246 



Professor George D. Liveing 



[March 9, 



Fig. 1. 



a matter of no small importance when faint light is to be observed 



(see Fig. 1). 



Now, although the intensity of the sun's rajs falls away rapidly 



beyond the Fraunhofer line H, and comes to nothing about as far 



above H as F is below it, it is far otherwise with the radiation of 



our terrestrial elements when 

 heated up in the electric spark 

 or arc, or even in some cases in 

 flames ; some of those elements 

 which we know to be abundant 

 in the sun, such as iron aud 

 magnesium, exhibit their most 

 intense radiation, their strongest 

 and most persistent rays, in the 

 ultra-violet region, in waves 

 which succeed one another at the 

 I l/l shortest intervals. Indeed those 

 metals so readily take up certain 



L._...^ 



Solution of 





QUARTZ 

 PLATE 



ultra-violet vibrations, that when 

 there is much metal in the arc, 

 and it is confined in a crucible 

 of lime or magnesia, they often 

 give their characteristic lines 

 strongly reversed, dark absorp- 

 tion-bands being produced by the 

 slightly cooled vapour which is 

 outside the arc. This is seen in 

 the photographic plate Nos. 1 

 to 3. No. 2 shows the strongest 

 magnesium line, in a region 

 beyond the limit of the solar 

 spectrum, at wave length 2852, 

 expanded and reversed. No. 1 

 shows it enormously expanded, its 

 bright wings reversing inm lines 

 up to S. No. 3 shows a strong 

 group of iron lines, still more 

 refrangible, also expanded and 

 reversed by putting iron wire into the arc. The dark bands in 

 the photograph are due to absorption by the metallic vapour, and in 

 their places strong bright lines appear when less metal is present. 

 The spectrum of iron is of all metals the most complicated, and 

 those of the other elements which are most closely related to iron 

 in chemical characters come next to it in the number and com- 

 plication of their ultra-violet lines. Manganese and chromium are 

 especially remarkable for showing many groups of closely-set lines. 

 No. 2 shows a group of chromium lines between the solar lines S and U . 

 It is probably not without significance that this group of elements 



