28 Dr. E. N. da C. Andrade on the Electrical 



is required to produce a given luminosity, so that in further 

 experiments the pyrometer reading gives approximately the 

 rate of vaporization in mg8. per hour. 



As the luminosity is roughly, at any rate, independent of the 

 pressure for a given rate of vaporization of the salt, it was 

 thought that the flame temperature might be lower at high 

 pressure than at atmospheric, and that this might be the reason 

 why the streak is somewhat paler at high pressure, with the 

 same salted wire in the flame in each case. Measurements were 

 made of the temperature of the flame at different points at 

 different pressures by means of a platinum- platinorhodium 

 thermocouple. It was found that the flame was on the 

 average nearly 100° cooler at 4 atmospheres than at 1 

 atmosphere. 



§ 8. r lhe Migration of the Positive Carriers at 

 High Pressure. 



When the salt is brought into the flame at high pressure 

 in a central position, the coloured streak follows the form of 

 the flame, and is consequently divided into three points 

 above. It is brightest in the central portion of the flame, 

 and lighter at the sides. 



When the electric field is applied these points merge, as 

 stated in the last paragraph ; the streak of luminous vapour 

 is, however, always split into two in the electric field, 

 showing the existence of both positively- and negatively- 

 charged luminous metal atoms. At three atmospheres excess 

 pressure, e.g., if the bead is central the deflexion of the two 

 parts are approximately equal, the positive and the negative 

 streak being roughly equally bright. The deflexion of the 

 flame alone is hard to estimate since it is broadened out 

 (§7), and seems to be deflected to some extent in both 

 directions. It was attempted to estimate the deflexion of 

 the (lame by bringing coal-dust particles into it. as was 

 done to determine the ilame velocity *, but this was un- 

 reliable, since the particle- probably become charged. 



The difficulty was got over by bringing the bead into the 

 flame close to the edge nearest the negative electrode : the 



deflexion of this edge of the flame was taken a- the deflexion 

 of the flame at the region occupied by the streak of positively 



charged vapour. The deflexion oi' the positive edge of the 

 streak of metallic vaponr was observed, thus giving the maxi- 

 mum velocity of migration of the metallic atoms forming the 

 luminous si reak. The deflexion- were read on a glass Bcale, the 



• See p. SO. 



