690 Prof. 0. W. Richardson on the 



In the previous experiments the length of time the salt 

 had been heated before the observations were taken was not 

 recorded. It varied from 4 to 8 hours,, according to the 

 convenience of the observer and the time required for steady 

 conditions to be obtained. In the present experiments a 

 record was kept of the time which elapsed in this way. 



Current-E.M.F. curves were taken after about 10 hours' 

 heating. They showed a better approach to saturation than 

 any which had been obtained previously as the following 

 numbers testify. 



Voltage 40 80 120 160 200 240 320 400 



Current pre SSUre 1 



= O'OUlo nun. J 



do. 0-8 mm 1 1-03 Ml 113 1 21 124 1*42 160 



do. 9-4 mm 1 M2 M8 125 1-31 1*36 146 



The currents which are given in the three lower rows have 

 been arbitrarily reduced to equality at 40 volts. The order 

 of magnitude of the unit is 10~ 7 amp. Although the 

 temperature was only 785° C in this experiment it was 

 observed that there was a negative emission of about 5 x 10 ~ 7 

 amp. This negative emission is very much larger than the 

 platinum alone could give rise to at that temperature, and it 

 seems probable that the heating in a vacuum had partially 

 converted the barium sulphate into barium oxide which was 

 giving rise to the large negative emission which Wehnelt 

 showed to be characteristic of that substance. It is curious 

 that the positive emission should be easier to saturate in the 

 presence of this large negative emission, but I have frequently 

 noticed that the current saturates more easily after continued 

 heating. This is one of several features of the relation 

 between the current and electromotive force in the case of 

 heated salts which seems worthy of further investigation. 



The positive emission after the barite had been heated for 11 

 hours is shown in curves 1 and 2 of rig. 5. Curve 1, points 

 thus, x, corresponds to increasing, and curve 2, points 

 thus, ©, to diminishing pressures. The temperatures in this 

 experiment fell steadily from 783° C. to 772° 0. This curve 

 is quite different from the one obtained with the chemical 

 barium sulphate. I think, however, that the difference is 

 due to the difference in the previous history rather than to 

 the different composition of the material, as the same sample 

 of barite later on gave a curve which was similar to those 

 which had been obtained from the barium sulphate. Curves 

 1 and 2 show a greater sensitiveness to changes of pressure 



