371 



KNOW Li:i)r,i:. 



OCTOitKR, 1912. 



Didcr, fvi-n aiiproachiiig tliat of li^;lu : aiul tlicsc 

 discharges concerned particles whose masses were in 

 the ratio 1 : 2 : 4, and tliiis explain a very remark- 

 able feature of inuitiple emanations at these 

 distances. For tliis reason it appears to me probable 

 that this was the real nature of the phenomenon.* 



When we recognize that stars which are at the 

 same distance vary enormously in luminosit\', we 

 have to admit that, unless the object selected happens 

 to be very near, the method of measuring relative 

 parallaxes by comparing positions of neighbouring 

 stars viewed from the ends of a radius of the ICarth's 

 orbit, and selecting faint com[)arison stars on the 

 plea that they are probably farther away, is very 

 unsatisfactory. The [parallaxes of a few hundredtiis 

 of a second obtained in this wa\' are often fictitious. 

 Even negative parallaxes may be found, showing 

 that the faint stars were really the nearer. In spite 

 of the great accuracy of instruments and observers, 

 the means for discriminating between the objects 

 selected remain insufficient. Hence it has seemed 

 to me wiser to acce[)t such an indication as has been 

 furnished b\- the outburst of nebulosit\- around Xova 

 Persei, rather than to trust to large numbers of 

 apparent parallaxes of a few hundredths of a second. 

 which are affected iiy tlu- unknown parallaxts of tin- 

 comparison stars. 



Of twenty-eight stars between the fourth nnd 

 tenth magnitudes, with annual proper motions i:iiig- 

 ing between 0"-01 and ,)"-7.t. three whicii iuul 

 proper motions of 0"-01, but wliirh were an:ong the 

 brightest stars on the list (magnitudes J-8 to 4-7). 

 all gave to Dr. .Schlesinger negative parallaxes large 

 enough to show that they were more distant than 

 the comparison stars.' The tollowing are tiic mean 

 parallaxes : — 



4th iiiat;nitiide (2 stars) 



5th .. (4 stars) 



Sth ., (10 stars) 



qth ,. (H stars) 



10th .. (2 stars) 



IT = -0"-002 

 + -045 

 + -085 

 4-0 -145 

 + -173 



Most of these stars were selected on account of 

 their large ()roper motions, but the four brightest 

 ones were chosen to give examples of the Orion t\pe. 

 The series is interesting because it relates to fainter 

 stars than are usually picked out for parallactic 

 measurement, and because, like Eastman's list of 

 proper motions cited above, it does not show any 

 falling off with diminishing brightness. That a 

 fourth-magnitude star of small proper motion ma\- 

 be farther away than a ninth-magnitude star of 

 large proper motion, is a result which justifies the 

 use of large proper motion, rather than that of 

 brightness, as a criterion of nearness ; but the 

 phenomenon of star-drift suggests that proper 

 motion, in turn, may fail as a test, especially for 



stars b(4onging to an assemblage which shares with 

 our .Sun in a common motifin through space. 



If the Sun were reirioved until it appeared as a 

 seventh-magnitude star, it would still be three times 

 as distant as the mean of the ninth-magnitude stars 

 measured by Dr. Schlesinger. We are not making 

 any large demand on i)robability if we assume that 

 a particular strand of the Milky Way consists of 

 seventh to fifteenth-magnitude stars no farther away 

 than our hy[)othctical seventh-magnitude Sun, all 

 moving in the same general direction and having 

 neither recognizable [larallax nor [iroper motion. If 

 the Sun shares the motion only in part, a small 

 stream-drift, such as is indicated by Kancken's 

 investigation, remains. 



\n annulus has been assumed as a first approxi- 

 mation to the galactic shape by most investigators, 

 because w'e have no means of differentiating between 

 this shape and the more general two-branched 

 nebidar spiral, since tlu' more remote turns of the 

 spiral are superposed and projected on the same 

 great circle of the sphere. The opposition of two 

 narrower and brighter parts of the Galaxy, together 

 with a similar opposition of the more wideK' diffused 

 parts of the ring at points 180 apart, indicate that 

 the Milky Way is reall\- a two-branched spiral. If 

 the Sun, instead of being in the central lumen of an 

 annulus, is situated between two branches of a spiral, 

 or as shown in Figure .559, the total lireadth of the 

 spiral figure may easily be five times the diameter of 

 a simple ring, or the radius of the outer boundary 

 of a circular disc may be five times the distance 

 from the Sun to the nearer condensations; and some 

 such unknown factor is implicitly understood in an 

 estimate which purports to give only the order of 

 magnitude involved. Hut if the linear dimensions 

 are increased five-fold, the assumed luminositv of 

 Nova .Andromedae must be twenty-five times greater. 

 This is perhaps not beyond the bounds of possi- 

 I)ilit\- ; but to increase the distances fifty-fold, as 

 Dr. Crommelin would have us do, increasing the 

 brightness of the nova in the ratio 2,500 : 1, puts a 

 strain on the probable luminosity of even such a 

 remarkal)le object as a nova, because my estimate 

 had already assigned to this star the status of one 

 of the brightest of the novae. This dilemma was 

 also instanced by Gore+ as an argimient against the 

 ascription of so large a distance to the Andromeda 

 nebula, and it has seemed to me conclusive. 



If the Gala.xy is a unit with two great branches 

 which are the relics of the original movements, then 

 two great opposing drifts, modified by minor local 

 drifts, ought to surpass all others. The hypothesis 

 hitherto accepted is thus stated by Newcomb ^ : 

 (Supposition a) "There are scattered around us in 

 the stellar spaces, in every direction from us. a large 



Frank W. Very, ".\ii In(|Uiry into the cause of the Nebulosity around Nova Persei." Am. Joitnial of Sci. l4l Vol. WI. 



page 49, July, 1903. 

 Frank Schlcsin),'er, ■'Photographic Oeterniinatious of Stellar Parallax made witli the \'erkcs Refractor." Astroplivsidil 



Journal. Vol. XX.XIV, page 26, July, 1911. 

 ', " KNOWLKDGli," July, 1909. 

 - " Tlic Stars," page 293. 



