Mabch 1, 1900.] 



KNOWLEDGE 



65^ 



impossible to soo the coiucidcuce as a whole ? That is, 

 is it only bv the coinpai'isou of each line, ouc aftei" 

 another, with its couuterpart iu the solar spectrum, 

 that the assurance is obtained that the element's 

 spectrum is reallv contained iu and forms part of the 

 solar spectrum '. 



Devonport, E. E. Makkwick, Uol. 



14th January, 1900. 



[Unless Col. Markwick has written " dark " when 

 intending to write " bright " in the sentence " none of 

 the other (dark) lines in (1) seem to have their counter- 

 pai-t iu (2), " I fear he is labouring under a slight 

 misapprehension. In the case of carbon, as for other 

 substances, its identification depends upon the coin- 

 cidence of bright lines obtained terrestrially with the 

 dark ones of the solar spectrum. Viewed iu this way, 

 a considerable number of the bright lines which build 

 up the carbon fluting, commencing at Lambda 3883, are 

 seen to correspond with dark ones in the sun, and the 

 presence of carbon among the solar elements is cleai-ly 

 demonstrated bv the photograph. The apparent darK 

 lines between the bright members of the carbon flutings 

 are simply dark interspaces and naturally do not agree 

 with solar lines. 



I may add that photogra])hic demonstrations of the 

 presence of many chemical elements in the sun arc just 

 as convincing, and certainly more satisfactory, than the 

 diagram to which Col. Markwick refers, but believing 

 these to be sufficiently well known, I purposely selected 

 an illustration of the detection of a substance requiring 

 more careful investigation. In the ease of carbon the 

 component lines are crowded together and in the sun 

 are superposed upon lines belonging to other substances, 

 so that the correspondence is not so obvious as in 

 many other illustrations which might have been 

 given. — A. Fowler.] 



OBSERVATIONS OF VARIABLE STARS. 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — My estimates of the movement of S. S. Cygni 

 at Its last appearance are as follows, viz. : — • 



Mag. 



... .S-6o 



... 8-8.J 



... 8 85 



Nov. 1 & 2. Cloudy. 



„ ' ( 10(>0 or 



•■ **• ~" ■[ less. 



1899. H. Miifr. 1899. H. 



Oct. 21. — ...Invisible. Oct. 30. 7.45 p.ni 



„ 25. 7.30 p 111. . 805 „ ;jl. 7.45 „ 



„ 26. 7.30 8 35 „ 31. 9.00 „ 



., 23. 7.00 ,. ... 835 



., 28. 8.00 „ ... 8-35 



., 29. 7.30 „ ... 855 



Since August 11 I have seventy-one observations of 

 S. XJ. Cygni (Muller and Kempf's variable), period 

 3d. 20h. 15m. 21s., varying from 6m. 57 to 7m. 37, an 1 

 though at first apparently irregular, experience has 

 shown it to be an evenly moving star, within the range 

 of the field glass at all phases, and an attractive 

 object for amateurs. At the minimum phase it appears 

 on many occasions to remain at the same light for 

 twenty-four hours, and then rise rapidly to maximum. 

 The fact is that in its near approach to minimum and 

 rise from that phase it takes twenty-four hours to fall 

 and to rise, or to change, less than one step 0.1. That 

 is to say when the star reaches about 7m. 34 it takes 

 twenty-four hours to fall to 7m. 37 and rise again to 

 7m. 34. So one who has no photometer may announce 

 a minimum some time before that j'hase is reached, or 

 after the star has passed it. This is shown by Prof. 

 Pickering's curve. There have been similar occuiTcnces 

 at maximum more than once, but they have not been ob- 

 served so often nor have thev been so marked. 



S. Virginis was an uubieady star at its last appear- 

 ance; its changes were estimated as follows: — 



Mill,-. 



;iu-oo 



10 (HI 

 9-10 

 'J 23 

 ',) 10 

 8-50 

 8-35 

 8-30 

 8 13 

 800 

 7 90 

 7 70 

 7 9J 

 7-85 

 7-75 



l'<9fl. 



Jiilv 8. 



„ ' 9. 



„ 10,11. 



„ 12, 14, 1.-), 17. 



„ 22, 23 maximiiiii. 



„ 29, 30, 31. 

 Aup. 1. 



•J 



" 4^ 5. '.'.'. 



6. 

 „ 9, lo. 

 „ 11. 

 „ 20,21. 



„ rs. 



Miiir. 

 7-60 

 7-40 

 7-23 

 G95 

 »)-80 

 7 03 

 7 15 

 713 

 725 

 7 35 

 7-30 

 7-90 

 7-80 

 7 87 

 805 



2G, so 



The computed date of maximum was June 

 the star was 26 days late. The weather during nearly 

 the whole of the summer period was not favourable for 

 observations in the southern skies, and so R. Hydrse 

 being further south was more difficult of observation. 

 The best I could do is appended : — 



The maximum was due August 8, and compared willi 

 former ajipearances is jjrobably late again. 



D.wiD Flanehy. 

 Memphis, Tenn.. U.S.A., 



• 18 November, 1899. 



IS THE STELLAR UNIVERSE FINITE ? 



TO THE EDITORS OF KNOWLEDGE. 



Sirs, — I have followed lately with great enjoyment, 

 the interesting discussion on this most interesting 

 subject. There is, however, one mo.st fundamental point 

 which has not yet been elucidated, and in connection 

 with this I beg leave to say a few words. 



Mr. Burns stated and Mr. Maunder and Mr. Hill 

 have challenged the statement that — " If the number 

 of stars were infinite, we should have the whole skv a 

 blaze of light.' The unqualified statement is certainly 

 a bold assumption, and a little careful consideration, 

 combined with a little mathematical analysis, will soon 

 show to what extent the assumption is justifiable. 



The problem resolves it.self into the integration or 

 summing up of an indefinitely great number of indefi- 

 nitely small quantities. Will the yield be necessarily 

 indefinitely great? May it not bo a finite quantity? 



There is a kind of struggle going on between the 

 greatness of the number and their sniallness in size. 



To take an actual example, if wc consider a 

 series of quantities, one, a half, a quarter, an eighth, 

 and so on, each term becomes smaller and smaller, 

 and oven if we take an infinite Dumber of tenns 

 of this series and add them together the total 

 cannot exceed two. In this case the infinite small- 

 ncss of the terms has so counteracted the infinite 

 greatness of their number that the total yield 

 is a finite quantity. Now may it not be likewise in 

 the case of an infinite stellar universe? May not the 

 infinite number of the stars be so counteracted by their 

 infinite smallness in apparent size, that, no matter 



