Feb. 1 8, I 



NATURE 



573 



which Lockyer was the first to observe. It appears that 

 lines which, judging from terrestrial experiments, belong 

 to one element, sometimes appear in the solar spectrum 

 as distorted or displaced in dii?erent directions, indicating 

 perhaps that the substance producing the one line is 

 moving towards the eye with great velocity, while that 

 producing the other is moving in a contrary direction. It 

 is thus certain that the substances producing the two lines 

 must be in different places, and under these circumstances 

 we can hardly come to any other conclusion than that 

 these lines are given out by different molecular groupings 

 which have become separated from each other by the 

 sifting process already mentioned. 



From Fig. 26 it will be seen that different rates of solar 

 motion are exhibited by different iron lines. 



It would thus appear that a study of the solar spectrum 

 is likely to furnish us with much information regarding 

 the modes of vibration of molecular structures. It would 

 also seem that in view of these facts we should revise our 

 nomenclature. What, for instance, do we mean when we 

 say that iron occurs in the sun ? Clearly nothing more 

 than that certain molecular structures in the sun's atmo- 

 sphere are the same as certain terrestrial molecular 

 structures momentarily formed when we obtain the spec- 

 trum of iron. But if we could seize upon the various 

 particles that unite in giving out some one iron line, and 



put them into a bottle, we might perhaps find that they 

 were not iron, and they might even be different from the 

 thing obtained by treating some other iron line in the 

 same way. 



The following statement of Mr. Lockyer's views is 

 taken from the Report of the Solar Physics Committee : — 



" The view of the construction of the solar atmosphere 

 to which Mr. Lockyer has been led, may be stated as 

 follows : — If the atmosphere of the sun were quite 

 tranquil, and if we could see the spectrum of a section of 

 it, we should see it divided into an almost innumerable 

 number of layers, each with its appropriate spectrum. So 

 far from each substance (with some notable exceptions), 

 as determined by a spectral line, extending very far above 

 or below its normal position, it would be confined to one 

 heat-level, and the spectrum, taken as a whole, would get 

 simpler as we approach the photosphere from without. 

 The metallic elements, instead of existing as such in a 

 so-called ' reversing-layer,' resting on the photosphere, 

 are entirely broken up there, and their germs are distri- 

 buted throughout the atmosphere, the molecular groupings 

 getting more complex as the distance from the region of 

 greatest heat increases. The Fraunhofer spectrum, as 

 regards any one element, does not result from the vibra- 

 tion of the molecules of that element existing as such at 

 any given height in the sun's atmosphere, but results 



S / R J U S 



^wm^a^La^^ 



POL L U X 



Ot H E RCC L/ S . 



•,i ' J 



from an integration of the vibrations of the germs of that 

 element existing, perhaps distributed, from the top of the 

 atmosphere to the bottom." 



It may be said, however, that, while we have strong 

 evidence of a splitting up and also of a sorting or sifting 

 process of the various molecular structures going on in the 

 sun, we are yet without evidence that the molecular struc- 

 tures of two different elements can be split up into the 

 same component. The reply to this will be found in a 

 statement made by Mr. Lockyer {Proc. R.S., December 

 15, 1881) that the greater part of the lines seen at the 

 bottom of solar spots and in solar flames are lines ap- 

 parently common to two or more terrestrial substances 

 with the dispersion employed — in other words, basic 

 lines, as these have been termed. He has exhibited 

 some of these results in a diagram which will be found 

 in Nature, vol. xxiv. p. 323. Now, even if we imagine 

 that the coincidence in position of these lines, as derived 

 from two or more substances, is not absolute, yet the fact 

 that such lines form the greater portion of the spectra 

 belonging to the hotter regions of the sun is a very 

 significant one, and surely implies something more than 

 a mere chance correspondence. 



Rutherfurd and Padre Secchi were the first to attempt a 

 preliminary classification of the stars into groups, but for a 

 spectroscopic analysis of these bodies we are especially 

 indebted to Dr. Huggins. Lockyer thus sums up the 



information which we have thus obtained (December 12, 

 1878), the symbols indicating the metallic lines visible in 

 the various spectra : — 



Hottest Stars "^ (n + Ca-t-Mg 



Sun S i H-fCa-f Mg-fNa + Fe 



Coolest Stars •- / — _ Mg + Na-|-Fe + Bi + Hg. 



I may here remark that the hottest stars in the above 

 table are chosen because of their superior brilliancy, and 

 the cooler stars because of their inferior lustre. We are 

 thus entitled to say that the most brilliant and presumably 

 the hottest stars are those in the spectra of which the 

 prominent black lines are the lines of hydrogen, calcium, 

 and magnesium, while in those stars of which the sun is a 

 type we have in addition lines of sodium and of iron. In 

 the cooler stars hydrogen and calcium have disappeared, 

 and we have lines of magnesium, sodium, iron, bismuth, 

 and mercury, while in the coolest stars of all we have no 

 metallic lines but only fluted bands of metalloids. In 

 these last we may imagine that all the metallic lines hav 



