August 29, 1895 J 



NATURE 



423 



We now know that this classification will not do, since, all 

 ufercncc to bright lines is omitted, and everyone now agrees 

 ih;it they must take the first place, and this is one of the great 

 I L-achings of the views I have been bringing forward for the last 

 ten years. 



- The idea which one arrives at by a discussion of all the 

 spectroscopic facts is that we begin with a condition in which 

 meteorites in swarms and streams are very far apart, and we get 

 from the collisions of these a spectrum which gives us bright 

 Hutings and lines, in other words the spectrum of the nebuhu ; 

 when they get a little more dense, we get the bright-line stars; 

 and as they get still more dense, we find the star with a mixture 

 of bright and dark flutings. Then we get still more condensa- 

 tion and dark lines, and then the highest temperature of all ; 

 I Her which begins a descent on the other side, till at last we end 

 HI cool, dark bodies like the earth and moon. 



This seems to be the classification which is necessitated by the 

 1 . .nsideration of all the facts, and it is, moreover, one which 

 Mcius to give us possibililiesof an explanation of the phenomena 

 il new stars and variable stars, and many other things without 

 j'ing into the region of the unknown and impossible. 



It also lands us in the so-called temperature curve along which 

 I \ entured to place the various classes of nebulx and stars some 

 nine ago. I am glad to say that so far no valid objection has 

 I R-en made to it. 



It will be noticed that in the classification I have suggested 

 I use the word "group," first employed by Rutherfurd ; itis one 

 which ought never to have been changed. 



With regard to this subject. Prof. Keeler, one of our most im- 

 portant authorities in this matter, agrees that a classification 

 which depends on this temperature curve certainly has ad- 

 vantages over other systems. He writes ' : — 

 . " Prof. Lockyer's system of stellar classification provides for 

 both an ascending and a descending branch of the temperature 

 curve, and in this respect it certainly has advantages over other 

 systems which claim to have a rational basis." 



I am also more glad than I can say that Prof. Pickering, who 

 has now given many years, with the aid of appliances beyond 

 all precedent, to the study of these questions, has arrived at 

 conclusions .strikingly similar to my own. 



In the first |ilace he includes the nebul.-e as well as the stars 

 in his system ; but it is right that I should add that he does not 

 commit himself to any statements relating to the relative tem- 

 jierature of the different groups, although he distinctly accepts 

 the idea of evolution, or what he terms an order of growth. 

 He writes {Astronomy and Astro-Physiis, 1893, p. 722) : 

 " In general, it may be stated that, with a few exceptions, all 

 the stars may be arranged in a sequence, beginning with the 

 planetary nebidic, passing through the bright-line stars to the 

 Orion stars, thence to the first type stars, and by insensible 

 changes to the second and third type stars. The evidence that 

 the same plan go\erns the construction of all parts of the visible 

 universe is thus conclusive." 



Prof. Pickering's results may be shown in tabular form, but 

 first it will be well to show the general differences between the 

 more recent classifications : — 



Nebub; 



Uright-linc stars 



Mixetl rtuting stars 



Dark line stars (ascending) 



Broad hydrogen stars 



Solar stars 



Carbon absorption stars... 



Sccctii. 



Vogel. 



Not clas- 

 sified. 



Typi 



HI. 



II. 



I. 



II. 



IV. 



Lockyer. 



Group I. 



„ II. 

 „ III. 

 „ IV. 

 „ V. 

 „ VI. 



In his classification. Prof. Pickering begins with the earliest 

 stages, taking the planetary nebuku and such nebula; as that of 

 Orion ; he tlien comes to the bright-line stars, and then to such 

 stars as those of Orion, and ultimately places the Sun, as I also 

 do, after the s|)ectrum of such a star as Sirius. There are prac- 

 tically two departures in his classification from that given by 

 myself. One is that what I call the bright and dark fluting 

 group of stars, represented by several of the red, and brightesr, 



1 Astronomy and AstrO'FhyiicSy 1894, p. 60. 



NO. I34S, VOL. 52] 



stars in the heavens, he m.akes cooler than the Sun. And the 

 class of stars which I group together and call Group VI., in which 

 we get mainly the absorption of carbon in the atmosphere, he 

 omits altogether, possibly for a very wise reason, as they are 

 certainly the most dihlicult stars to tackle ; but you see the diver- 

 gences in his classification from mine are small as compared with 

 those between Dr. \'ogel and myself, and he, I repeat, like my- 

 self, attributes the variation to an "order of growth." 



This premised, the difierences of sequence between Prof. 

 Pickering and myself may be Shown as follows : — 



TxK;kycr. 



I. 



II. 



III. 



IV. 



V. 



VI. 



Pickering. 



I. 

 III. 



V. 



II. 



Prof. Pickering, in the Draper Catalogue, combines like stars 

 under the different letters of the alphabet. The distribution of 

 these letters in relation to my Groups is as follows : — 



It will be seen that certain groups are represented by more 

 than one letter, but it is to be noted that here again Prof. 

 Pickering and myself have arrived at very nearly similar results, 

 for generally a different letter with him represents a sub-group 

 with me. This will be gathered from the subjoined table. 



Table showing the subdivisions of Groups III. and V. 



Group. 

 III. a 

 III. & 

 III. 7 



Pickering. 



H. 



I. (some Q. ) 

 B. 



V. a 

 V. 8 

 V.o- 



1-. 

 K. 



L. 



With regard to Prof Pickering, then, I have chiefly to justify 

 the place I have given to the stars of my Group H., which I place 

 after the nebuke and bright-line stars, and he jilaces after the Sun. 



I fancy that one of the reasons which has led Prof. Pickering 

 to this conclusion is to be found in the assumptiim that strong 

 indications of calcium and iron can only mark one stage of 

 growth, while I think it is certain they must mark two. 



We know they mark the present stage of the Sun's histor)*, 

 and taking meteoiites as we find them, a relatively low tempera- 

 ture would provide us with more calcium and iron vapours to 

 act as absorbers round each one than anything else. 



Now we have strong inilications of calcium and iron absorp- 

 tion in such stars as a llerculis as well as in the Sun, but the 

 general appearance of the spectra of these stars is so different 

 that both Secchi and Vogel have classified them ajMrt, and so 

 indeed does Prof. Pickering. 



B\it the reason that I classified these stars also in different 

 groups, and one on the rising and the other on the descending 

 arm of the temperature curve, was that in those like a llerculis 

 we have enormous variability as well as bright lines and flutings 

 indicative of sparse swarms, while in those like the Sun the pro- 

 duction of such phenomena is almost unthinkable. The special 

 variability of stars of my Group II. (Secchi's type III.) and the 

 production of bright lines at maximum is now freely acknow- 

 ledged. On this jKjint Prof. Pickering remarks ' : — 



" Long ])eriod variables in general are of the third type, and 

 have the hydrogen lines bright when near their maxima, as 

 stated above. This property has led to the discovery of more 



I Astrohomy and Astra-Physics, 1893, p. 721. 



