March 4, 1886] 



NATURE 



427 



Angstrom, and Thalen, who were the first workers in this field 

 of inquiry. 



Table A. — Elements present in the Sun according to KirchJioff, 

 Angstrd>n^ and ThaUn 



KirchhoO" 

 Sodium 

 Iron 

 Calcium 

 Magnesium 

 Nickel 

 Barium 

 Copper 

 Zinc 



Angstrom and I'hale: 

 Sodium 

 Iron 

 Calcium 

 Magnesium 

 Nickel 



Chromium 



Cobalt 



Hydrogen 



Manganese 



Titanium 



Another gives the substances which were added to the preceding 

 list by taking a special consideration into account. Some time 

 after the first work on the chemical composition of the solar 

 atmosphere was accomplished, a method was introduced by 

 which it was easy to determine the existence of a small quantity 

 of any particular vapour in a mixture ot vapours, so that tlie 

 substances indicated in the second table are those substances 

 which possibly exist in the sun's atmosphere in a small quantity 

 only. 



Table B. — Elements the Longest Lities of ivhieh coineide with 

 Fraiinhofer Lines 

 Certainly coincident Probably coincident 



Aluminium Indium 



Strontium Lithium 



Lead Rubidium 



Cadmium Caesium 



Cerium Bismuth 



Uranium Tin 



Potassium Silver 



Vanadium Glucirum 



Palladium Lanthanum 



Molybdenum Yttrium or Erbium 



It is important to call special attention to the fact that 

 Angstrom and Thalen, who followed Kirchhoff, did not agree 

 with regard to barium, or copper, or zinc, and they added 

 chromium, cobalt, hydrogen (a very notable addition), man- 

 ganese, and titanium, the existence of which Kirchhoff had not 

 discovered in the solar atmosphere. 



A detailed study of the facts recorded by Angstrijm gives a 

 good idea of the immense difficulty of the research, and also of 

 the doubts and of the difficulties which were suggested in the 

 very first part of the inquiry. For instance, in the case of 

 sodium, what Angstrom did, of course, was to get the vapour of 

 sodium incandescent in the laboratory, and he got the eight 

 familiar lines. He then observed whether there were dark lines 

 corresponding with all of them. He found that there were. 

 With regard to cobalt he got nineteen lines, and he found nine- 

 teen lines in the sun coinciding with them. But when he studied 

 the spectrum of barium in his laboratory he got twenty-six lines, 

 but of these in the solar spectrum he found only eleven. When he 

 came to aluminium, of the fourteen lines seen in the spectrum of 

 the metal only two existed among the Fraunhofer lines. In 

 zinc it is not yet quite decided whether we even have two out of 

 twenty-seven ; so that we see it was not all perfectly plain 

 sailing. 



A New Vu-u 



So much, then, for the chemistry of the solar atmosphere, 

 taken as a whole. Two observations suggest themselves : the 

 first is, that it is perfectly clear that if we have in the sun's 

 atmosphere incandescent iron vapour, and calcium vapour, and 

 magnesium vapour, and the incandescent vapours of many other 

 substances which we generally know here as solid bodies, there 

 must be tremendously strong convection-currents somewhere ; 

 for were these vapours at rest they must cool on the out- 

 side, and if they get up high enough they will condense, first 

 into liquid particles, and then into solid particles, and then they 

 are bound to go down. So that we see there is a new world of 

 motion in full front of us the moment we are driven to the 

 conclusion that we are really dealing with a mixed mass of 



gases so intensely hot that its constituents exist in it, 

 except in the coldest parts of it, in a state of vapour. 

 To enable us to think this out a little, let us consider 

 a small part of the sun where we will imagine that the statical 

 condition is as nearly secured as possible, and that then 

 we suddenly upset the temperature equilibrium. When we get 

 any solid particles, say of iron, falling into a region where 

 they will be gradually melted, and then driven into iron 

 vapour, the vapour is bound to reascend — it will not continue in 

 its downward flight ; whereas if vapours, by ascending, gradu- 

 ally get cooler and cooler, they must afterwards redescend, falling 

 first, as I h.ive said before, as mist and then in big liquid drops, 

 and finally as solids — as meteorites, if you like : they are bound 

 to go doH n. To put this in the most general form, we may say 

 that in the sun's atmosphere complex molecules are bound to go 

 down, and simple molecules are bound to go up ; so that we 

 shall have convection-currents, as I have already hinted, pro- 

 duced in this kind of way, and these convection-currents must 

 exist wherever the temperature equilibrium is bro'.ven. Of course 

 we must assume that these more or less vertical convection- 

 currents may be modified to a certain extent by the rotation of 

 the sun, in the same way as the up and down currents of our 

 own air, and even the currents pole-wise and equator-wise are 

 modified by the rotation of the earth. 



The other observation which suggests itself is as follows : — 

 We need not limit ourselves to the general chemical ideas «'e 

 have acquired ; the chemistry of each part of the sun (always 

 above the photosphere) can be examined bit by bit. The photo- 

 sphere has spots in it ; it has the chromosphere above it with 

 the included prominences, the inner corona above that, the outer 

 corona above all. As a matter of fact, all these have now been 

 examined, bit by bit, by the spectroscope — that is from the 

 chemical point of view. 



The next point of importance to urge is that a view of the 

 solar constitution has been arrived at in consequence of this new 

 wealth of facts, on which something must be said before we 

 go further. 



The old view put the absorbing atmosphere above the photo- 

 sphere, the various chemical substances being arranged in the 

 order of their vapour-densities, so that hydrogen would be 

 highest, then sodium, then magnesium, till fin.ally we arrive at 

 iron and platinum, and so on. 



Now, if that view were correct, it would be perfectly easy to 

 prove it at once by the new method of local examination. We 

 have only during an eclipse, or even without an eclipse, to put 

 the slit of a spectroscope on the edge of an image of the sun 

 thrown by an object-glass, and observe the spectrum from each 

 part of the sun ; from the photosphere to as far above the photo- 

 sphere as we can get. If that view were true, we should get, 

 as short bright lines close to the photosphere, the lines of sub- 

 stances having a high atomic weight. Then higher and higher 

 we should get longer lines indicating the existence at greater 

 heights in the solar atmosphere of those substances which have 

 a sm.aller atomic weight. Further, as we have evidence to show 

 that the spots exist in a low part of the sun's atmosphere (how 

 low we shall see by and by, when we come to consider them in 

 detail), we should expect in those spot^ to find all the familiar 

 line^ of the substances having a high atomic weight to be 

 affected. When we examine the chromosphere and the base of 

 the prominences which arise out of it, we should find that at the 

 same height or about the same height where the spots give us 

 lines of the substances of high atomic weight, the chromosphere 

 itself should be full of the same substances of high atomic 

 weight. 



Now the fact that not one of these expectations is realised — 

 that none of these things are so — has necessitated the putting 

 forward of the new view to which I have referred. 



This can be stated in a very few words. It is that the 

 temperature of the sun is not only sufficient to drive all 

 our most refractory metals into vapour, as we can do in 

 our laboratories on the earth, but that it goes very much 

 further ; it continues the work of our laboratories, and 

 drives them into something else altogether finer than anything 

 that we can separate with our terrestrial conditions. Accord- 

 ing to that view, what would happen would be this : — If we 

 could lay hold of a solar meteorite, say a hundred thousand 

 miles from the photosphere, and watch it in its downward 

 flight, the solid would first become liquid, it would then be 

 vaporised, and we should have the spectrum with which we are 

 familiar in our laboratories ; but after that the vapour would still 



