1862.] 293 



spectrum of dark lines answering to the bright lines observed in a 

 flame, and referable to a known element, to infer with a high degree 

 of probability the presence or absence, in a state of vapour, of that 

 element in the atmosphere of the sun. Thus the presence of the 

 elements sodium, potassium, iron, and some others in the sun's 

 atmosphere has been rendered in the highest degree probable. 

 And the same reasoning that applies to the sun applies also to 

 the fixed stars, which are known to exhibit in their spectra fixed 

 lines of their own, while many of their lines are identical with 

 those of the sun ; so that the enormous distance of these bodies 

 does not prevent us from drawing some conclusions as to their 

 chemical constitution. 



It is not, however, to be expected that all the fixed lines of the 

 solar spectrum should thus be traced home to particular elements. 

 It is well known that when the sun is near the horizon dark lines 

 make their appearance in the spectrum which are not seen when the 

 sun is high, and which plainly owe their origin to absorption by the 

 earth's atmosphere; though what the particular compounds are 

 which absorb light in this manner is still unknown. It is very pos- 

 sible that gases, having a similar property of absorbing definite rays, 

 may exist in the outer portions of the solar atmosphere, which, at the 

 very high temperature necessary for incandescence, would either be 

 decomposed, or would have their mode of absorption so changed that 

 there would no longer be a perfect correspondence between the bright 

 lines which might be exhibited by the light emitted at a high tem- 

 perature, and the dark lines produced by absorption at a low tempe- 

 rature. 



In mentioning these striking results due to Professor Kirchhoff, it 

 seems right to refer to some earlier researches bearing closely on the 

 subject. In 1848 M. Foucault made the remarkable discovery that 

 the voltaic arc was at the same time a source of light, giving out the 

 double bright line D, and a medium capable of absorbing light of 

 that precise refrangibility ; but he did not extend the result to ordi- 

 nary flames, nor connect the absorption and emission by the theory 

 of exchanges ; and indeed his observation, which was only published 

 in ' Vlnstituty seems to have attracted little attention, and was not 

 known to Professor Kirchhoff when he discovered the general in- 

 version of bright lines. In two remarkable papers published in the 



