366 The Principles of Spectrum Analysis. 



These ingenious and indefatigable workers found that the 

 bright lines in the metallic spectra corresponded closely with 

 the dark lines of the solar spectrum. Why was this ? and 

 how could it be explained ? They found, by experiments be- 

 fore cited, that each colour was opaque to rays of its own colour. 



To illustrate the point more clearly, we may suppose two 

 pendulums of equal length to be placed side by side. If the 

 one be made to vibrate, it will, after a time, cause its com- 

 panion to do the same in consequence of its equal length or 

 isochronous condition; and so it is supposed that the rays of 

 one colour will be taken up by another whose vibrations are of 

 equal length, and so be arrested in their voyage. Now look 

 at the application of this result. If, in the rays of light from 

 an artificial source, this principle be correct, it may also be 

 correct with the rays of solar light. Hence they have inferred 

 that the dark bands of the solar spectra are produced by their 

 passage through an atmosphere containing certain metals in a 

 state of high combustion or vapour. 



Upon these grounds they have concluded that in the outer- 

 most solar envelope exist all those metals in a state of vapour, 

 whose colour-bands coincide with dark lines of the solar spec- 

 trum, as sodium, potassium, iron and nickel ; and that it is by the 

 more powerful light of the photosphere shining through this only 

 feebly-luminous layer that the dark bands of Fraunhofer are 

 produced by the process before described. They have also 

 inferred the source from which these metals are derived to be 

 the mass of the sun. A bold assertion, perhaps a correct one; 

 but we are' certainly not at present justified in accepting it as 

 if it were proved ; and we may advantageously reflect upon the 

 words of Dr. Miller : — 



" Fascinating as this theory is, it must be remembered that 

 it is yet upon its trial, and that it does not explain the facts at pre- 

 sent known respecting the vapours of hydrogen, mercury, chlorine, 

 bromine, iodine, and nitrogen. M. Morren even questions the accu- 

 racy of some of KirchhofTs observations. Thus, he states that in 

 a measurement which he made of the red band of potassium, con- 

 jointly with Pliicker, they found that it did not correspond with the 

 solar line A, but that it is considerably more refrangible." 



There are many other important facts which will have to 

 be considered before we can arrive at a complete theory of the 

 spectral phenomenon. For example, chloride of lithium pro- 

 daces a single crimson line in, the flame of a Bunsen burner ; 

 the greater heat of a hydrogen name enables it to emit an orange 

 ray, and the voltaic arc adds a brilliant stripe of blue. In like 

 manner, iron and other metals furnish spectra which advance 

 t in complication as the temperature of their vapour is increased. 

 Professor Roscoe says that " the general rule is that lioui- 



