424 



NATURE 



[January 25, 191 2 



easily 



i \n^ \vH«i rathrr diftiriilt, but on Deccmli- : 

 s'-fn and its granular structure detected. 



The " ANNfAiRE " of thk Bureau uks Lonoitudp.s, 

 1013. — Thin year's i««uc of the " Annuaire " contains the 

 usual astronomical information, ephcmcrides, and tables, 



md deals with chemistry and physics similarly. It also 

 lont.iins thf complete list of minor planets, for 714 of 

 V liirh it gives the orbital elements. .Among the " notes " 

 ih' I'- appear an interesting risum6 of solar physics by M, 

 l)i«i|.-mdres. .» long discussion on the various calendars, an 



iirirl' hv M Biijoiirdan on seismology, a brief description 

 titution of the moon by M. Puiseux. 

 he mean temperature in various parts 

 ,.i li.iriK' Dv .M. Bigourdan. All the times in this 

 " .\nnuaire " are given in accordance with the new law 

 of March I'^f 111" priro cif ih.> " Annu.nire " is 1.50 

 francs net. 



A' 



TffR n A DIAL VELOCITIES AND SPECTRAL 

 TYPES OF STARS. 



l.lHOUCili the determination of radial v.ii.<..ii.^ is, 

 as a practical proposition, a development of com- 

 paratively recent years, the data already secured by the 

 •. ivt-ral observatories doing line-of-sight work promises 

 lo be of inestimable value in the study of cosmological 

 problems. ii is no longer merely a question of 

 •' approach " or " recede " — a far greater vista has been 



i[)ened up as the worli has proceeded. Combined with the 



•searches of Kapteyn, Hddington, Dyson, and others on 



\\\f >tr<amin.'j Uiidencies disclosed by the discussion of 



• ))n>|irr iimiinii^, ■' it promises a rich mine of as yet un- 

 iliM lu-.ti I:m t> < niKirning the evolution of enormous 

 ■■idereal syst^ ni>. 



.A glance tlirun^h I'rot. Campbell's second catalogue of 

 spectroscopic binaries ' tells of much work already accorn- 

 plished, and the discussion discloses how much there is 

 still to be done. 



One outstanding result of this discussion of more than 

 ;oo binary systems, so far as it applied to the comparative 



■ \v for which the periods of revolution and other orbital 

 • lements have been determined, was the fact that the 

 motions of the stars are intimately related to the spectral 

 types which may be taken as indices of stellar ages. 

 Briefly, it appears that the older a binary system becomes 

 the greater bJecomes the eccentricity of the orbit and the 

 longer grows the period of revolution. 



The existence of the relation between radial velocities 

 and stellar types was also brought out in a later paper" 

 •dealing; with '-(une peculiarities in the motions of the stars, 

 where ih. tollowing table was given, the spectral types 

 being given under the Harvard designations : — 



Spectral types 

 O and B 



A 



F 



G and K 

 M 



No. of stars 

 ... 141 .. 

 ... 133 •• 

 ... 159 .. 

 ... 529 ., 

 ... 72 .. 



Average radial velocities 

 ... 8-99 km. 

 - 994 „ 



... 1390 M 

 ... 1515 M 



... 16-55 .. 



In the general discussion it was found that the B-type 

 (" helium ") stars called for special treatment, and Prof. 

 Campbell discussed the motions of the brighter stars of 

 this type in a further paper.' 



The main conclusion to which we wish to direct atten- 

 tion here is that " .An error, of obscure source, causes the 

 radial velocities of Cla.ss B stars to be observed too great 

 by a quantity, K, amounting to several kilometres. For 

 stars of Class B-Bj the value of this error is approxi- 

 mately K=+4.7 km. per second. The value obtained for 

 Class B-B, stars is +41 km. It is therefore probable 

 that K is less than +4 km. for stars of Class B,-B,." 



The result was based on the consideration of 225 Class 

 B stars, K being an assumed, unavoidable, but systematic 

 error inherent to the observed velocities, and disclosed in 

 calculating from these the velocity of the solar system in 

 space. This systematic difference, peculiar to the B-type 

 stars, has led to some very interesting and important 



1 Lick Observatory Bulletins, No. 181. 2 Ibid., No. 196, ' Ibid., No. 195. 



NO. 2204, VOL. 88] 



Miggestion* as lo the characteristics of the stm 

 sewcs ; Prof, Camplxdl make* several tentative si 

 as to its source. Of thes*,', we would direct atitniion i.i 

 that in which it is suggested that in these stars the 

 absorption takes place in the k>wer layers of the atmo- 

 spheres, and thenforc under greater pressure, thus modify- 

 ing the efli-ttive wave-l<-ngth and producing the error when 

 the measures of wave-length are compared with terrcstri.il 

 standards. Another suggestion points out that the helinri: 

 lines frequently used in the measures are double, witii 

 the red component, in the laboratory, much the faini'i. 

 If the conditions in the star increase the relative intenMitN 

 of these red components, the wave-length of the centre of 

 gravity of the whole line would be shifted, and the observed 

 difference be thus produced. 



.A most interesting contribution to the discussion of ih' 

 motions of this type of star is published by Dr. Luden- 

 dorfT * in the form of some remarks on \\\f classification 

 of helium stars. He takes from Campbell's list of 224 

 stars all those which have an absolute radial velocity, V,, 

 > 8-0 km., and shows that there is a distinct differentia- 

 tion of their velocities if they are arranged according to 

 Lockyer's classification * of the helium stars. 



•As is generally known, this classification is unique in 

 that it aims at arranging the classes of stars in an evolu- 

 tional order, difference in the chemical characteristics, 

 accompanying differences in age and temperature, being 

 the criterion. It also essentially takes into account the 

 idea that stars probably increase before decreasing in 

 temperature, and on these lines arranges the helium stars 

 at the top of the temperature curve thus : — 



AIniiamian 

 Argonian 



Crucian 



In 



Taunan 

 (Rieeliao) j S 5 



\%. \ .Achcmian 



^^^ \ 

 1^/^ \ Algolun 



In the Lockyer classification the Rigelian class is not 

 given as essentially a helium class, although helium is 

 mentioned as one of the prominent elements ; for this latter 

 reason Dr. Ludendorff includes it in his discussion. 



Of the sixty -three stars selected from Campbell's list, 

 he finds eighteen classified in the South Kensington cata- 

 logue, and tabulates them as follows, the velocities being 

 given in round numbers : — 



From this table evolves the striking fact that, without 

 exception, those stars placed by Lockyer on the ascending 

 arm of the temperature curve all have -r velocities, and, 

 with one e.xception, those on the descending side have 

 — absolute radial velocities. It w^ould appear extremely 

 unlikely that this remarkable division is due to chance ; 

 but Dr. Ludendorff seeks further evidence by taking from 

 Campbell's catalogue all (seventy-one) those stars classified 

 at South Kensington as helium stars — including the 

 Rigelian class — and arranges them as follows, - giving the 

 mean absolute radial velocities Vj* of each class : — 



< Astronomistht NachrickttM. No. 4547, vol. cxc, p. 19^. 

 * Catalogue of 470 of the Brighter Stars Classified according to their 

 Chemistry. 



