﻿884 On the Carriers of Electricity in Flames. 



positive carriers are metal atoms, as my experiments have 

 shown, so far as their accuracy allows. But as the obser- 

 vations of the brightness of the Sr lines produced by the Sr 

 deposited on the platinum strip cannot be carried out very 

 exactly, we can only assert with certainty, that the carriers 

 cannot possibly consist of more than two atoms. We have 

 very good grounds for assuming that the light emission is 

 effected by single atoms, on which assumption Lenard has 

 based his researches on the subject *. The results of experi- 

 ments on the diffusion of sodium vapour in the flame also 

 indicate this f . It is, however, not absolutely certain that 

 the faster carriers are centres of light emission : they are 

 present in such relatively small numbers that they could not 

 be expected to be visible. 



§ 10. Conclusions. 



(1) The positive carrier in a flame containing metal 

 vapours is of metallic nature, very probably the metal 

 atom. 



(2) The positive carrier has two different definite velocities 

 of migration in the electric field, according as it is present 

 in the streak of metallic vapour or in the free flame : in both 

 cases it alternates the positive with the neutral state. In the 

 former case it is charged for about ¥ J- D of the time, in the 

 latter for J of the time. 



(3) For strontium the greater velocity of migration is 



rt > cm. /volt mi , , 



2'i) / • — lnere seems no reason to doubt the results 



sec./ cm. 



of other experimenters, that the velocity is the same for 



different metals, which agrees with our theory. Hence this 



value probably gives the velocity of migration of an atom 



of any metal in the free flame. 



(4) The presence of glowing platinum in the vapour causes 

 the liberation of electrons from it in large quantities. 



(5) There exist, besides the free electrons, negative carriers 

 of the second kind which are metallic, in relatively small 

 numbers. This gives an analogy with the canal rays. 



This work was carried out in the Radiologisches Institut 

 of the University of Heidelberg, to the Director of which, 

 Professor P. Lenard, I am indebted for much help and 

 encouragement. I also wish to express my thanks to 

 Dr. Ramsauer for assistance and many valuable suggestions. 



* P. Lenard, Ann. der Pkysik [4] xvii. p. 198 (1905). 



+ A. Becker, " Uber die Diffusion leuclitender Metalldampfe," 

 He ide Wej -f/er Acad. 1911. 



