CELESTIAL CHEMISTRY. 233 



to pass through a prism of glass, it is decomposed into the beauti- 

 ful colors which are seen in the rainbow. These colors, when in 

 this way separated from each other, form the spectrum of the 

 light. Let this white plate reiDresent the transverse section of a 

 beam of white light travelling towards you. Let now a prism be 

 interposed in its path. The beam of white light is not turned 

 aside as a whole, but the colored lights composing it are deflected 

 ditferently, each in proj^ortion to the rapidity of its vibrations. 

 An obvious consequence will be, that, on emerging from the prism, 

 the colored lights which formed the white light will separate from 

 each other, and in place of the white light which entered the prism 

 we shall have its spectrum, that is, the colored lights which com- 

 posed it, in a state of separation from each other. Wollaston and 

 Fraunhofer discovered that when the light of the sun is decom- 

 posed by a prism, the rainbow colors which form its spectrum ai'e 

 not continuous, but are interrupted by a large number of dark lines. 

 These lines of darkness are the symbols which indicate the chem- 

 ical constitution of the sun. It was not until recently, in the year 

 1859, that Kirchhoff taught us the true nature of these lines. He 

 himself immediately applied liis method of inter^jretation to the 

 dark lines of the solar spectrum, and was rewarded by the discov- 

 ery that several of the chemical elements which exist upon the 

 earth are present in the solar atmosphere. 



"I present the results of the extension of this method of analy- 

 sis to the heavenly bodies other than the sun. These researches 

 have been carried on in my observatory during the last four years. 

 In respect to a large part of these investigations, — viz., those of 

 the moon, the planets, and fixed stars, — I have had the great 

 pleasure of working conjointly with the very distinguished chem- 

 ist and philosopher, Dr. William A. Miller. Half a century ago, 

 Fraunhofer recognized several of the solar lines in the light of the 

 Moon, Venus, and Mars, and also in the spectra of several stars. 

 Recently, Donati, Janssen, Secchi, Rutherford, and the Astronomer 

 Royal, have observed lines in the spectra of some stars. Before I 

 describe the results of our observations, I will state, in a few words, 

 the principles of spectrum analysis upon which our interpretation 

 of the phenomena we have observed has been based, and also the 

 method of observing which we have employed. 



" When light which has emanated from different sources is de- 

 composed by a prism, the spectra which are obtained may differ 

 in several important respects from each other. All the spectra 

 which may present themselves can be conveniently arranged in 

 three general groups. 



"1. The special character which distinguishes spectra of the 

 first order consists^'in that the continuity of the colored band is un- 

 broken either by d;irk or bright lines. We learn from such a spec- 

 trum that the light has been emitted by an ojDaque body, and 

 almost certainly by matter in the solid or liquid state. A spectrum 

 of this order gives to us no knowledge of the chemical nature of 

 the incandescent body from which light comes. 



"2. Sjiectra of the second order are very different. These con- 

 sist of colored lines of light separated from each other. From 

 20* 



