184 



THE AMERICAN MONTHLY 



[October, 



sis may be readily illustrated by a few 

 simple experiments. We have here 

 the colorless flame of a Bunsen burner. 

 I take a platinum wire with some 

 common salt upon the end This 



Fig. 40. Slit of Zeiss Spectral-ocular. 

 colors the flame yellow. Examined 

 with a spectroscope we find only a 

 single yellow line indicating the pre- 

 sence of sodium. Strontium and lime 



it their characteristic colors. The 

 spectroscopic examination of these 

 flames shows that the lines produced 

 by each element occupy different 

 positions in the spectrum, so that by 

 measuring the places of the lines in 

 the spectrum one can readily deter- 

 mine the composition of the flame 

 that produces them. Most metals, 

 however, require the more intense 

 heat of the electric arc to develop 

 their characteristic lines, and a few 

 elements, as carbon and boron for 

 example, are only volatilized in the 

 intense heat of the sun. 



The temptation to dwell upon the 

 applications of the spectroscope in 

 solar and stellar physics is great, for, 

 this subject is, to me, one of the 

 greatest interest. I would be pleased 

 to review the progress of this branch 

 of investigation during the past tvventy 

 years. In no department of physical 

 research have important discoveries 

 followed in such rapid succession ; 

 in no department have results of such 



Fig. 41. Sorby-Browning Microspectrocope. 



give a red flame, baryum a green, 

 potassium a violet, and several other 

 metals are sufficiently volatile at the 

 temperature of the flame to impart to 



a surprising character been so readily 

 revealed. I would be glad to speak 

 at some length of the recent researches 

 upon the chemical and physical con- 



