Behaviour of Metallic Vapours in Flames. 17 



wander towards the negative side ; hence the concentration of 

 negative electricity grows from the positive to the negative 

 side, and the probability of an atom being positively charged 

 decreases. If there is only a little metal present, then 

 firstly the streak is narrow, and secondly there are few free 

 electrons. But if much metal is vaporized, then, with the 

 broad streak and the many free electrons which wander 

 through it, we may have large differences of concentration 

 between the two sides of the streak, until finally, many 

 metallic atoms in the negative side may become negatively 

 charged for a fraction of the time. 



§ 3. Methods of Measuring the Conductivity. 



The method first adopted for measuring the conductivity 

 was to bring two small parallel electrodes of platinum-foil into 

 the body of the metallic vapour, and to measure the current 

 sent through this resistance by potentials of from to 100 

 volts. With a given amount of salt vaporized, the curve of 

 current against potential was found to be of parabolic form, 

 as has already been shown by Wilson *. As this form of 

 curve does not offer readily a constant which measures the 

 conductivity, this method was abandoned. 



The electrodes were then made larger and placed further 

 apart, usually at a distance of about 1*2 cm., their dimensions 

 being 0*8 x 2 cm. The vapour was introduced between the 

 electrodes as a luminous streak from a small bead ; it did 

 not touch either electrode. In this case the curves of current 

 against voltage are straight lines, or Ohm's law holds within 

 the observed limits, and a definite conductivity may be attri- 

 buted to the salt-vapour for a given portion of the streak. 

 The current spoken of above is the difference of the current 

 through the pure flame, and the flame containing the salt. 

 The curve of current against voltage for the pure flame is 

 roughly parabolic. 



Although we thus get a definite conductivity for a given 

 position of the streak between the electrodes, this conductivity 

 is changed by changing the position of the luminous streak 

 relative to the electrodes, although these be kept fixed at a 

 constant distance apart. As we bring the streak nearer to 

 the negative electrode the current rapidly increases ; if the 

 vapour is in visible contact with the glowing electrode the 

 current becomes comparatively enormous. The curves in 



* H. A. Wilson, Phil. Mag. [6] x. p. 476 (1905). 

 Phil. Mag. S. 6. Vol. 24. No. 139. July 1912. 



