Brush Discliar ge in Weak Acids and Solutions*. 811 



sulphide. In all the cases in which it makes its appearance 

 the conditions oE excitation are feeble and therefore the 

 temperature probably low. Measurements on the breadths 

 of spectrum lines in Geissler tubes show that the temperature 

 must be approximately room temperature, and that of the 

 glow discharge in the experiment of Himstedt and von 

 Dechend is not likely to be higher. It seems unlikely then 

 that the sulphur should produce a spectrum at room tem- 

 perature in the part of the discharge only where the conditions 

 of excitation are feeblest, i. e. farthest from the point. 

 Sulphur produces its spectra in the vacuum-tube, when it is 

 heated sufficiently to produce a high enough vapour-pressure 

 for the discharge to pass. 



The spectrum appears in dilute sulphuric acid, solutions of 

 sulphates, hydrogen sulphide, sulphur dioxide, and in no 

 other solutions where sulphur was not present- It only 

 occurred in those cases where a bubble formed at the point 

 and was the spectrum of the light emitted by the bubble. 

 Phosphoric acid gave a green bubble showing bands in the 

 green which, however, were too faint for spectrographic 

 work. These are the only two acids which were tried which 

 gave coloured bubbles. 



Spectra of the Brush Discharge, 



The spectra which occur in the brush-discharge are those 

 of:— 



Hydrogen : Series spectrum and secondary spectrum. 



Solute : Spectrum of the metal of dissolved salt. 



Oxygen : Series spectrum and elementary line spectrum. 



Water vapour : Ultra-violet band spectrum. 



Electrode : Arc spectrum (platinum). 



Impurities : Nitrogen (positive and negative band 

 spectra), Calcium, Sodium, Silicon (line 

 and ultra-violet band spectrum), and Lead 

 (arc) spectra. 



The spectrum characteristic of distilled water differs some- 

 what in the two discharges. The negative brush, which is 

 blue in the most highly resistant water, shows a continuous 

 background stretching from the red to about A 3300. It 

 is strongest photographically between X 4861 and A 3800 

 (fig. 4 6, PI. XL), and it is to this that the blue colour is due. 

 As the conductivity rises, the brush becomes pinker and the 

 continuous spectrum fades (fig. 4 c). The spectrum of the 

 positive brush contains very little continuous background 

 (fig. 4 a). 



