72 



KNOWLEDGE 



[April 1, 1891. 



the iis|iritivc ages in the evolution of a normal star which 

 the diUereut types might represent. For it was naturally 

 felt that there must be growth and development amongst 

 stars as amongst animals and plants. A star must have 

 its birth, its periods of growth, full vigour, and decay, 

 closing in the — 



Last scene of all 

 That ends this strange, eventfnl history, 



a dark, cold body, incapable of giving light and heat to 

 other worlds, or of sustaining life upon its own surface. 

 And naturally the attempt was made to connect these 

 different types of spectrum with the various stages of 

 stellar life-history. Zollner seems to have been the first 

 to suggest that the white stars were the hottest, and that 

 the yellow colour of the stars of the second type showed 

 that they had advanced some way in the process of cooling 

 down. Angstrom suggested that the shaded bands of the 

 orange stars indicated the formation of compoimd bodies 

 in their atmospheres, consequent on the lowering of tem- 

 perature ; and Lockyer, a little later, expressed the same 

 views still more definitely. " The hotter the star, the 

 simpler its spectrum," he said ; " and the older a star, 

 the more does the free hydrogen disappear from it." 

 Vogel, more recently, being dissatisfied with Secchis mode 

 of grouping stellar spectra, devised a more elaborate one, 

 the principal change from Secchi's system of Types being 

 that \'ogel makes Types 111. and IV. varieties of the same 

 Class 111. The leading thought in this new classification 

 was that all stars have at one time or another been of the 

 first or Sirian Type, all will pass through the second or 

 Solar Type ; but after passing this stage the star may either 

 show a spectrum hke that of a Herculis, or like that of 

 the small red stars. The road diverges here ; the majority 

 follow the path which Antares, Betelgeuse, and the btcidn 

 of Hercules have taken ; but a few, especially the stars in 

 two small groups on the Jlilky Way, prefer to gi\e spectra 

 in which the bands are sharpest to the red, and shade off 

 towards the violet. 



It was well objected to this scheme that it made no 

 pro\-isiou for a period of increasing temperature in a star. 

 Stars are certainly not brought forth from nebulie, like 

 Pallas Athene from Jupiter, fully developed and equipped 

 with their whole armoury of light and heat. There is 

 evidently a time during which the surface brilliancy in- 

 creases, and this probably corresponds with a time during 

 which the mean temperature must be increasing. Besides, 

 the place assigned to the red stars was a perfectly arbitrary 

 one. We know that the First and Second Types are con- 

 tiguous stages, whichever be the earlier ; for stars hke 

 Procyon, Rigel, Spica, Polaris, a Cygni, and others, supply 

 examples of almost every possible gradation, from the 

 unmistakable Sirian to the complete Solar form. So, 

 again, Aldebarau, Betelgeuse. and many others form a 

 series of links connecting the Second and Third Types ; 

 but until recently the Fourth Type stood alone. There 

 are no instances of stars whose spectra leave us in doubt 

 as to whether the Fourth Type or some other is the more 

 strongly represented ; there are no intermediate forms. 

 It is therefore a pure assumption to assert that this pai-- 

 ticular kind of spectrum is either the next stage to the 

 Solar Type or that it is an alternative stage to the 

 a Herculis Type. 



The most recent classification is that of Lockyer, and 

 he certainly avoids one of the objections to which Vogel's 

 is open. The Sirian Type is still taken as that of the 

 hottest stars, but he breaks up the Solar Type into two 

 groups ; the one showing rising, and the other falling 

 temperature. The course of evolution, according to his 

 plan, is : Group I., the nebular stage ; Group 11., the 



orange star, or a Herculis stage ; Group HI., the a Cygni 

 stage, including some spectra of Secchi's Type 11. ; Group 

 IV., the Sirian stage ; Group V., the Solar stage, including 

 the rest of Secchi's Type II. stars ; Group VI., the red 

 star stage; Group VII., the dark stage. The position 

 assigned to the red stars is, of course, as much a matter 

 of assumption on Lockyer's plan as on Vogel's. The 

 former has, however, this advantage over his predecessors, 

 in that it corresponds to the probable stages of growth as 

 well as those of decay. The Sirian stars, which he, in 

 common with almost every theorist," regards as the 

 hottest and largest, mark not only the commencement of 

 a fall in temperature, but the conclusion of a rise. We 

 are saved the difficulty of assuming the Sirian phase to be 

 that in which all stars were originally created, and from 

 which they have only changed by degradation. 



But in all these systems the supremacy of the Sirian 

 Type is assumed, not proved, and it has been well pointed 

 out that it is quite possible to read the record the other 

 way, and to argue that the ruddy stars are the hottest, 

 and that the orange, yellow, and bluish tints are o\-idences 

 of a progressive decline in temperature. It is a question, 

 therefore, of much importance to see if any positive 

 information can be given us on the subject. 



Two circumstances have greatly operated to the present 

 view. Firstly, that a solid body raised to incandescence 

 first glows with a ruddy hue, and then, as the tempera- 

 ture increases, so the colour changes to orange, yellow, 

 white, or blue ; and it was very natural, though scarcely 

 scientific, to extend the analogy to the stars. For a stellar 

 spectrum shows us by the continuous band of colour, in- 

 terrupted by dark bands, that the light we receive is not 

 the whole of the Ught the star emits, but that it has 

 sufl'ered absorption in the atmosphere of the star itself ; 

 and it is m the difi'erences of the quality and amount of 

 this absorption, and not in the differences of the original 

 light of the star, that we find the tests for distinguishing 

 one type from another. That is to say, it is by the dark 

 lines or bands, and not by the continuous spectrum, that 

 we classify the stars. The difference, therefore, between 

 the Sirian and Solar stars lies not in their photospheres, 

 but in their absorbing atmospheres. For Sirius such ab- 

 sorption is almost confined to hydrogen, the influence of 

 which is excessively marked ; but for the Sun, Arcturus, 

 and their congeners, twenty or thirty elements have im- 

 pressed the spectrum with the evidences of their presence. 

 Given that two stars of these two Types are at the same 

 distance from us, and that they appear to shine with the 

 same amount of light, surely the star which displays the 



* In this, as in other departments of science, the inductive method 

 is the only safe one, and a few observed facts are preferable to any 

 number of theories founded on assumed conditions. The physical 

 connection of the trapezium stars with the Orion nebula, and the 

 stars of the Pleiades cluster with the Pleiades nebula, can hardly be 

 doubted : and if the nibular stage is the first one in a star's history, we 

 have evidence that three diflferent classes of spectra are exhibited by 

 stars involved in nebulous matter. In all three the hydrogen lines, 

 or some of them, are conspicuous, but the spectra are considerably 

 more complicated than those of stars usu.ally ranked as belonging to 

 Secchi's First Type. The spectra of the trapezium stars appear to 

 be crossed with bright lines similar to. but more intense than, the 

 bright lines of the Orion nebula around them ; the hydrogen lines are 

 not hazy and diffused as in Secchi's First Type, but comparatively 

 narrow and sharp. The two classes of spectra in the Pleiades gi-oup 

 each include stars of various magnitudes ; and if we may assume 

 that all the stars of the Pleiades group are of the same age, we must 

 conclude thiit the two classes of spectra do not correspond to different 

 stages of cooling. The star which appeared in the Andromeda nebula 

 in 1885 seems to have had a spectrum terminating abruptly at the 

 red end, as well as, according to some observers, very faint bright 

 lines, which are asserted to exist in the Andromeda nebula spectrum. 

 The ifacts brought together by Mr. Maunder in this paper should check, 

 too precipitate theorists. — A, C. Ranyaiu>. 



