1922] on Problems in the Variability of Spectra 529 



which are due to atoms, though the arrangement of the lines which 

 constitute each band is of the type usually found in band spectra. 

 When powerful condensed discharges are passed through helium a 

 spark spectrum is developed. Two series in this spark spectrum 

 are known as the 4686 and the £ Puppis series, and their discovery 

 by Professor Fowler has led to some of the most important develop- 

 ments of theoretical spectroscopy. These spark lines of helium arc; 

 found in the nebula? and early type stars, and are attributed to 

 helium atoms which have lost an electron. 



The energy required to produce spark spectra varies widely with 

 the nature of the gas under investigation, and for elements of the 

 same chemical group is, as a rule, smaller the greater the atomic 

 weight of the element. Thus, in the case of helium powerful dis- 

 charges are required for the production of the spark spectrum, and 

 the lines of the arc series are always bright. In the case of argon a 

 much less intense discharge is required to produce the spark lines, 

 and with very powerful discharges the arc lines disappear almost 

 entirely from the spectrum. In addition to the production of these 

 spark spectra one of the effects of powerful condensed discharges is 

 to alter the relative intensities of the arc lines. Generally speaking, 

 the effect of an increase of energy on a particular series of lines is 

 to enhance relatively the more refrangible members of the series, but 

 the effect varies in degree for different series. Experiments of this 

 kind enable us to imitate to some extent in the laboratory the dis- 

 tribution of intensity amongst the lines which is found in nebular 

 and stellar spectra. It will be seen that whilst many variations in 

 spectra can be referred to different compounds, to molecules, and to 

 uncombined atoms in successive stages of ionisation, there are a 

 number of other changes for which there is at present no obvious 

 theoretical explanation. The possibility of some specific influence 

 of one gas on the spectrum of another must now be recognised apart 

 from the formation of chemical compounds, which, in the action of 

 helium on the spectrum of hydrogen, appears to be excluded. There 

 is other evidence, based on a study of the broadening of spectrum 

 lines, of a specific action on neighbouring atoms. We are still 

 awaiting a satisfactory theoretical explanation of phenomena of this 

 kind, though it is now forty years since what is perhaps the first 

 known example, the action of sodium on the absorption spectrum 

 of magnesium vapour, was observed by Professor Liveing and 

 Sir James Dewar in this Institution. 



[T. R. M.] 



2 P 2 



