations were laborious, and very great skill 

 was necessary in the manipulation of the instrument. However, 

 in the hands of Gill and Elkin at the Cape, of Elkin, Chase and 

 Smith at Yale, and of Peter at Leipzig, a number of valuable 

 results were obtained by means of it. 



The recent rapid advances have been due to the introduction 

 and perfecting of the photographic method, by which, with a 

 few simple precautions, results can be obtained of an accuracy 

 equal or superior to that reached in the best heliometer ob- 

 servations, and with much less trouble. In 1910, Schlesinger 

 published the parallaxes of twenty-five stars from photographs 

 taken with the 40-inch refractor at Yerkes, and he concluded 

 that " the number of stellar parallaxes that can be determined 

 per annum will, in the long run, be about equal to the number 

 of clear nights available for the work." This statement shows 

 the possibilities of the photographic method. During the past 

 year, the parallaxes of over fifty stars, also determined at 

 the Yerkes observatory, have been published by Slocum and 

 Mitchell {Astr. Nachr., Nos. 4709 and 4760), which are of the 

 same high order of accuracy. The latest contribution to our 

 knowledge comprises the parallaxes of about forty stars deter- 

 mined at the Royal Observatory, Greenwich (Monthly Notices, 

 R.A.S., 191 5, June), and equal in accuracy to the Yerkes results, 

 the mean probable error being + o' /- oo9. These latter results 



