xxiv.] EFFECT OF DIFFERENT CONDITIONS. 335 



solar storms, as already stated, the iron lines sometimes make 

 their appearance in the chromosphere. Now, if we were deal- 

 ing here with one molecular grouping, we should expect the 

 lines to make their appearance in the order of their lengths, and 

 we should expect the shortest lines to occur less frequently than 

 the longest ones. But, precisely the opposite is the fact. 

 Professor C. A. Young records in his observation of the chromo- 

 spheric lines, made at Sherman, that two very faint and short 

 lines close to the triplet near G were observed thirty times, 

 while one only of the lines of the triplet was seen twice. 



The intensity of the lines then is greatly changed when we 

 pass from one set of observations taken under one set of con- 

 ditions, to another set taken under other conditions. It is not 

 a mere question of dropping out the lines when we pass from 

 the temperature of the arc to the temperature of the coil, but 

 in addition there is a considerable intensification of certain of 

 the lines visible under one or another condition. 



Closely adjacent lines sometimes show very great variations ; 

 three lines at wave-lengths 4918, 4919'8, and 4923'2 afford a 

 case in point. 



In the solar spectrum 491 9'8 is thickest, in the oxhydrogen 

 flame neither is visible, in the electric spark 4923*2 is thickest, 

 while it is almost invisible in the electric arc ; under no con- 

 ditions are all intensified at once; each one seems intensified 

 at the expense of the other. 



In photographs of long reaches of the spectrum in which the 

 results of the spark and arc temperatures are shown side by 

 side, this wonderful variation comes out in its strongest form. 

 This does not depend upon my own work only, for Messrs. 

 Liveing and Dewar have recently published such photographs. 1 



We shall have to return to these lines afterwards. 



The strongest contrasts of intensity are seen when the 

 greater differences of temperature are employed ; thus the 



1 Phil. Trans., vol. 174, Part I., p. 212. 



