720 Dr. S. R. Milner on the Current-Potential 



and the sheath might be expected as a result of the different 

 ionizing agents in the two cases. The impact of a positive 

 ion on a molecule might be expected, in consequence of its 

 greater mass and the probably greater amount of energy 

 which it communicates to the molecule, to produce a different 

 type of vibration in the ion formed from that resulting from 

 the impact of a corpuscle. That the essential feature for the 

 production of the core lines is the impact on and ionization of 

 the air molecules by positive ions is proved by an observation 

 made in 1884 by Schuster *, who showed that the line spectrum 

 of air (i. e. that of the core of the spark), is seen in the glow 

 which covers the cathode in vacuam-tubes. This is a region 

 in which ionization by positive ions is now known to be 

 occurring and in which the production of the lines cannot 

 be attributed to the current density or the temperature, 

 which are both comparatively low. With regard to the 

 second air-line spectrum Hemsalech discovered that, unlike 

 the core spectrum, it occurred in the oscillations subsequent 

 to the first discharge. This effect allows us to associate it 

 definitely with the glow discharge (ionization by negative 

 ions) which we have seen occurs in the interval between the 

 oscillations. 



3. Nature of the Streamers. — The inclination with respect 

 to the axis of the spark of the luminous streamers which are 

 seen when the spark is viewed in a rotating mirror, shows 

 that the luminosity of the electrode-metal vapour occurs at 

 the centre of the spark at a later instant than near the poles. 

 Schuster and Hemsalech from their researches f came to the 

 conclusion that the reason of this delay was that the metal of 

 the poles was vaporized and rendered incandescent by the 

 heat of the spark, and that the vapour takes an appreciable 

 time to diffuse from the poles to the centre of the gap. 

 Schuster and Hemsalech's observations, however, only apply 

 to the inductionless spark, where the streamers corresponding 

 to the successive oscillations are all massed together and 

 superposed on each other in the photographs. Their con- 

 clusion therefore does not strictly apply to the individual 

 streamers in the successive oscillations of the discharge. 

 Eoyds if, from an extension of their work to sparks with 

 induction in circuit, concluded that it is in the case of the 

 first streamer only that the inclination represents the velocity 

 of diffusion of the vapour, and that in the subsequent ones it 

 represents the rate of propagation through a comparatively 



* Schuster, Bakerian Lecture, Proc. R. S. xxxvii. p. 317 (1884). 

 f Phil. Trans. A. cxciii. p. 189 (1900). 

 | Phil. Trans. A. ccviii. p, 333(1908). 



