200 Prof. J. A. Cunningham on the 



as in fig. 3. From these curves, for a temperature of about 

 1140° 0., we see that the effect of increasing the current is to 

 continuously increase the steepness of the potential-gradient 

 at the cathode, and at the same time steadily to drive back 



Fig- 3. 



R T 



Temperature if 1 =-C»2G (= 1140° C 



) Pressure =0'55 mm. 



9 8 7 6 



D/stance on Scale (C/ith ode at d^/o-s /hvooE fir c< o-s cms) 



the positive column towards the anode. Exactly similar 

 results were obtained at other temperatures. This driving 

 back of the positive column was observed by Skinner *. It 

 is also exemplified for high temperatures in Table I., where 

 the bright bands seemed to disappear into the cathode. 



My principal attention was, however, given to variations in 

 value of the Cathode Fall of Potential accompanying changes 

 in the current through the tube, and in the temperature of 

 the cathode. Such a series of measurements is plotted in 

 fig. 4 (p. 202), which shows the connexion between current 

 and cathode-fall for a number of high temperatures. (On the 

 same scale all those for low T er temperatures would be crowded 

 together as almost vertical lines, making only a very small 

 angle with the axis of potential. In order to be able to 

 include the measurements at the highest temperature, I have 

 had to contract the axis of current by cutting out certain 

 portions along which the slope of the curve was practically 

 uniform.) The ratios of the resistances of the filament hot to 



* Phil. Mag. [5] vol. 1. p. 563 (1900). 



