342 



TRANSACTIONS OP SECTION A. 



2. The Dependence of the Spectrum of an Element on its Atomic Weight. 

 By Professor W. M. Hicks, D.Sc, F.R.S. 



In addition to the well-known series lines of a spectrum, a large number of others 

 are directly related to them in the same way as the second or third set of a doublet 

 or triplet series depends on the first, or as the satellite fines of a D series depend on the 

 more intense set. The relation may be represented in the following way. The wave 

 number of a fine is given by a formula of the form N/Jl +D,p — N/{«* + D } 2 which 

 may be denoted for, say a S r series by S^D,) - VS (m). The S 2 series will then 

 be denoted by S 1 (D,-W 1 )-VS(m). As instances of one type of the relations 

 in question, the following may be taken as examples amongst many others. 



In the above the first column gives the wave length, the second the dependence 

 on the atomic weight term W, or W.,, the third the difference between calculated and 

 observed, on the supposition that the observed value for S, (1) is exact, and the last 

 the limits of possible error. The examples are chosen from the sharp series — but 

 the other series show similar connections. 



3. On the Arc Spectra of certain Metals in the Infra-red Region (\ 7600 

 to X 10,000). By Major E. H. Hills, F.R.S. 



In 1880 Abney showed that it was possible to prepare a collodion-bromide 

 emulsion which was sensitive in the infra-red region about as far as M0, 000, and 

 a few years later he published his now well-known map of this portion of the 

 solar spectrum. 



With one or two exceptions he did not extend his researches to include 

 metallic spectra and since that time it does not appear that any other experi- 

 menter has followed up this line of work. For such observations of infra-red 

 spectra as have been made either some form of bolometer or thermopile has 

 been employed or, where photographic methods have been used, recourse has 

 been had to plates stained with an organic dye. The latter method is not avail- 

 able beyond A8,000, while with any form of thermopile method it is extremely 

 difficult to attain adequate resolving power. It would appear that Abney 's 

 emulsion furnishes the best means of recording lines in this part of the spectrum 

 and it is difficult to understand its total neglect by spectroscopic workers. 



The importance of observing infra-red spectra lies not in the addition of a 

 few new lines to the known metallic spectra, but in the fact that these new 

 lines are just those most favourably situated for obtaining the laws governing 

 series of spectrum lines, and therefore leading to fresh light on the interior 

 dynamics of the molecule. 



For the present research a spectroscope of two dense flint prisms was 

 employed. Wave lengths were determined by comparisons with the sharp lines 



