390 THE CHEMISTRY OF THE SUN. [CHAP. 



Again, the observations have been used by Messrs. Liveing and 

 Dewar to invalidate my conclusions with regard to the three 

 iron lines at 4918, 49 19 '7 and 4923 discussed in chapter xxiv., 

 I hold on the contrary, that their observations form another 

 link in the chain of arguments which can be adduced in favour 

 of my view. 



They say : 



" We have more particularly observed the effect of a current of 

 hydrogen on the iron lines at wave-lengths 4918, 4919'7 and 4923. 

 These lines as seen in the arc in a magnesia crucible usually have 

 about the same relative strengths as are shown in Angstrom's map 

 of the solar spectrum ; Thalen gives their intensities respectively as 

 2, 1, and 3. They are all developed simultaneously when iron is 

 dropped into the crucible, the first being sometimes reversed, the 

 second frequently reversed for some time, the third much strength- 

 ened but not reversed. After a time these effects die out, but if 

 now a very gentle current of hydrogen is led in through one of the 

 carbons perforated for the purpose, the line 4919*7 is again strongly 

 reversed, that at 4918 expanded, while that at 4923 becomes very 

 bright but remains sharply defined. These effects of the hydrogen 

 were observed several times. In all cases the line at wave-length 

 4923 seemed to maintain a.bout the same relative strength compared 

 with the two other lines, and never showed any variation l at all 

 corresponding to the prominence it holds in Young's catalogue of 

 chromospheric lines, where it has a frequency of 40, while that at 

 4918 has only half that frequency, and the strongest line of the 

 three does not figure at all." 2 



With regard to the latter part of this quotation, I may remark 

 that the observations indicate a difference of kind between the 

 two more refrangible lines in one case, and the less refrangible 

 one on the other. It is not a question of degree as is suggested 

 by the use of the word " variation." 



If Messrs. Liveing and Dewar had seen the line at 4923 



1 The italics are mine. J. N. L. 



2 Proc. Roy. Soc. vol. xxxiii. p. 430. ' 



