124 



NATURE 



of good students. If either supply fails, it Is not 

 geography alone, but all sciences and studies, that 

 will be damnified ; for all require the best of the 

 help she can give in proportion as her science 

 grows and improves. History will be able to call 

 but indifferent geography to her assistance if this 



[September 22, 192 1 



science has been understaffed and discouraged by 

 official ijeluctance to allow it a place of its own 

 in the sun. Is there not still some such reluctance 

 on the part of the Board of Education, of some 

 of our universities, and of the Civil Service Com- 

 missioners? 



Stellar Parallax. 



By J. Jackson, Chief Assistant, Royal Observatory, Greenwich. 



THE determination of stellar distances is funda- 

 mental to the investigation of the sidereal 

 universe. When once the distance of a star is 

 Known, we can calculate its transverse speed in ! 

 kilometres a second from its proper motion in | 

 seconds of arc per century, and we can determine j 

 its absolute brightness from its apparent bright- | 

 ness. For binary stars at known distances we can ■ 

 determine the separation of the components in j 

 kilometres, and this, together with the period, 

 enables us to compute the mass of the system. 

 Recent work at Mount Wilson has shown that it 

 is practicable to determine the angular diameter 

 of the larger stars, and for such stars a know- 

 ledge of the parallax will enable us to compute 

 the linear diameter. Many of the investigations 

 about the sidereal universe made during the last 

 twenty years have been possible only through the 

 increase in the number of stars the distances of 

 which are known with reasonable accuracy, and 

 the results obtained have been of such importance 

 that in an increasing degree the energy of astro- 

 nomers is being directed to supply the required 

 data. 



The direct determination of stellar distances 

 depends on triangulation from the earth at dif- 

 ferent positions in its annual path round the 

 sun. The apparent motion of a star can be 

 analysed into a linear component due to the rela- 

 tive motion of the sun and the star, and a periodic 

 motion due to the motion of the earth round the 

 sun. The parallax of a star is the angle sub- 

 tended by the earth's radius at the distance of the 

 star, and is equal to the semi-major axis of the 

 apparent ellipse described by the star as a result 

 of the earth's orbital motion. It is therefore 

 determined from observations made as nearly as 

 practicable at the times when the star is at the 

 ends of the major axis of the "parallactic ellipse." 



The principle to be used in the determination 

 of stellar distances was obvious as soon as the 

 Copernican theory of the solar system was recog- 

 nised. The difficulty in applying the principle is 

 due to the extreme minuteness of the change in 

 angle which is to be measured after a six months' 

 interval. Had it been known or assumed that 

 the stars w^ere comparable in real brightness with 

 our sun, although they appeared about a million 

 million times fainter, it could have been calcu- 

 lated that their parallaxes were less than a second 

 of arc. Successive attempts to measure the paral- 

 laxes of selected stars for long necessarily met with ' 

 NO. 2708, VOL. 108] 



failure, although they led to many important dis- 

 coveries. In the first half of the eighteenth cen- 

 tury Bradley made a remarkable series of observa- 

 tions of the meridian zenith distance of the star 

 ■y Draconis — a star which passed the meridian 

 near the zenith so that the angles to be measured 

 were relatively small, while errors introduced by 

 varying atmospheric conditions were reduced to a 

 minirnum. He discovered aberration and later 

 nutation through these observations, and proved 

 that the parallax of this star was less than a 

 second. Observations of the same kind might 

 later have led to the discovery of latitude varia- 

 tion. The attempt made by Sir William Herschel 

 towards the end of the eighteenth century may 

 also be noted here. Instead of attempting to 

 determine the absolute parallax of separate stars 

 which involves the measurement of large angles 

 from the vertical or some other direction which 

 it is supposed can be accurately identified after 

 a six months' interval, he attempted only to dis- 

 cover relative parallaxes from the relative dis- 

 placements of stars in nearly the same direction, 

 but probably at very different distances. The 

 method is essentially that now used almost ex- 

 clusively ; but where Herschel applied it to pairs 

 of stars actually at different distances, his observa- 

 tions were not sufficiently accurate to reveal the 

 parallactic motion. His most extensive series of 

 observations were made of fairly bright pairs of 

 stars within a few seconds of arc, and the motion 

 he actually discovered was orbital motion of the 

 stars, which proved that they were really close 

 together in space and revolving round one 

 another under gravitational attraction. This dis- 

 covery led to the systematic study of double stars. 

 Success in the determination of stellar parallax 

 was obtained almost simultaneously about 1838 

 by three observers employing different methods on 

 different stars. The principal credit is usually 

 given to Bessel for his determination of the paral- 

 lax of 61 Cygni — a pair of faint stars with large 

 proper motion — relative to faint neighbouring 

 stars by means of the heliometer. This instru- 

 ment consists of an ordinary telescope with the 

 object glass cut in two along a diameter, and 

 means are supplied for rotating the object glass 

 and for sliding the two halves along their 

 common diameter. With this instrument each 

 star forms two images, and the observation con- 

 sists in bringing an image of one star into coinci- 

 dence with the other image of the other star. The 



