THE SIGNIFICANCE OF SPECTROSCOPY 229 



in our laboratory. All we require is that it shall send us light. 

 Spectrum analysis is not, it is true, a complete substitute for 

 chemical processes. Quantitatively, it gives, at present at any 

 rate, only a rough approximation to the truth. But, so far 

 as its scope extends, it is of inestimable value. Nor are its 

 uses entirely chemical. The distinction between flame, arc, 

 and spark spectra is a very fundamental one, and leads us to a 

 knowledge, not only of the elements producing a spectrum, 

 but also of their physical state. The most extensive applica- 

 tions of this distinction are astronomical. It forms the basis 

 of a theory of evolution of the stars, and has already given us 

 a great deal of information concerning the physical condition 

 of the sun at different depths below its surface and in the 

 interior of sunspots. The spectrum is, in fact, the most pro- 

 foundly inspired utterance of the celestial bodies that has yet 

 descended to us, and the secrets it has revealed, incredible as 

 they would have appeared a hundred years ago, impress us 

 to-day even more as an earnest of greater revelations to come 

 than with a sense of their own intrinsic wonder. 



It is interesting to note that the great majority of astro- 

 nomical spectra are due to absorption. The solar spectrum, 

 for instance, consists of a continuous background crossed by 

 a large number of fine lines and bands — the Fraunhofer lines. 

 As a result of our laboratory experiments, there is no difficulty 

 in interpreting this. The sun consists of a central source of con- 

 tinuous radiation, surrounded by a cooler envelope of luminous 

 vapours, which by themselves would give a bright spectrum 

 coincident with the dark lines observed. The constitution of 

 this outer envelope may therefore be inferred by comparing the 

 dark lines with the emission lines of various substances produced 

 terrestrially. As a confirmation of this reading of the spectrum, 

 we can actually obtain the bright lines, with a dark background, 

 by analysing only the light from the sun's edge. A solar eclipse 

 offers very favourable circumstances for the observation. 



We have compared a radiating atom to a peal of bells, 

 representing each wave-length by a musical note. It is natural 

 to ask whether the waves emanating from a particular atom 

 are a haphazard, chaotic collection, or whether there is any 

 regularity in them corresponding to melody. To answer this 

 question, we turn to the simplest spectrum known, namely 

 that which is given under certain conditions by hydrogen. 

 As long ago as 1885 Balmer noticed that the wave-lengths (X) of 

 the lines in this spectrum could be calculated with very great 

 accuracy by giving m the values, 3,4, 5 in the formula 



X='?646'i'^. — 1 . The unit of wave-length here is the 



" angstrom," or " tenth-metre," which is equal to 10"^^ 



