(58 
PROF. H. A. WILSON ON THE ELECTRICAL CONDUCTIVITY AND 
Hence 
which gives 
dx — dx' 
-Xdx, 
v 
x—x' — - f X dx. 
V Jo 
But X dx denotes the potential difference between the lower grating and the point 
Jo 
a distance x higher up. Let this potential difference be denoted by V so that 
Jc = v{x-x')/V. 
Here x—x' is the distance through which a puff of salt appears to be moved 
downwards, as seen through the rotating disk, when the lower grating is connected 
to the battery. 
Since v — n {x 2 —x^) we have 
^ _ 7i (x.j—Xj) (x—x') 
If the puffs contain much salt vapour they have a greater conductivity than the 
rest of the flame so that the electric field will be less in the puffs than in the rest of 
the flame. In this case the salt inside the puffs will not be acted on by the whole of 
the potential fall, Y, so that the deflexion of the puffs will be smaller than that 
represented by the equation just given. In the case of sodium and strontium salts, 
which have small conductivity and intense luminosity, it was possible to observe the 
puffs when the amount of salt in them was not enough to change the conductivity 
appreciably so that with these salts no error could be produced by the conductivity of 
the puffs. The fact that the conductivity with small amounts of these salts was not' 
appreciable was shown by measuring the current with the stop-cock open all the time 
so as to get a continuous streak of the salt vapour. With small amounts of the salts 
the current was not increased by the salt more than 50 per cent. With the 
apparatus described it was found that the puffs of salt vapour did not appear to move 
at all when the battery was connected or disconnected. A motion of 0‘5 mm. could 
have been detected, and Y was increased to over 2000 volts in the case of some salts. 
This shows that k was less than 
100 x 3 x 0'05 
2000 
0‘0075 cm. per second per volt per centimetre. 
Salts of sodium, strontium and potassium were tried and no deflexion of the puffs 
of vapour could be detected in any case. The amount of salt in the puifs was varied 
as much as possible in each case. With sodium and strontium Y was increased to 
over 2000 volts, but with potassium it could not be raised above about 1000 volts 
without an arc striking between the electrodes. With sodium and strontium in 
