March 1 8, 1886J 



NA TV RE 



47i 



chemistry, to make a more or less detailed examination of the 

 materials — of the various chemical substances — in them. 



The Chemistry of the Spots 



Now from what has been already said we at once see that if 

 we get a downrush of solid bodies, such as cooled iron masses 

 which are derived from the condensation of the iron vapour, 

 and which we may regard as solar meteorites, from the 

 top of the sun's atmosphere down to the photosphere, we 

 shall have, if that quantity be great enough in the spot region, 

 a considerable dimming of the sun's light, by reason of the 

 fact that we have an infinite number, or a very large number, of 

 solid bodies stopping the light from that particular part of the 

 sun. We shall have in that way then a continuous absorption of 

 the Sim's light ; that is to say, the red light, the yellow light, the 

 green light, the blue light, and so on, will all be more or less 

 stopped, and that part of the sun will, from that cause alone, 

 look dimmer. Further, if we have any vapour apf roximating in 

 molecular structure to the chlorine vapour to which I drew your 

 attention before, we shall have an absorption of another kind ; 

 we shall have a special stopping of the blue light of the sun, 

 making, therefore, the sunlight yellower than it otherwise would 

 be. Again, if we have other substances associated with these 

 spots in the state of fine vapour in a state of incandescence, we 

 shall have such absorption indicated by the darkening of certain 

 of the Fraunhofer lines, and by the widening of them as well 

 as the darkening of them if the quantity of any particular vapour 

 is considerable, and we shall also get new lines if new substances 

 are produced or their existence revealed by these conditions. 



Now, I propose with regard to this point, seeing that our 

 time is limited, to confine myself to the observations that have 

 been made on the sun's spots at South Kensington since the year 

 1879. From that period some 700 spots have been observed. 

 Of that number some 200 have been discussed ; that is to 

 say, maps have been made, and we have endeavoured to 

 see along which line it was best to push on the work. But 

 this climate of ours is not aparticularly good one for observations 

 of this kind, especially in London, because there is a good deal 

 of smoke, and it is almost as difficult to Fee through a smoke 

 cloud as it is through one of the ordinary kind. Still, in that time 

 a considerable number of observations have been made altogether; 

 and, to guarantee us as far as possible against bad climatic con- 

 ditions, what has been done in the case of each spot has not been 

 to observe all the lines which are specially affected, but to con- 

 tent ourselves with getting the results with regard to twelve lines, 

 six in the green and yellow, and six in the blue. In that way, 

 as these observations can be recorded in about an hour, all the 

 observations made from 1879 to the present time are strictly 

 comparalile. 



Before I state exactly what results we did get it is worth while 

 for one moment to consider what results we should get if the old 

 view of the chemical structure of the solar atmosphere were 

 correct. This view is tliat most of the absorption which pro- 

 duces the Fraunhofer fines in the solar spectrum, especially tliose 

 due to the absorption of the vapours of chemical substances of 

 high atomic weight, takes place close to the photosphere, which 

 is practically the place where the spots live. Thertfore we should 

 expect to see : — 



(1) The :ame lines constantly thickened, as the ordinary solar 

 spectrum is constant. 



(2) T he same lines of a substance widened in all spots, in 

 which we have evidence that that particular substance is present, 

 or the lines thinning out in all cases in the same order. 



(3) No lines but those visible in the general spectrum. 1 



(4) Motion indicated by one line of a substance indicated 

 by all. 



If the old view were true, that we have iron vapour among 

 the other vapours in the atmosphere of the sun— nickel vapour, 

 magnesium vapour, and so on, it would not be at all out of the 

 way to suppose that some spots might chiefly consist of iron 

 vapour, whereas another spot might be chiefly filled with nickel 

 vapour, or with magnesium vapour, and so on. 



With reference to (2) it may be explained that in the case 

 of a spot which we can imagine to contain a very large 

 quantity let us say of iron vapour, and a very small quantity of 

 magnesium vapour ; by a laboratory process A\'hich had been 

 worked out before this work was commenced, it was easy to make 

 a rough, but stil! a very useful, quantitative guess as to their 

 relative proportions, because it has been found in laboratory 

 work that if we only have a very small quantity of one vapour, 

 let us call it vapour a, in a mixture of other vapours, b, c, and d, 

 then we shall not get all the lines of n ; we shall only get some 

 of them, the longest lines ; and if the quantity a is very small 

 indeed, then we shall only get one line — the longest, in the 

 spectrum of the vapour. 



The method of mapping adopted may next be stated. In 

 order to get as much light as we could out of the work, 

 first of all the Fraunhofer lines were mapped in a manner which 

 enabled anybody who took the trouble to look at the map to see 

 which was darkest. In addition to those Fraunhofer lines infor- 

 mation is given showing the origin of them. 



I should tell you that it has been found since Kirchhoff's first 

 researches that it does not do to talk about the spectrum of a 

 substance as if it were an unchangeable thing. To be precise 

 we must refer to the spectrum at the temperature of the arc ; or 

 at the temperature of the electric spark with a jar or without a 

 jar ; or at the temperature of the oxy-hydrogen flame. These 

 spectra are very different indeed, not with regard so much per- 

 haps to the actual lines which we see in any case, but chiefly 

 with regard to the intensity of the lines as seen in one spectrum 

 and in the other. 



Thus it was useful to compare the lines of a substance as seen 

 in the arc and spark with that seen in the spots, and that, so 

 to speak, formed the ground-plan of the maps. The work to 

 be done was to observe all the lines most widened in a region 

 of the spectrum, and see whether they were absolutely unchange- 

 able or whether they were not. 



The diagram shows a part of one of these maps dealing with 

 the first 100 observations. The Fraunhofer lines are at the top ; 

 the lengthening of the lines representing the intensity, that is to 

 say, the longest line is the darkest. Below are the lines of the 

 substance being specially studied seen in the electric arc, the 

 longest being the brightest ; and lower still the lines seen in 

 the electric spark, the longest line al-o being the brightest. 



We ob -erve in the first place that there is a very considerable 

 diSerence in these two spectra. We have a considerable number 

 of lines seen in the arc, and seen in the solar spectrum among the 

 Fraunhofer lines, which are not seen in the spectrum of the 

 spark, the temperature of which of course is assumed to be very 

 much higher than the temperature of the arc. 



Again we may have a very faint line at the temperature of the 

 arc, which is considerab'y intensified when we pass to the tem- 

 perature of the spark. There are, again, other lines seen nearly 

 of the same length both in the arc and in the spark. 



Now, when we first mapped the spjt observations, the maps 

 did not indicate the origin of the lines, and when there was a 

 great variation in the lines discovered it was a fair thing to fay 

 that the explanation lay in the fact that some of the lines be- 

 longed to one substance, and some to another. Let us call those 

 two substances a and b. In some spots we have more of a and 

 in other spots we have more of l>. That of course was very good 

 reasoning, so good that it was necessary to undertake a complete 

 discussion in the case of each element. In this case the spot 

 lines studied are not all the lines which were seen in spots, 

 but the lines of one substance. Now the moment this work was 

 begun strange results appeared, and the matter became difficult, 

 because we should have imagined a prhri that if the same sub- 

 stance were always present in the spot we should always have 

 got the same spectrum, or, at all events, a spectrum along that 

 line to which I have already referred, viz., that if the relative 

 quantity of the vapour were less the number of its lines would 

 be reduced, and at last when the quantity was the least possible 

 the number of lines would be the least possible. I must call 



