A.— MATHEMATICAL AND PHYSICAL SCIENCES, 89 



The Nebular Spectrum. 



In the nebuloe are spectrum lines which have never been observed 

 terrestrially. These are not faint members of otherwise complex spectra, 

 such, for instance, as we have in nearly all remaining unidentified lines of 

 the solar spectrum, but they stand out, bold and challenging, on a dark 

 background, presenting a puzzle that was the more intriguing from its 

 apparent simplicity. According to spectroscopic experience, now made 

 precise and rational, simple spectra are due to light elements. This, 

 taken with the fact that lines known to be due to hydrogen and helium 

 accompanied the nebular lines, strongly suggested that they too were 

 due to light elements of the class which terrestrially are known as perma- 

 nent gases. But the fact remained that no one had succeeded in observing 

 them in the laboratory, and as time went on the originally convenient 

 resource of relegating them to an unknown element had become less con- 

 venient. For the scheme of the elements became definite, and there was 

 no room in it for new light elements. This was one of the many cases in 

 science where the method of frontal attack has been exhausted in vain. 

 More systematic knowledge of spectra in general, and of the spectra of 

 the light elements in particular, was wanted before the question could be 

 resolved. 



The clue was afforded by the circumstance that important nebular 

 lines occur in pairs, obviously associated by their closeness and their 

 constant relative intensity in different nebulae and in different parts of 

 the same nebula. The consideration of such pairs or multiplets has more 

 than once proved an advantageous point of attack on spectroscopic 

 problems. It was in this way that Hartley, examining the diffuse triplets 

 of magnesium, first established the constancy of frequency intervals, thus 

 suggesting for the first time that addition and subtraction of frequencies 

 was the proper method of analysing spectra — an idea which appeared at 

 that time sufficiently paradoxical. Again, the recognition of the frequency 

 intervals of multiplets afforded the clue by which complex spectra such 

 as manganese and iron were first unravelled. 



It is found then that the frequency difference of the green pair of lines 

 originally discovered by Huggins, and known as N, and N^ is 193 waves per 

 centimetre. I. S. Bowen, to whom we owe the final elucidation of this 

 enigma, sought for an equal interval in the spectrum of doubly ionised 

 oxygen which he was analysing and found it in the interval between the 

 low-lying levels designated as TF^ and PPj. 



This is hardly enough in itself to establish the suggested origin ; to do 

 that it is necessary to fix, not only the interval between the nebular lines, 

 but their position as well. The lines were attributed to intercombination 

 between one singlet upper level and two lower levels belonging to a triplet, 

 the third being excluded by the rule of inner quantum numbers. To fix 

 the differences of the terms concerned it was necessary to connect the singlet 

 and triplet levels by an intercombination line observed in the laboratory 

 spectrum of doubly ionised oxygen. This was done by A. Fowler, who, 

 combining Bowen's laboratory data with his own, was able to get a fairly 

 satisfactory check on the observed position of the nebular pair. Practically 

 no doubt remains, in view of the fact that other less well-known nebular 



