38 



KNOWLEDGE 



[Februaky 1, 1895. 



the Galaxy ; a transition from a more homogeneous condition 

 to one of greater agglomeration in certain regions. And this 

 transition is one which is going on, not only with the 

 nebulous matter, but also with the stellar. 



Hcttcrs. 



[The Editor does not hold himself responsible for the opinions or 

 statements of correspondents.] 



• 



To the Editor of Knowledge. 



Deak Sir, — The remarks on the spectrum of /3 Lyras, 

 made by Father Sidgreaves in the last number of 

 KxowLEDGE, are of especial interest just now, when the 

 nature of that star is one of the moot points in astro- 

 physics. Nor does there seem to be any good ground for 

 calling in question the validity of his reasoning as regards 

 the possible projection of a partially dark upon a bright line 

 in such a compound spectrum as /3 Lyrse is reasonably 

 inferred to show, liut the line produced in the way he 

 suggests would still be fairly strongly illuminated ; only a 

 differential effect could be produced. In p Lyrte, never- 

 theless, the projected rays seem to be absolutely black. 

 One of them certainly is. Dr. Vogel expressly states that 

 there is a total absence of chemical action along the line 

 enclosed within the briUiant band representing the first 

 of the ultra-violet hydrogen series (H ^), in which the 

 spectral changes of the star are most characteristically 

 shown. A hmt is thrown out that this remarkable 

 circumstance may be a photographic effect due to rapid 

 development ; but the fact remains that the stripe is 

 totally dark. . Hence there appears no alternative but to 

 conclude that where it crosses the spectrum the gaseous 

 light is entirely cut off. This can only be explained, so far 

 as we can see at present, as a " reversal." 



A. M. Clerke. 

 — I • I — 



THE STEUCTURE OF THE UNIVERSE. 

 To the Editor of Knowledge. 



Sir, — With regard to the views of Prof. Kapteyn, 

 expounded in your last by Mr. Gore, there are some 

 remarks which occur to me. That the stars which 

 surround the sun are chiefly of the solar type has, I think, 

 been substantially proved, though the precise degree of 

 preponderance has not yet been ascertained. But when 

 it is alleged that at a greater distance this relative 

 preponderance is reversed, and that the Galaxy consists 

 chiefly of Sirian stars, I think we are going entirely 

 beyond the evidence. 



According to the best estimate at present available, a 

 Sirian star is probably, on the average, abouttwo magnitudes 

 brighter than a solar star, whose mass and distance is the 

 same. Supposing the stars to be uniformly distributed, the 

 number would increase about fourfold for each magnitude. 

 Hence, if we use the number of Sirian and solar stars of 

 any given magnitude as an index to their proportions 

 throughout the sky, the former would appear to be about 

 sixteen times as numerous as they really are. In point of 

 fact, however, the stars do not increase in number, as the 

 magnitude diminishes, so rapidly as this. The true dis- 

 proportion would probably be nearer to ten times than 

 sixteen times. But this difference is sufficient to upset 

 our calculations. In comparing the Sirian stars of the 

 sixth magnitude, for instance, with the solar stars of the 

 same magnitude, we are comparing two sets of stars the 

 average distance of one of which is at least double that of 

 the other ; and if we compare all the stars of each kind 

 up to and including that magnitude, we are comparing the 



stars comprised in two spheres, one of which has half the 

 radius of the other. 



Let us pass, then, from magnitude to proper motion. 

 Suppose, for instance, that we find a preponderance of 

 Sirian stars among those whose proper motion lies between 

 0-04 ' and 007" annually. Do we know all the stars 

 whose proper motions lie between these limits '? Certainly 

 not. We seldom know the proper motions of very faint 

 stars, though the case of a binary star with a faint 

 companion often indicates that there are faint stars whose 

 proper motions are considerable. In fact, we only know 

 the brighter stars whose proper motion lies between the 

 limits in question, and we might naturally expect that 

 the majority of these brighter stars would be Sirians. 

 Statistics based on a particular amount of proper motion 

 are of no value unless we know a large majority of 

 the stars with that proper motion, or at least the stars 

 which we know may be considered as impartial representa- 

 tives of the class. But as regards stars with small proper 

 motion neither of these conditions is fulfilled. We do not 

 know all the stars which possess the motions in question, 

 and those which we do know are selected on account of 

 their brightness, thus giving an xmdue advantage to the 

 Sirians. 



I think if any one wiU examine the solar stars down to, 

 say, the fifth magnitude, tabulating their proper motions, 

 and then estimate what the magnitude of a solar star with 

 a proper motion of 0-04" might be expected to be, he will 

 find an easy explanation of why we know of so few solar 

 stars with this proper motion. 



Nor do I think the photographic results attained in the 

 Galaxy conflict with these considerations. Sirian stars 

 are even more superior to solar stars in their photographic 

 power than in their light. If they appear to be multiplied 

 tenfold to the eye, they probably appear to be multiplied 

 fiftyfold in the photograph. But, generally speaking, long 

 exposure brings out more and more stars in the Galaxy 

 without any limit so far as we know at present. What are 

 these stars '? Are they Sirian stars situated at greater 

 distances or of smaller mass ? Or are they solar stars 

 mixed up with Sirians, but which require much longer 

 exposure in order to impress their images on the plate ? 



The whole question is evidently one in which superficial 

 appearances may deceive us. The evidence requires not 

 only to be stated but to be carefully weighed, and it is by 

 no means the only instance in which astronomers seem to 

 me to have deduced conclusions from insufiicient premises. 

 When we have ascertained the spectra and proper motions 

 of all stars down to the eighth or ninth magnitude, we shall 

 be in a better position to offer an opinion ; and it is by no 

 means impossible that the solar stars may be found to 

 retain their relative preponderance as far as our list of 

 proper motions can be regarded as complete. I believe they 

 do so at present. 



In conclusion, I wish to make one or two objections to 

 the theory that solar stars develop into Sirians. In the 

 first place the difference in brilliancy is, I think, too great; 

 but, in the second place, the solar stars whose spectra 

 approach nearest to the Sirian type — the Capellans, with 

 the spectra designated E, F, and G in the Ih-ajwr Catahigiu 

 — appear to be the least brilliant of solar stars. On the 

 theory of development, therefore, solar stars must diminish 

 in brilliancy as they approach the Sirian type, although 

 the increase in brilliancy is great when they attain it. 

 Indeed, the fact that the spectral types B and F stand at 

 opposite ends of the scale of proper motion seems to me 

 fatal to any theory of development. 



Truly yours, 



W. H. S. MoNOK. 



