February i4> 1895] 



NA TURE 



575 



irrto its constituent parts Trom ihe red to the violet; between 

 these colours we get that slrinij of orange yellow, green, and 

 blue which you are familiar wiili in the lainbow. For six|)ence 

 any of you may mike for your^rlves an instrument which will 

 serve many of the purpo-^es of demonstrating some of the more 

 beautiful fields of knowledge which have been opened up 



Fii.. I, — A simple form of spectroscope. 



t" as by its use. From an optician you can get a tiny 

 prism for sixpence; glue it at one end of a piece of 

 wood about 12 x i x A inch, so that you can see through 

 it a coloured image of a needle stuck in at the other end of the 

 piece of wood (Fig. i). This you must dj by looking sideways 

 tliroughit. Allowyour needle to be illuminated by a candle or a gas 



Fig. 2. — Star spectroscope, arrangcii for photograpliinij. attached to eye-end of reflecting telescope. 



Ihme, taking care that the direct light from the car.dle does 

 not fall upon the face of the prism ; you will then get a com- 

 plete band of colour from red to blue. If you go into the sun- 

 light — taking care again to protect the prism itself from the 

 entrance of any foreign light — and allow the sunbeam to 

 illuminate your needle, you get a spectrum of a different kind, 

 full of black lines. 



NO. 1320. VOL. 51] 



By such experiments as that, certain spectroscopic axioms 

 have been formulated : three of them are very impoitani. 



First, when solid or liquid or densely gaseous bodies are in- 

 candescent, they give out continuous spectra. 



Second, when a solid or lujuid body reduced to a slate of gas, 

 or any gas itself, is giving light, the spectrum consists of bright 

 lines, and these lines are different for different substances. 



Third, when light from a solid or liquid body passes through 

 i;as at a lower temperature, the gas absorbs those particular rays 

 ( f light r f which its own spectrum consists. 



We will next suppose, then, a spectroscope placed at the eye- 

 end cf a telescope (Fig. 2). The question put to the combined 

 instruments is : What is starlight like ? It was found that the 

 stars give a spectrum very much like the spectrum of the sun, in 

 most cases at all events, and that this spectrum could be defined 

 in the light cf the third axiom, that certain of the light was ab- 

 sorbed, there were dark lines in the spectrum (Fig. 3); and thus we 

 knew that light had been absorbed by an atmosphere surround- 

 ing something which was very much hotter than itself, and in 

 that way the science of solar and stellar physics was founded. 



Suppose another question put to this instrument: What is 

 the light of the nebula; like? 



I have already told you that Laplace held that in these bodies 

 we were dealing with gas at a high temperature. From the 

 time of Tycho Brahe down«ards, people had an idea that the 

 nef.uJK were "fiery." What should we expect to get in our 

 instiument? The second axiom tells us that, if we are dealing 

 vvith matter in a state of gas, or anything vapourous at very high 

 temperature, we shall get bright lines only. The question as to 

 the nebulae was put in 1864, and, curiously enough, when the 

 observation was made. Dr. Muggins remarked: "I suspected 

 some derangement of the instrument had taken place, for no 

 spectrum was seen, but only a line." "Only a line" was exactly 

 what I suppose Laplace would have given all he possessed to see, 

 if spectrum analysis had been invented in his day. That line 

 settled the question. There was certainly a tremendous spectro- 

 scopic difference between stars and nebulae, and this difference 

 has been emphasised by subsequent researches. (See 

 Fig. 3.) It is evident, therefore, that Lord Kosse's sus- 

 picion that the rebula; might, alter all, be found to be resolvable 

 into star clusters when greater op- 

 tical power was used, was proved 

 to be erroneous. 



Now we come to the second 

 point. I indicated in the previous 

 course of lectures that th re were 

 differences among the stars, de- 

 pending possibly upon chemical 

 constitution, or temperature, or 

 even upon their ages, and that the 

 stars had been classified by several 

 very diligent inquirers. Also, that 

 in all the classifications that had 

 been attempted, it was universally 

 taken for granted, for some reason 

 or other — possibly in view of the 

 idea of Laplace — ihat all the stars 

 in the heavens began in the con- 

 dition of highest temperature, and 

 that all that the stars did after that 

 was to spend their millions and 

 billions of years of life in getting 

 colder; so that, if we could at the 

 present moment find out which was 

 the very hottest star in the heavens, 

 wc might be perfectly certain that 

 every star in its beginning re- 

 sembled exactly in spectrum, and 

 therefore in physical constitution, 

 that particular star which we sup- 

 pose to be the hottest. It so 

 happened that in that very course 

 of lectures I pointed out, for the 

 first time, I think, in reference to the separation of stars 

 into classes, that such an idea as that would never do; 

 for if we form any conception of nebula; changing into stars, 

 we begin by knowing that the slats are veiy much denser than 

 the nebula;— taking the sun as an instance, the star practically 

 close to us— and that as the stars are denser than the nebulB, 

 they must be hotter than the nebulae, instead ol being colder. 



