August 1 8, 1881] 



NATURE 



365 



' SOLAR PHYSICS— THE CHEMISTRY OF THE 



SUN' 

 The Test iupplied by Change of Refrangibility 

 VSTE have then got so far. Limiting our studies to iron we 

 find that the prominence spectrum is made up of one set of 

 lines seen in the terrestrial spectrum, and the spot spectrum made 

 up of another set. And more than this, if «e add the lines seen 

 inthe prominence and spot spectra together we do not then by 

 any means make ufi the complete spectrum. 



It is fair to ask the foUuwing question : — Have we any other 

 means of establishing this extraordinary fact of the separation of 

 the iron lines in spots and storms ? We have. Reference has 

 already been made to the change of refrangibility of the lines 

 brought about by the change of velocity of movement of tlie 

 various solar vaji lurs. But if, as already hinted, the lines of 

 iron behave to each other in precisely the same way as the lines 

 of two perfectly distinct substances behave to eacli other ; then 

 if we obsei-ve changes of refrangibility in the iron lines, both in 

 spots and flames we should get the same differentiation as we 



have already got in the lines thickened or intensified in the 

 spectra of spots and flames. 



We will now see the results which have been obtained along 

 this line of research, and it should be pointed out that it is not 

 a method by which it is easy in a short time to accun.ulate a 

 large number of observations, because metallic prominences are 

 very rare except at the sun-spot maximum, and in the case of 

 spots we not only want a spot, but we w ant that spot to be in a 

 very considerable state of commotion, n\ order that the change 

 of refrangibility may be obvious enough to enable us to record 

 the phenomena. 



So far as this inquiry has gone at present we have only ob- 

 served the lines contorted in spots. 



In the diagram (Fig. 37) the zig-zag lines indicate the iron lines 

 which changed their refrangibility in a number of spots observed 

 at the end of last year, T he point is that, although we have a 

 great many of the iron lines bent, twisted, contorted — with their 

 refrangibility changed, yet some of the iron lines mixed with 

 them give us no indication of niovemenl. All these observations 

 have been made upon lines seen at the same moment in the 



Fig. 37.— Different rates of motion registered by different iron lines. 



same field of view. Observations of-^this nature exist twelve 

 years old, but no importance can be attached to them, for the 

 reason that the phenomenon was not understood, as I hope it is 

 understsod now, and jjrecautions were not laken in the obser- 

 vations then made to show that no motion of the slit across the 

 spot took place in tlie interval between the two observations. 

 For, of course, it is not fair to compare a line which one sees 

 in one part of the spectrum with a line seen in another, unless 

 one is absolutely certain that the slit has not moved on the sun's 

 miage; because one-thousandth part of an inch on the sun's image 

 means a good many miles on the sun. Referring to Fig. 37 

 we have, at wave-length 5366-70, three lines, two in motion, 

 and one at rest, all belonging to a well-known group of iron 

 Imes. At a later date we have the line at 5382 .at rest, while 

 that at 5378 is in motion. Thus it will be seen that these points 

 and ethers prove there is just as much individuality in the way 

 in which the lines of iron change their refrangibility as there is 

 m the way in which one particular line, and then another, is 

 thickened in a sun-spot or brightened in a prominence ; and if 



olar Physics at South Kensington (see p. iso)' 

 Continued from p. 324. 



we go further we find this very interesting'and additional fact, 

 that the lines which are not contorted are in a great many cases 

 precisely those lines which are seen in the flames, but not in the 

 spots. 



It is seen therefore that the evidence afforded by change of 

 refrangibility is of like nature to that afforded by the thickening 

 of lines in spots and brightening of lines in flames. 



The explanation which lies on the surface is that the vapours 

 in the flames produce one set of lines in one place or at a certain 

 temperature, and the vapours in the spots produce another. 



Sometimes these vapours are mixed up by up-or-down rushes, 

 and sometimes therefore the lines are common. 



Bearing of these Observations on the Origin of the Fraioihofer 

 Lines 

 At the end of the last lecture it was pointed out that the obser- 

 vations we are now discussing seem to indicate that in time we 

 may be able to say that the absorption to which any particular 

 Fraunhofer line is due takes place in a certain region of the solar 

 atmosphere, whereas formerly we could only say that it was 

 produced by absorption someiohere. 



