86 



KNOWLEDGE 



[April 2, 1900. 



(2) a rapid thinning-out of the stars as we proceed to a 

 great distance from the sun, or (3) a rapid decrease in 

 the luminosity of the stars — which would have the same 

 effect as a decrease in their number. 



Taking the successive members of our series as thp 

 stai-s of each successive magnitude, it would be an in- 

 creasing, not a decreasing, series, on the assumption tha^ 

 the distribution of the stars was uniform, and that there 

 was no absoi-ptive medium. Adopting Pogson's scale, 

 the total light of "^he stars of the n-f-lth magnitude 

 would always be equal to 1.585 times the total light of 

 the stars of the nth magnitude. Can any reasonable man 

 who looks at the sky imagine that this process is carried 

 as far as the stars of the 1000th magnitude? 



What light an infinite number of stars situated at an 

 infinite distance might or might not give us is an in- 

 soluble problem. I only profess to deal with stars at 

 finite distances. The main question is whether there is 

 a gi-adual thinning-out of these as we proceed to great 

 distances from the sun. I hope your future corre- 

 spondents will confine themselves to this issue. 



The distance of the sun from its nearest neighbour 

 and Centauri is so vast that many persons seem to con- 

 clude either that we do not belong to the Galactic 

 System or that we are situated in a void space in the 

 Galaxy. I cannot see that either of these alternatives 

 has been proved. The sun, so far as I can judge, is 

 very probably a Galactic Star situated in a region which 

 is neither unusually dense nor unusually rare. If so, 

 a great diminution in the density of the stars may be 

 expected when we once get outside of the Galactic 

 Cluster. The stars outside of it may, notwithstanding, 

 extend to infinity or even be infinite in number, but 

 then much more sparse distribution outside of the 

 Galactic Cluster would accoiuit for what we see. 



The Galaxy, as we see it. is nearly a great circle on 

 the sphere This fact suggests that the sun (and earth) 

 is in it, not outside of it — if indeed the galaxy is i.ot a 

 hollow ring near the centre of which we are situated. 

 But the gradual decrease in the density of the stars 

 as we proceed towards the Poles of the Galaxy is hardly 

 consistent with the theory that the sun occupies an ojien 

 space in the centre of the ring. There may have been 

 something in the old idea that the eai-th was the centre 

 of the Universe — viz., that it is one member of vast 

 central constellation, and may therefore be regarded as 

 the actual centre when the distance considered is suffi- 

 ciently great W. H. S. Monck. 



P.S. — When I first wrote on the subject I fell into the 

 same error as Mr. Inglis, viz., that on the hypothesis 

 of uniform distribution the total light of the stars of the 

 nth and n-)-lth magnitude would be the same. I made 

 this mistake in consequence of considering the surface 

 of the sphere instead of its solid contents. I was fortu- 

 nate enough to be the first to point out the error into 

 which I had fallen. But I am afraid that I am 

 theorising too much for vour reviewer. — W. H. S. M. 



IS THE STELLAR UNIVERSE FINITE? 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — Mr. Monck writes : " Bright stars lose as mucli 

 by absorption, atmospheric or telescopic, as fainter 

 ones." Individually they do not lose the same propor- 

 tion of their total light, which is the essential point so 

 far as my argument is concerned. A faint star becomes 

 invisible near the horizon, while a bright star still re- 

 mains visible at a similar altitude. Although the com- 

 bined light of a number of stars too faint to be 



separately visible may certainly, as he points out, pro- 

 duce the sensation of light, yet if the atmosphere absorb-; 

 80 much of the light of each individual star as to render 

 it invisible as such, then none of those stars are included 

 in the counts or gaugings of stars on which Mr. Burns 

 relied, although they might actually exist. Conclusions, 

 therefore, based upon these counts, absorption being 

 iieglected, appear to me to be erroneous. 



Wm. Anderson. 

 Madeira, Febnian- 12th, 1900. 



THE CONSTITUENTS OF THE SUN. 



TO THE EDITORS OF KNOWLEDGE. 



[I was wrong, no doubt, in writing " dark," seeing 

 that the laboratory spectrum of carbon consists of bright 

 lines. But even if we deal with these bright lines or 

 flutings, I still maintain that the coincidence of all of 

 them with dark solar lines is not made clear to me in 

 Fig. 3. I admit that the fluting beginning at 3879 

 and ending at 3883 coi-responds generally with that in 

 the solar spectrum, but the break or change in con- 

 tinuity shown in the latter at about 3882 does not appear 

 in the carbon spectrum ; neither does the abrupt ter- 

 mination of the fluting coincide exactly in position with 

 that of the dark. Then, again, as to the beautifully 

 regular fluting c-ommencing at 3874 and ending at 3878, 

 there may be a dark line coiTesponding in position to 

 each of the brighter, but the general character of the 

 fluting is by no means represented on the dark lines — 

 that is to say, the vei-y slight but regular crowding 

 together of the lines towards the right. Similarly at 

 wave lengths lower than 3874, no particular corre- 

 spondence suggests itself to my ej^e. I do not deny tha<- 

 there is a correspondence between the lines of caibon 

 generally, with dark lines in the sun ; I merely reiterate 

 that the figure in question does not, to me, give that 

 convincing visual coincidence which the textrbooks lav 

 stress upon as proving beyond possibility of doubt the 

 existence of certain elements in the sun. — E. E. M.j 



[To my former remai-ks I need only add that in the 

 solar spectrum the flutings of carbon are superposed 

 upon lines due to various other elements, so that excep' 

 in the case of the fluting commencing at 3883, the corre 

 spondence with the arc spectimm of carbon is not very 

 Striking. Nevertheless, Rowland finds that 145 of the 

 .sciar lines between Lambda 3883.5 and Lambda 3864.0 

 agree with the component lines of the carbon flutings 

 fu the diagram I indicated only the most obvious of 

 these coincidences, and showed also that the break in 

 continuity at 3882 was due to the presence of a line of 

 robalt.— A. F.] 



iaottccs of Books. 



" The R.aces of Man ; an Outline of Antliropolojjy and Ethno- 

 graphy." By J. Deniker. Contemporary Science Series. (Walter 

 Scott, Limited.) A popular work on his favourite science by 

 such a distinguished anthropologist as Dr. Deniker, the Chief 

 Librarian of the Paris Museum of Natural Histoiy, can scarcely 

 fail to be a masterly treatise on the subject, and should merit the 

 best attention of his fellow-workers in this country, even if they 

 be disposed to doubt the correctness of some of his views. As its 

 title implies, the work treats of man not only from a zoological, 

 but likewise from a physiological and sociological standpoint, so 

 that it deals with anthropology in its most comprehensive sense, 

 the greater part of a chapter "being devoted to language and its 

 evolution. Xo less than 170 illustrations— some of full-page size- 

 are employed to convey an adequate idea of the leading types of 

 mankind, and the mode of dress (or " undress ") of the various 

 races. The care with which these illustrations have been selected, 

 and the excellence of their execution, wiU scarcely fail to be 



