432 



NATURE 



[April 2, 1874 



therefore still nearer the sun there is an enormous quantity of 

 vapours of the different elements existing in the sun, in what we 

 may term a reversing layer, and it is to the absorption of tlie 

 elements in this layer that the absorption of the sunlight, and 

 therefore, so to speak, the creation of the spectrum of the sun, is 

 in the main due. 



I will now direct your attention to two photographs of the solar 

 spectrum, and reminding you that the complexity of a spectrum 

 depends upon the number of elements, and upon the pressure at 

 which the vapours of tliose elements exist in the atmosphere of 

 any star, you will gather from these photographs a pretty good 

 idea of the extreme complexity of the sun's reversing layer to 

 which I have referred. 



In a photograph of any part of the sun's spectrum each of the 

 lines of course has its story to tell, not only so lar as the substance 

 which is at work doing that particular part of the absorption 

 is concerned, but also even so far as the quantity of that sab- 

 stance is concerned ; because not only will a certain substance 

 absorb particular waves of light, but it will absorb many waves, 

 or few, or none at all, according to the quantity of that particular 

 substance in the envelope surrounding the sun. Now there 

 is a great deal of calcium in the sun, and therefore the absorption 

 lines of the calcium are very thick, the absorption lines of the 

 other metals which do not exist in such great quantities in the sun 

 being very much thinner. 



Having brought before you these various points connected 

 mainly with the sun, so far as its physical constitution goes, let 

 us consider what is the chemical constitution of the sun. 



I have already told you that surrounding the sun is an enve- 

 lope composed in the main of hydrogen, and of a new element, 

 and that nearer to the sun is a region of vapours of great com- 

 plexity, containing at leist one new element. This region con- 

 tarns, besides hydrogen, and dealing with known elements, 

 magnesium, sodium, titanium, calcium, nickel, chromium, iron, 

 manganese, aluminum, copper, zinc, barium, cobalt, and so on, 

 and latterly we have had reason to suppose that some six or 

 seven new elements must be added to the list — potassium, lead, 

 cerium, ur.anium, strontium, and cadmium. Further, if instead 

 of the new " atomic weights " of the elements we take the old 

 " combining weight " we find that the arrangement of these 

 layers round the sun follows the vapour densities of the various 

 substances either absolutely or very closely. 



This then is the verdict of the prism with regard to the chemi- 

 cal constitution of the sun, the nearest star that we can get at ; 

 and I think you will acknowledge that if the prism had done 

 nothing else it would have done good work. But 1 think it has 

 done very much more, becMse it has enabled us not only to 

 chronicle those things as existing in the sun, but in connection 

 with the other facts which I have already brought before you it 

 has enabled us to place the sun in its proper pasition amongst 

 the stars. For instance, I have already called your attention to 

 the first, second, and third classes of stars. Is Ihe sun in the 

 first, the second, or the third class ? Does its spectrum contain few 

 or many lines ? or are there channelled spaces or bands ? Its spec- 

 tn;m is not excessively simple ; there are no channelled spaces 

 or bands ; and thsrelore the sun is to be placed in the second 

 class of stars. Can we then go beyond this chronicle of iacts, 

 which I am afraid some of you may have considered rather 

 dry? 



Vou know that what a scientific man has to do in any reseirch 

 is not merely to add fact to fact, and to go blindly looking after 

 facts irrespective of order. What he has to do after he has ac- 

 cumulated a certain number of facts is, to try whether it is pos- 

 sible to arrange them in order. If you wish to get a law out of 

 any accumulation of facts in physics, in chemistry, or astronomy, 

 you must first get your facts into order or you will never do it. 

 Is there any possible order into which we can group these various 

 Iacts to which I have referred ? I venture to think there is. 



Call to mindthe three clas^es of stars. Is there any other physical 

 quality tacked on to tho^e differences ? Yes. The stars with the 

 simplest specua are on the whole the brightest stars in the heavens; 

 and the channelled spaced stars are on the whole the dimmest 

 stars in the heavens. Of about 500 stars which have been already 

 observed, over 300 are of the complicated second order or type. 

 There are a great many bright stars of the first order, but an 

 extremely small number, only, I think, about 27 of the third 

 order with the channelled space spectrum. Now, if this be true, 

 and if it be fair to assume that the star which is the brightest is 

 on the whole the hottest, and I think it is fair to say so, if you 

 take all other things as equal, then yoti come to a generalisation 



of this kind, that the brightest and hottest stars in the heavens 

 have the simplest spectra, and the dimmest, reddest, coldest stars 

 have a spectrum entirely different. If this be so, cm we connect 

 these facts ? I think so. Grant these facts (and the future alone 

 will show whether they are facts), and the thing; is clear. We 

 may group them all together by supposing that in the stars of the 

 first and second classes there are dissociating forces at work 

 which, from considerations which I have not time to bring before 

 you now, we can imrtgine to be infinitely higher, or at least con- 

 siderably higher, than any dissociating force that we can get 

 here even with the electric spatk. if you imagine in these 

 stars an atom-severing force greater than we can obtain here, you 

 can at once group in a working hypothesis all the facts which I 

 have brought before you, and in a simple way ; and let me add 

 that simplicity in Science is a very great evidence of truth. 



If you assume that at the highest possible temperature — here 

 I use the word temperature for want of a belter — of a star 

 you have work done in continuation of the work done in terres- 

 trial furnaces that is to say, if instead of having 63 elements 

 which we have here with our furnaces, there is a mucii smaller 

 number, taking into consideration the increase of temperature, 

 yon will see at once that the brightest and hottest star in the 

 heavens should have the simpleit, spectrum, because there you 

 have the fewest elements, and that the coldest, reddest, dimmest 

 star should have a spectrunr indicating metalloids and com- 

 pounds because you have there a low temperature, at which the 

 metals do not exist except in combination. And if you think 

 this matter over yoa will see that this suggestion of a 

 higher temperature giving us a simpler c-nduion ol .v hat we with 

 our paltry temperature^ call c io,iiio-il clcrncnis, ins'.ea ! of making 



FiG S.— Tlie Coro 



I (Inali 



Eclipse, 1871). 



these stellar spectra complicated, difficult to recoilect and to 

 understand, puts them all in one line easily to be grasped, and a 

 line which I venture to think is somewhat coincident with the 

 probabilities of the case. 



Does the work stop there ? Has it nothing to say to comets 

 and meteorites ? Here again it has a question to ask. The 

 beautiful hypothesis of Schiaparelli, which is accepted by all 

 astronomers, has made it perfectly clear that a comet is nothing 

 more than one of an infinite number of meteors or meteorites, 

 or whatever you like to call them, travelling in cometary orbits 

 round the sun, and to which the showers of shooting stars are 

 due. We know that there is a comet connected with that parti- 

 cular meteor sw.irm which gives us the November meteors, and 

 we know that there is another comet connected with that parti- 

 cular meteor sw.irm which gives us the August meteors ; and we 

 assume that in all probability there .are millions, or any enormous 

 number that you like, of meteors, or meteorites, or shooting stars 

 peopling a part or the whole length of that concrete orbit so to 

 speak. 



How is it then that there is only one comet amongst all that 

 infinite number of potential meteors or meteorites ? Here again 

 I am sure that the future will enable the prism to throw an im- 

 mense light, and we already have a glimmering. If, for inst.ance, 

 you assume th.rt out of a star of the second clas.s, in the reversing 

 layer of which there are no melalloids, portions of the atmosphere 

 are, by forces which we know to be at work, thrown boililyfrom 

 the sphere of the star's attraction into space, those vapours on 

 being cooled would give us very much the same kind of chemical 

 composition that we get in the well-known masses called iron 



