June 1, 1891.] 



KNOWLEDGE 



101 



^^ AN ILLUSTRATED ^»V [ 



MAGAZINE OF SCIENCE | 



SIMPLY WORDED— EXACTLY DESCRIBED 



LONDON: JUNE 1, 1891. 



CONTENTS. 



stellar Spectra. [Sopond Pa)ii'r.] By E. W. irvrNPEit, 



K.K.A.S. " . 101 



The Origin of the Chalk. By .Tonx T. Kemp, M.A.Cantab. 103 

 What is the Cause of Volcanic Action ? By the Rpt. II. N. 



llrTcniNsoN, B.A., F.G.S. . 101 



Astronomy as taught by Academy Pictures 107 

 The Travels and Life-History of a Fungus. I!y J. 



Tknthm. Smith, M..\., li.Sr., ,»(,■ 107 



Notices of Books 110 



The Coal-Sack Regions of the Milky Way. \\\ \. ('. 



R.VNV.Min .. ... ... ... ... Ill 



Flying Animals. By R. Lydukkeu, B.A.C'aiitab 112 



Letters :— J. Gwen Davidson ; W. T. Lynn ; W. F. 



Denning ; W. H. S. Monck 115 



The Face of the Sky for June. By Herbert Sadler, 



F.R.A.S 117 



Whist Column. By W. Montagu Gaitie, B.A.Oxon. IKS 



Choss Column. By C. D. Locock, B.A.Oson 119 



NOTICE. 



After this number, " Knowledge " will be published from 

 it.s own office, H2G, High Holborn, W.C. {near to the end of 

 ( '/iiinceiii Lime). Cheques and Postal Orders should be 

 made payable to Messrs. Witherhy & Co., at the above 

 address, to whom all communications with respect to 

 advertisements, subscriptions, and other business matters 

 should be addressed. 



STELLAR SPECTRA. 



[Secnnd I'ujier.] 



I'.Y K. W. MAiiNnuR, F.R.A.S. {Assisttnit siiperintendinij the 



SjKrtronrojiie Ikjniitiitent of (rreniukh Ohservatori/). 



THE last paragraph of my paper on this subject in the 

 April number of Knowledge was cut abruptly short, 

 and would bear further explanation, and the Editor 

 having courteously invited me to resume the subject, 

 I accordingly do so. 

 The method of comparing the mass-brightness of double 

 stars, which Mr. Monck devised (Ohsentitcri/, vol. x., p. 



96), depends upon the circumstance that the period of a 

 binary is not affected by its distance from us, and thorigli 

 its mass will vary according to the distance from us which 

 we assign to the star, its total brightness will vary in the 

 same proportion. Given then for two pairs of double 

 stars the apparent brightness, the apparent angular radii 

 of the relative orbits of the two components, and the 

 periods of each, and their relative mass-brightness — or to 

 use Prof. Young's more graphic phraseology, their relative 

 " candle power per ton " — can be computed. The formula 

 only breaks down in the improbable case of a pair in 

 which one star is at once very much fainter, and yet 

 nearly as heavy as the other. 



The following table shows the values of the relative 

 mass-brightness of the stars of the two types as given by 

 Mr. Gore in his " Catalogue of Binary Stars " : — 



The table shows two striking features. First, an 

 average superiority of the Sirian over the Solar stars, far 

 too marked and too frequent to be the effect of accident, 

 for whilst the mean mass-brightness of the First Type stars 

 is 12-00, that of the Second Type is not one-fifth so great, 

 being only 2-30. Indeed, one-third of the Sinan stars are 

 brighter per mass unit than the brightest Solar star, and 

 one-half the Solar are fainter than the faintest Sirian. So 

 that -we cannot resist the conclusion that, in proportion to 

 their light-giving power, stars resembling our own sun in 

 spectrum are considerably heavier, and if we assume a 

 uniform brilliancy for equal photospheiic areas, much 

 denser than those like Sirius or Vega. But -n'hen we take 

 this conclusion m connection with the one already arrived 

 at, that the total light-giving power of the average member 

 of the Second Type is greater than that of the First, the 

 superiority in mass of the Solar stars becomes yet further 

 enforced. If it be legitimate to ascribe the title " giant 

 suns " to any one of the classes into which we divide stars, 

 as distinguished by their spectra, then undoubtedly it must 

 be the Solar and not the Sirian stars which must bo so 

 designated. 



But the second conclusion to be drawn from the table 

 -would show that spectrum alone, at least as at present 

 classified, is insuHicieut to deterinino whether the mass- 

 brightness of a star is relatively high or low. Just as 

 with the table given on p. 7ci, sttirs of every degree of 

 absolute light-giving power were found in either type, so 

 it is now with mass- brightness. We could not then suppose 

 that any given dimensions, or any given total-light 

 radiation, were necessarily associated with a particular 

 form of spectrum ; and so now we are precluded from 

 supposing that any given degree of condensation corre- 

 sponds to the one type or the other. 



This conclusion has a more important bearing on the 

 question of the " .-\ge of Stars" than the former; using 

 the word " ago," of course, not to signify actual length of 



