﻿Electricity in Flames containing Metallic Vapours. 879 



electrons of the flame get absorbed by neutral metallic atoms, 

 which diminishes their velocity considerably: hence a certain 

 number of electrons which before, in consequence of their high 

 velocity, travelled almost horizontally out from the lower 

 part of the streak of metallic vapour, now, combined with 

 metal atoms and travelling slowly, arrive at the earthed 

 electrode at a place higher than before : hence we get a 

 diminished current for low positions of the platinum strip, 

 but a considerable extra current above. 



§ 9. Theoretical. 



It has now been clearly established that when metallic 

 vapour is present in a flame there exist positive carriers of two 

 classes, distinguished by their having very different velocities 

 of migration, the velocities of the one class being about 



cm. /volt ag compared with . 04 S /Y^Llor the other 

 sec./ cm. f sec./ cm, 



class in the case of strontium. Both kinds of carriers are 

 metallic ; the slower kind form a luminous streak. The 

 faster kind produce no visible coloration in the flame when 

 driven sideways out of the luminous streak ; this is to be 

 anticipated, since they are present in relatively very small 

 numbers, as can be deduced from the time required to 

 deposit a quantity of metal on the strip sufficient to produce 

 strong coloration in the flame. The faster carriers are pro- 

 duced throughout the height of the streak, and hence their 

 presence is not connected with the contact of glowing 

 platinum, but inherent in the luminous vapour itself. 



As a result of much experimental and theoretical work, 

 Lenard has put forward with strong support the view that 

 any single carrier of the class whose presence is detected by 

 the visible deflexion of the luminous streak is not positively 

 charged always, but only for a short fraction of the time *., 

 We may now explain the difference of velocity of the two 

 classes by supposing that in both cases the carrier is only 

 charged for a fraction of the time, alternating the positive 

 state with the neutral state, but that the fraction is very 

 much greater for the one than for the other. We can cal- 

 culate the theoretical velocity of a permanently charged 

 carrier of atomic size from the formula given by Lenard 

 already cited (p. 871), which has proved very successful f, 



* E. g., P. Lenard, Ann. der Physik [4] ix. 1902, p. 648, and [4] xvii. 

 1905, p. 238. 



t For a short account of the formulae for the mobility of carriers in 

 a gas, see A. Becker, Ann. der Physik [4 J xxxvi. p, 217 (191 1). 



