156 DR. FRANK HORTON ON THE DISCHARGE OF 



The variations in Q are not greater than can be accounted for by the errors of 

 experiment. The mean value is lower than that given by WILSON (viz., 131,100), but 

 this is probably due to the fact that the present experiments were performed with a 

 gas pressure of several millimetres in the apparatus, whereas WILSON'S result was 

 obtained in a good vacuum. 



Taking Q = 121,900 and the current per square centimetre at 1610 C. as 

 5 '91 x 10~ 7 ampere, the value of the constant A in the formula x = A^e"^ 2 * is 

 1'55 x 10', so that the equation for the current x (amperes) at the absolute 

 temperature 6 becomes 



The following is a comparison of the currents calculated by means of this equation 

 and those found experimentally. 



TABLE III. 



Temperature, Negative leak in amperes per square centimetre. 



C. Observed. Calculated. 



1331 l-95x!0- 1- 96xlO- 9 



1468 3-96xlO- 4-06x10-" 



1542 1- 75xlO- 7 -l-73x!0- T 



1571 2-99x10-* 2-95x10-' 



1610 5-91 xlO- 7 5-91 xlO' 7 



The observed and calculated values of the current agree very well, showing that 

 the formula expresses the experimental results with considerable accuracy. 



(3) The Negative Leak from Calcium in Helium. 



Having shown that the negative leak from the platinum strip was of the normal 

 amount, and that it varied with the temperature according to the established law, 

 calcium was sublimed on to it and the alteration of the leak caused thereby was 

 investigated. The sublimation of the calcium was performed by connecting the thick 

 copper leads of the calcium wire (F, fig. 1) to the alternating current from a 

 transformer and gradually decreasing the resistance in the circuit until the calcium 

 became red hot. It then sublimed, and the bulb was covered with a fine metallic 

 mirror, and the electrodes would be similarly covered with calcium. With practice it 

 was possible to regulate the current so that the wire did not fuse through on the first 

 heating. After observations of the negative leak had been taken, more calcium could 

 be sublimed on to the cathode, and the observations repeated. The gas pressure in 

 the apparatus increased during the process on account of the gas evolved by the 

 calcium. The discharge was therefore started in the potassium-sodium tube and kept 

 going until the whole of the evolved gases were absorbed by the alloy and the 

 apparatus contained only helium gas at the same pressure as before. In order to 

 see if the evolved gas increased the leak from the platinum strip, an experiment was 



