June 1. 1894.] 



KNOWLEDGE. 



129 



long ellipse, while that of Beta Lyrfe can be inferred from 

 the equal duration of its phases to be sensibly circular. 

 Nest, because the ebb aud flow of light displayed in them 

 is so tranquil as to exclude the possibility of their con- 

 nection with any kind of explosive action. 



We seem, then, thrown back upon some description of 

 optical rntio)inh'. The most obvious is that of a double 

 eclipse, the brighter object being obscured at the principal 

 minima, the satellite at the subsidiary minima. This view 

 has in fact been adopted both by Father Sidgreaves and 

 by M. Belopolsky." Yet the light-curve of Beta Lyrfe 

 is totally different from those of Algol and other ordinary 

 eclipse stars. It represents continuous change, instead of 

 a settled maximum, interrupted at iixed epochs by a 

 comparatively swift decline and recovery. It would thus 

 appear that the Lyra variable can be associated with Algol 

 and its congeners only on one condition — that it is made up 

 of two enormous orbs circulating in such exceedingly close 

 contiguity that their mutual occultations form an almost 

 unbroken series. If each were of the solar mass, the 

 distance from centre to centre would then be roughly six 

 and a half million miles. Admitting further the separation 

 of their surfaces by an interval of one million, the diameter 

 of each of the conjoined bodies, which, for the sake of 

 simplicity, may be taken as equal, would be six millions of 

 miles. They would accordingly be three hundred and 

 thirty times rarer than the sun, their density exceeding 

 that of air at the sea-level only about three and a third 

 times. That globes thus tenuous should shine with stellar 

 brilliancy does not seem probable, but one should hesitate 

 to pronounce it impossible. 



Assuming their existence, we can safely assert that, 

 although they might escape eclipse, they could not escape 

 distortion through the inequalities of their mutual 

 attraction. They would, in fact, not be globular, and 

 their persistent extension along the line joining their 

 centres, and corresponding compression in the opposite 

 direction, would cause them to present to our view (if the 

 plane of their orbit made only a small angle with the 

 visual ray) a larger luminous area when hroad-siile-on than 

 when end-on. And since the range of variability in Beta 

 Lyrffi is of one magnitude, the joint extent of surface 

 visible respectively at the elongations and conjunctions of 

 the pair should be in the proportion of two and a half to one. 

 The disparity of their alternate minima would, indeed, 

 offer a difficulty, but it might be successfully met by an 

 ingenious contriver of hypotheses. All that we are at 

 present concerned to establish is that the periodical loss 



the entire orbital velocity is directed along the line of 

 sight, but coincide when it is directed across it. These 

 regularly alternating effects are, nevertheless, very im- 

 perfectly, if at all, recognizable in Beta Lyrse. The 

 extraordinarily complex nature of the problem with which 

 astronomers are thus confronted is emphasized by the 

 splendid series of spectrograms taken at Potsdam, mainly 

 by Dr. Wilsing, in 1892-3, and discussed, in his usual 

 masterly style, by Dr. VogeLf The prismatic section 

 depicted in them lies between wave-lengths -150 and .380, 

 so that they include the remarkable line at \ 4-17 together 

 with a considerable extent of the ultra-violet radiations. 

 They are, besides, unprecedentedly numerous, and appear to 

 be of pre-eminent definition, constituting altogether a 

 stock of materials of the very highest importance. 



The spectral changes in this star are particularly well- 

 accentuated in the line H s. as the first member of the 

 ultra-violet hydrogen-sequence is termed in the new system 

 of nomenclature introduced by Dr. Vogel. The accom- 

 panying figures, reproduced by his kind permission, show 

 its characteristic aspects at the four critical epochs of light- 

 fluctuation. In Fig. 3 we see its condition at the principal 

 minimum. A brilliant wide ray lies on the red or less 

 refrangible side of an equally wide dark ray. The ensuing 

 maximum (Fig. i) brings little alteration ; only a slight 

 brightening comes in on the violet side of the absorption. 

 At the second minimum, however, a totally different state of 

 things is seen to prevail (Fig. 5). The black line is projected 

 upon a much broader bright band. And the next maximum 

 (Fig. 6) unexpectedly follows suit. The chief minima of 

 the star, then, take place when the H ^ lines are furthest 

 separated ; and the remaining hydrogen and other lines 

 tell the same story. This is as much as to say that these 

 mysterious obscurations are not due to eclipses, which are, 

 of course, geometrically impossible except when the 

 occulted and occulting bodies are travelling straight 

 athwart the line of sight. An additional circumstance of 

 great perplexity is the spectroscopic correspondence of 

 adjacent, and not, as would naturally be looked for, of 

 opposite phases of luminous variation. It obliges us to 

 discard utterly the supposition of motion-displacements in 

 a nearly circular track ; unless, indeed, we are permitted 

 to bring to the rescue a swift translation through space of 

 the entire system. This would also account for the 

 unsymmetrical character of the recorded line-shiftings, for 

 the bright lines travel very much further from the dark 

 ones towards the red than towards the blue, the com- 

 pensatory swing being almost nuU. But we can arrive at no 



H( 



hS 



Fig. 2. — General Spectrum of R Lyraj from wave-length A. 380 to A 450. Reproduced from paper of Dr. Togel, published in the 



Sitzuiigslerichte of the Berlin Acad. d. Wiss, February 8th, 1894. 



of light in this star, if an optical effect of revolution in an 

 orbit, can only occur when its components are crossing 

 our line of sight. 



The spectroscopic evidence on the same point may now 

 be glanced at. Visually indivisible binaries, moving in a 

 plane passing nearly through the earth, show a double 

 spectrum at elongations, a single spectrum at conjunctions. 

 That is to say, the coupled lines in the dispersed light 

 of the companion-suns are separated to the utmost when 



• Memorie degli Spettrosco^isti Italiani, t. XXII., June, 1893. 



reliable conclusion on this point until Dr. Vogel carries 

 out his intention of measuring these shiftings with the 

 help of a comparison-spectrum. Information as to their 

 absolute values, their progress and plan, will then be 

 forthcoming. It is true that the width and close juxta- 

 position of the lines interpose serious obstacles to the 

 success of such operations ; but at Potsdam, if anywhere, 

 they will be overcome. 

 An inspection of Dr. Vogel's beautiful drawings suggests. 



t Sitzungsberichte, Berlin, 8th February, 1894. 



