510 Mr. W. S. B. Woolhouse on certain Triadic 



again recur. Amongst the large number of those salts already 

 examined by us, which, owing to their volatility in the flame, 

 render a spectrum-analysis possible, we have not found, in spite 

 of the great variation in the elementary bodies combined with 

 the metal, a single one which failed to produce the characteristic 

 bright lines of the metal. Considering these numerous obser- 

 vations, made under the most widely differing circumstances, we 

 might be led to suppose that in all cases the bright lines given 

 out by a body occur quite independently of the other elements 

 chemically combined with that body, and that therefore the 

 relation of the elements, as regards the spectra of their vapours, 

 is exactly the same, whether they are free or chemically com- 

 bined. Yet this supposition is by no means founded on fact. 

 We have repeatedly insisted that the bright lines in the spectrum 

 of a luminous gas must coincide with the absorption lines which 

 this gas produces in a continuous spectrum of a sufficient degree 

 of luminosity. It is well known that the absorption lines of 

 iodine vapour cannot be produced by hydriodic acid, and that, 

 on the other hand, the absorption lines of nitrous acid are not 

 visible in a mechanical mixture of nitrogen and oxygen. There 

 is nothing to show why an influence of chemical combination 

 upon the absorption lines, similar to that here noticed at low 

 temperatures, should not occur at a white heat. If, however, 

 the state of chemical combination alters the absorption lines of 

 a luminous gas, it must likewise alter the bright lines of its 

 spectrum. 



From these considerations one would conclude that in the 

 case of the spectra of two different compounds of the same 

 metal, different bright lines may appear ; it is, however, possible 

 that the salts which are volatilized in the flame cannot exist at 

 the temperature of the flame, and are decomposed, so that it may 

 be in reality the vapour of the free metal which produces the 

 lines ; and it would then appear quite possible that a chemical 

 compound may produce bright lines differing from those pro- 

 duced by its constituent elements. 



LXV. On the Rev. T. P. Kirkman's Problem respecting certain 

 Triadic Arrangements of Fifteen Symbols. By W. S. B. Wool- 

 house, F.R.A.S., FLA., F.S.S. tyc* 



IN the f Lady's and Gentleman's Diary ' for the year 1844, I 

 proposed the following mathematical prize question : — 



* Communicated by the Author. 



