NKOATIVE KLHCTBICITY FROM HOT CALCIUM AND FROM LIME. 15 



currents measured decreased rapidly as the heating of the cathode was continued, 

 especially at the highest temperatures. 



TABLE VIII. 



Temperature, Negative leak Temperature, Negative leak 



C. in amperes. C. in amperes. 



895 9-85xlO- 1380 4-5xlQ- 



1013 5-69xlO- 1465 1-4x10-* 



1220 3-12xlO- 1535 2'7xlO- 



1293 1-2 xlO- 1620 4'7xlO- 



It is thus seen that the negative leak from lime is considerably increased by intro- 

 ducing hydrogen gas into the apparatus. H. A. WILSON has shown that hydrogen 

 greatly increases the negative leak from platinum, and has come to the conclusion 

 that the negative leak from platinum in air, or in a vacuum, is almost entirely due to 

 traces of hydrogen in the metal. WILSON reduced the leak to j 5 ^ 66 part of that 

 oterved by RICHARDSON by taking precautions to remove such traces. It should be 

 mentioned that the currents in the above table are much larger than the negative 

 leaks from platinum in hydrogen obtained by WILSON, and the increase of current 

 cannot be merely due to the effect of the hydrogen upon the platinum. 



In another experiment a lime-covered cathode in oxygen at a low pressure 

 (0'002 millim.) with a potential difference of 40 volts gave a negative leak of 

 4'6x 10~* ampere at 965 C. On pumping out the oxygen and letting hydrogen 

 into the apparatus and then pumping down to the same pressure as before, the 

 negative leak at temperatures below 900 C. was only slightly greater than the leak 

 at the same temperature before the hydrogen was admitted, but at 980 C. a faint 

 luminosity was seen in the gas round the cathode, and the negative leak increased 

 tvithout the temperature of the cathode being raised t>r the difference of potential 

 between the electrodes being altered. This increase was slow for a few minutes, but 

 afterwards became more rapid, and, although the temperature was lowered by 

 putting more resistance in the heating circuit, the negative leak increased to 60 milli- 

 amperes at 885 C. The luminous glow was then quite bright, and filled the whole 

 bulb. It was at first thought that this sudden increase in the negative leak was due 

 to the temperature of the cathode increasing while the discharge was passing, but 

 experiments showed that the temperature of the cathode went up only a few degrees 

 when the electric field was put on, and the leak gradually increased, even though the 

 temperature, as indicated by the thermo-juuction, was diminished by putting extra 

 resistance in the heating circuit. 



On allowing the cathode to cool down, and then again testing at lower tempera- 

 tures, it was found that the negative leak at these lower temperatures was now much 

 greater than at the first observations. A measurable leak was obtained at a much 

 lower temperature than before, and the leak at 740 C. was 4'6x 10~* ampere about 

 10* times as large as before the glow had been obtained in the discharge tube. On 



