482 
DE. HAEOLI) A. ^VILSON OX THE 
compared with that of the salt, but as the tube got aged the conductivity of the air 
fell off to such an extent that it became quite small compared with that of the salt 
when solutions containing 10 grammes in a litre were sprayed. 
Diagram No. 11 shows the variation of the current with the temperature when 
spraying a solution of KOI, done when the tube wms nearly new\ 
The conductivity due to the KCl has a maximum value at about 800°. The 
current due to the salt is greater when the outer tube is negative, while the reverse 
holds good for air, so it is best to have the outer tube negative when measuring the 
conductivity of salt vapours. The dotted curves drawn w^ere got by subtracting the 
current due to the air alone from the current due to Ijoth salt and air. 
In all the measurements given below the outer tube wms negative, and the con¬ 
ductivity of the air alone was negligible compared wdth that of the salt. 
Diagrams Nos. 12, 13, and 14 show the variation of the current wdth the 
temperature when a solution of KI containing 10 grammes in a litre, was sprayed 
wdth E.M.F.’s of 800, 100, and 40 volts respectively. 
Diagram Xo. 12. 
It is evident from these curves that the relation between the conductivity of salt 
vapours and tlie temperature is of a much less simple character than is the case 
for air alone. 
As the temperature rises the conductivity at first increases rapidly and then 
attains a maximum value at about 900° C., above wdiich it falls again until about 
1150° C. is reached, when it suddenly begins to rise very rapidly. Above 1200° C. 
the current again becomes nearly independent of the temperature. 
With an E.M.F. of 800 volts, the sudden rise at 1150° just doubles the current, 
which suggests that at this temperature some kind of dissociation occurs and doubles 
the number of ions available. 
The amount of salt passing through the tube wars determined by substituting for 
