October, 1903.] 



KNOWLEDGE. 



^27 



ia the track I have made use of the " Nautical Almanac " 

 for 1860, as the eclipse of July 18 in that year is the 

 corresponding eclipse iu the 521 year cycle, and has a 

 similar track as regards latitude. 



It will of course bs understood that the tracks predicted 

 by the Megalosaros do not claim perfect accuracy ; there 

 is, in fact, an uncertainty of some 2.5° in longitude, (5' iu 



Diagram giving tlic Shift in Longitude and Latitude of Eclipse 

 Tracks in tbc Megalosaros. 



latitude. The cycle gives us, however, a close approxima- 

 tion with a miiiimnm of lal)3ur, aud enables us to select 

 the eclipses that are likely to prove interesting in any 

 given period. 



Thus the eclipse of 2-381 is total iu eeutral England at 

 lOh. Tim. a m. ( Rev. S. J. Johnson, " Eclipses and Tran- 

 sits in Future Years"), from which it appears that the 

 track, as predicted l)y the Meg., requires to lie shifted 

 westward aud southward through several degrees. 



It may be of interest to give an illustration of the 

 application of the Meg. to an eclipse in the remote past, 

 the eclipse of Ho and Hi iu China. Rev. S. J. Johnson 

 suggests, among others, 2126, October 13, as the date, aud 

 says, " on this date I titid au eclipse visible iu China did 

 o(-cur, but I have not looked into the circumstances of it." 

 The corresponding eclipse I Meg. later occurred, 321, 

 September 26. It was at tht; descending node, and 

 belonged to the long-auuularity class. The mean year is 

 — 12^4, aud from our diagram we olibiiu : —westward 

 shift, 170' ; southward shift, 8'. 



Whence we obtain the following, uslug Opi>olzer"s 

 T)osition for the later eclipse : — 



-311. -212l>. 



Snnvise point ... J07° West, 54° North ... 63' Kast, 02^ North. 

 No.m „ ... 36" „ 29° „ ... 134° „ .'57° „ 



Sunset 20° Kast, 5° «outh ... 170' West, 3° „ 



I have not inserted these tracks on the urap. which is 

 already sufficiently crowded, but it will be seen that the 

 track "of annularily in -2126 must have passed through 

 or near China, in accordance with Rev. S. J. Johnson's 

 statement. 



In couclusiou, 1 would e.vpress my great indebtedness to 

 Oppol/.er's great work the " Canou of Eclipses." The 

 amount of labour that its preparatiim must have involved 

 is well-nigh incredible, and the amount of assistance that 

 it gives in identifying auonent eclipses or studying the laws 



of their recurrence is immense. In fact it quite super- 

 sedes all cycles and other approxiuiate methods for the 

 period which it covers, and relegates their use to the 

 extremely distant past or future. 



But though the study of eclipse cycles has not now the 

 same imjwrtauce that it had to the ancient Chaldaeans, it 

 is stdl instructive and, I hope, not uninteresting, giving us 

 a clearer conception of the nature of the motion of the 

 sun and moou, and impressing us with a sense of the 

 majestic law and order which prevail throughout the 



universe. 



♦ 



THE CHEMISTRY OF THE STARS. 



V.-THE ORION STARS. 

 By A. Fowler, f.e.a.s. 

 The true character of the spectra of the Orion stars was 

 not realised until photographic methods were introduced. 

 Rutherfurd — one of the earliest observers of stellar 

 spectra — was unable to see any lines whatsoever, and this 

 also has been the experience of many subsequent observers, 

 even iu some cases where large instruments have been 

 employed. Nevertheless, Secchi, in 1863, published a 

 drawing of the spectrum of Rigel showing no less than teu 

 dark lines, among which the C and F lines of hydrogen 

 were conspicuous. Iu most of the Orion stars, how-ever, 

 the lines are so difficult of observation that even so 

 experienced an observer as Vogel described them iu 187-1 

 as " spectra in which the metallic liaes are few iu number, 

 and very faint or entirely imperceptible, and in which the 

 hydrogen lines are lacking." Later observers, including 

 Prof. Campbell, have been able to identify the D, line of 

 Helium as well as some of the lines of hydrogen, but the 

 fact remains that it is to photography that we owe 

 practically all our knowledge of this interesting and im- 

 portant group of stars. 



The photographic delineation of the spectra was for a 

 time far ahead of their interpretation. Up to 1895, the 

 lines of hydrogen, two lines of calcium, and a solitary line 

 of magnesium (X 4481) were the only ones recognised as 

 having anything iu common with terrestrial chemistry. 

 One important relation, however, had been arrived at, 

 namely, that some of the lin^s were identical with lines 

 associated with Dj in the spectrum of the solar chromo- 

 sphere, aud therefore presumably originated in the then 

 hypothetical gas, helium. Suspicion was converted into 

 certainty on the discovery of terrestrial helium by Sir 

 William Ramsay in 1895. It was then shown by Prof. 

 V^ogel, Sir Norman Lockyer, and others, that most of the 

 stronger Orion lines, always e.xcepting those of hydrogen, 

 corresponded in every particular with the lines of the 

 newly-discovered gas, and thus received a completely 

 satisfactory chemical expl uiation. The lines of helium are, 

 iu fact, the most characteristic feature of the Orion stars. 



The mystery still surrounding other important lines 

 was shortly afterwards dissipated by Sir Norman Lockyer's 

 identification of some of them with lines of silicium aud 

 carbon, and by Dr. McCleau's remarkal>le discovery that 

 several lines were due to oxygen. Not less uuexix'cted 

 was the subsequent discovery that nitrogen was responsible 

 for some of the remaining lines. The lines not yet 

 accounted for are mostly feeble, and their eventual identi- 

 fication with lines of terrestrial substances may be reason- 

 ably expected. 



The Ori<.m stars, however, are not by any means all of 

 the same pattern. Some of the lines are of different 

 intensities iu different stars, and lines found iu one may 

 even be absent from another. Signs are not wanting that 

 these dift'ereuces are due to the different stages which the 



