126 



NATURE 



[September 22, 1921 



shows the aperture and focal length of the tele- 

 scopes used : — 



Allegheny 



Dearborn 



Greenwich 



McCormick 



Mount Wilson 



Sproul ... 



Yerkes 



The telescope used at Mount Wilson is a re- 

 flector, while the others are refractors. It will 

 be seen that the focal length varies considerably 

 frono one instrument to another, but the probable 

 error of a parallax determined from about fifteen 

 plates is in all cases about 001'' or a little less. 

 The explanation given is that with average con- 

 ditions of working the images are larger for the 

 longer telescopes. The aperture is also a point of 

 some importance, as with larger aperture the 

 duration of the exposure can be cut down, and 

 with it the "guiding error." However, as some of 

 the larger telescopes are really visual telescopes, 

 and require a colour screen or special plates, the 

 advantage they would otherwise have is reduced. 



Considerable progress has already been made 

 in carrying out this scheme of co-operation, and 

 probably at least 200 parallaxes a year are being 

 determined. The value of this contribution to 

 ■our knowledge of stellar distances is realised when 

 we recall that m 1880 we knew the parallaxes of 

 only about twenty stars, while as late as 191 5 

 the number had risen only to about 200. As one 

 of the recent large publications we may instance 

 the 260 determinations made at the McCormick 

 Observatory in the five years 1914 to 1919.^ 



Let us consider the first star in this list. It is 

 /3 Cassiopeiae, a star of magnitude 24 with a 

 proper motion of 55" a century. The parallax was 

 found from the rather large number of twenty- 

 •eight plates exposed as follows : — 



1914 July 

 Nov. -Dec. 



1915 Aug. ... 

 Nov.-Dec. 



The parallax found was 0058" +0011^', in satis- 

 factory agreement with other determinations, of 

 which we may quote 0051'' ±0015" determined 

 by Smith with" the heliometer, and oo74"+o-oii" 

 found by photography at the Allegheny Observa- 

 tory. The proper motion in right ascension was 

 found to be +0524", as compared with +0529" 

 found bv Boss from meridian observations extend- 

 ing through about 150 years. But the agreement is 

 not always so satisfactory. Consider, for example, 

 the star y Ceti, which forms the double 2, 299. 

 The components are of magnitudes 36 and 6-8, 

 and are separated by about 3". The duplicity of 

 the star might well lie at the root of the trouble, 

 although with suitable exposures the fainter star 

 should not be seen. The parallax found with the 

 Yale heliometer was 0119" +0017", that by 



1 Publications of the Leander McCormick Observatory of the University 

 of Virginia, vol. 3. 



NO. 2708. VOL. 108] 



I 1916 Aug.-Sept. 

 5 Nov.-Dec. 



3 1917 Aug. ... 

 3 Nov.-Dec. 



photography at .\llegheny 0014" +0008", and 

 that at the McCormick Observatory 0037" + 

 0008". The photographic parallaxes may be 

 considered as in fair agreement, but the proper 

 motion in right a.scension was found to be -'O059'' 

 (from plates extending i^ years), as against 

 — 0147" found by Boss from observations ex- 

 tending 150 years. The photographic proper 

 motion was checked by plates giving an interval 

 of 3A years, during which time the difference in 

 proper motion amounted to no less than 03''. 

 This cannot be explained by the orbital motion of 

 the components, which is extremely slow, and it 

 is difficult to attribute it to all four comparison 

 stars, which on examination showed no appreci- 

 able proper motion. Results of this kind are by no 

 means uncommon, and the greatest caution has 

 to be used in applying small parallaxes from 

 single determinations. 



It may be considered that parallaxes greater 

 than 005'' may be used in calculations concern- 

 ing individual stars, t'or smaller parallaxes the 

 accidental errors will make the results untrust- 

 worthy. It has been estimated ^ that there are 

 about 2000 stars with a parallax as great as 0-05'', 

 although most of them will be as faint as the 

 loth magnitude and not attract notice. .As the 

 number of stars to this magnitude is of the order 

 of a million, it will be difficult to identify the stars 

 which are near, but faint. Stars chosen at ran- 

 dom will therefore generally give very small 

 parallaxes. Most observing programmes, there- 

 fore, contain specially selected stars, such as very 

 bright stars, stars with large proper motion, and 

 binary stars in rapid orbital motion for which a 

 fairly large parallax may be expected. 



On account of this selection of the stars great 

 care has to be exercised in discussions based on 

 the measured parallaxes. Many facts, however, 

 have been brought to light. The most important 

 of these correlate what may be called the apparent 

 qualities of a star with its absolute qualities. By 

 the former we mean those qualities which can 

 be found from observation without a knowledge of 

 the distance, such as the nature of the light a star 

 emits, or its angular motion, and by the latter 

 the intrinsic qualities, such as real brightness, 

 mass, and linear speed. Probably the most im- 

 portant results are those which have been reached 

 at Mount Wilson, where for the later-type stars 

 the relative intensity of certain spectral lines has 

 been correlated with the absolute brightness. It 

 is then a simple matter to deduce the distance 

 from our knowledge of the absolute and apparent 

 brightness. Parallaxes determined in this way 

 are called "spectroscopic parallaxes," and recently 

 a list of 1646 of these has been published by the 

 Mount Wilson observers. Again, for double stars 

 with known orbits and known parallax the mass 

 of the system can be computed. It is found that 

 the mass of all systems does not differ widely from 

 twice that of the sun. Assuming this mass, we 

 car compute "hypothetical" or "dynamical" 



2 Etidinglon, " Stellar Movements," p. 15. 



