584 



NA TURE 



[April 22, 1886 



representative), are in the initial stage of their life as suns. 

 Their energy is still unwasted ; their temperature is 

 enormously high ; their light is not sensibly modified by 

 absorption, hydrogen being the only constituent of their 

 atmospheres capable of strongly intercepting their radia- 

 tions. But with the lapse of ages, this early fervour cools 

 down, and absorption gains strength. Hydrogen no 

 longer stamps itself predominantly upon their spectra ; 

 metallic rays deepen and multiply ; a dusky veil is drawn 

 across each photosphere, stopping preferentially its more 

 refrangible emissions, and thus imparting a yellowish 

 tinge to the resulting light. The condition of our sun, as 

 well as of Capella, Pollux, and Dubhe, is, in short, 

 reached. Down to this point the history of all ordinary 

 stars is the same. Here, however, a bifurcation in the 

 path of development is reached. Two roads to extinction 

 are now open to them. For, according to Vogel, the two 

 varieties of banded spectra mark co-ordinate, not suc- 

 sessive, stages in stellar existence. The choice, so to 

 speak, once made, is definitive. Migration from one type 

 to the other is impossible. Hence Vogel's abolition of 

 Secchi's fourth type, and his distribution of such stars as 

 Betelgeux and a Herculis on the one hand, and 152 Schjel- 

 Icrup and 19 Piscium on the other, into two alternative 

 branches of his third. But let us look a little more 

 closely at facts before admitting conjecture. 



M. Duner's Catalogue includes 297 entries under the 

 heading Class 111. a (type of a Orionis), to which in all 

 475 stars are so far known to belong. A particular de- 

 scription of each spectrum, from his own and others' ob- 

 servations, is appended ; so that ample materials are 

 provided for some few safe generalisations. 



The first point to be noted is that the positions of the 

 leading bands in a/i spectra of this kind are absolutely 

 unchanging. The series is repeated with varying degrees 

 of intensity from star to star, almost as if in stereo- 

 type. The shadings are, it would seem, in reality made 

 up of fine lines very closely grouped. D'Arrest and 

 Huggins, at least, repeatedly succeeded in thus resolving 

 them, although to Vogel, even when employing most 

 powerful optical means, they persistently maintained a 

 nebulous appearance. Now a glance at the accompanying 

 figures will show a symmetry in the arrangement of these 

 bands suggesting that they result from the rhythmical 

 vibrations of one highly complex molecular system. In 

 other words, they betray the absorptive action of a single 

 substance ; particular identification is awaited ; nor is it 

 easily attainable. Great difficulty attends inciuiries into 

 the direct spectra of compound bodies, since the very 

 means employed to render them luminous, also tend to 

 destroy, by forcing them asunder into their constituent 

 elements. 



Besides this unknown substance, however, metallic 

 vapours exist abundantly in the atmospheres of Betelgeux 

 and its congeners. The grooved spectrum distinguishing 

 them might in fact be regarded as superposed upon a 

 modified Fraunhofer spectrum. Not only in its bright 

 spaces, but even across its dusky flutings, a crowd 

 of significant dark rays can be perceived. Their 

 number, as disclosed by the 27-inch Vienna refractor 

 in September, 1884, in the spectra of /3 Pegasi and 

 a Herculis, took Vogel altogether by surprise [Piiblicatio- 

 nen, Potsdam, No. 14, p. 22). Yet he and Dr. Huggins 

 had already measured no less than 95 such in the analysed 

 light of Betelgeux. Some of these can be identified with 

 terrestrial substances. Sodium, iron, magnesium, cal- 

 cium, and bismuth, ara without doubt incandescent above 

 the photosphere of that star. Lines of hydrogen have 

 also been made out, and its presence is certified by Dr. 

 Huggins's photographs. Its absorption is, however, in- 

 conspicuous in all, and imperceptible in most spectra of 

 this description. 



One of their most singular features, as yet unexplained, 

 is that dark metallic rays form the sharp boundary of 



many of the flutings. Thus calcium-lines (wave-lengths 

 6i6'4 and 585'6) respectively terminate, on their more 

 refrangible sides, the bands numbered 2 and 3 in the 

 figure ; strong contiguous lines of calcium and iron limit 

 band 4 ; band 5 ends with the well-marked iron lines of 

 wave-lengths 54S'o and 544-4, and band 8 with that of 

 495'S ; band 7 with the solar group I/; band 9 with a 

 deep furrow of unknown origin. These coincidences are 

 extremely puzzling ; for, as M. Duner remarks, they can 

 scarcely be accidental. 



Stars with fluted spectra are all more or less deeply 

 tinted with orange, owing to the stoppage, by a general 

 absorption, of by far the greater part of their blue rays. 

 Their actual emissions must then be very greatly in excess 

 of those reaching outer space. Stripped of its surrounding 

 atmosphere, our sun, it is computed, would leap up to 

 some three or four times its present lustre ; but in stars 

 like Betelgeux, absorption must at least c(ua Iruple its 

 solar effects. This consideration is of fundamental im- 

 portance in any estimate of the relative luminous power 

 of the stars. 



Fifty-five members of Class III. b find a place in the 

 Lund Catalogue. These are all that have hitherto been 

 discovered. Yet exploration, in their case, is more com- 

 plete than with the previous type, the broad, deep zones 

 of their spectra being distinguishable in objects much too 

 faint to show the narrower groovings of Class III. a. No 

 star of this kind is as bright as the fifth magnitude, while 

 eight between ninth and tenth are included in M. Dundr's 

 list. Thus, although the fluted spectra already examined 

 outnumber those in zones (as we may call them for the 

 sake of distinction) scarcely nine times, JVl. Dundr con- 

 siders that the real proportion of their excess is at least 

 fifty to one. 



The rare objects constituting Class III. b are amongst 

 the most interesting in the heavens. For they exhibit in 

 their spectra the unmistakable signature of that substance 

 which, more than any other, deserves to be called the 

 material basis of life. Father Secchi (their original dis- 

 coverer) regarded them from the first as " carbon-stars ; " 

 but Dr. Huggins in 1872 tested the supposition (for it 

 was then little more), and rejected it as disproved. There 

 is now no doubt that the Roman observer was in the right. 

 The three conspicuous bands of dense absorption visible 

 in such spectra agree in position quite closely with the 

 emissions of carbon-vapour glowing in the electric arc. 

 Dr. Huggins gives scanty details of his observation (see 

 Schellen's "Spectrum Analysis," ed. 1S72, p. 504) ; he is 

 rarely in error ; but on this occasion was perhaps misled 

 by the facile emergence of acetylene-bands, of which the 

 blue one falls just in the intermediate position indicated 

 by him as fatal to the suggested identity. 



Besides carbon, sodium is without doubt present in the 

 atmospheres of these remarkable bodies ; and there are 

 signs of further metallic absorption, notably by iron. 

 Their rays being, however, too faint to bear scrutiny with 

 a narrow slit, the finer features of their spectra remain, 

 for the present, unrecognisable. Yet we cannot avoid 

 being struck with the circumstance that their most pro- 

 minent constituent elements are precisely those which 

 kindled in the great comet of 1882 as it approached the 

 sun. 



The " zoned," like the " fluted " stellar spectrum, is, in 

 general outline, invariable, though capable of endless 

 individual modifications of tone and detail. It is as if 

 one fundamental sketch-plan were filled in with the most di- 

 verse depths of shading.' Another point on which all such 

 stars agree, is the redness of their light. The violet end 

 of their spectra is, uniformly, all but obliterated ; not 

 necessarily through original deficiency. The more re- 

 frangible emissions of 152 Schjellerup may, for aught we 



' See Figs. 4, 5, and 6 : the last a supposed example of a "transition" 

 spectrum given by a star in course of passing from Class II. rt (solar typt) 10 

 Class IIi;^. 



