QS2 Prof. 0. W. Richardson on the 



in the platinum tube. Before each experiment the tube was 

 cleaned by boiling in hydrofluoric acid, nitric acid, and dis- 

 tilled water in succession. Similar treatment was also 

 accorded to the platinum strips which were used in the other 

 type of experiment. Current-E.M.F. curves taken at a 

 temperature of about 940° 0. and a series of different pres- 

 sures gave the numbers in the following table: — 



Volts 40 80 120 160 200 240 280 320 360 



Pressure =-0075 mm 1 1-28 1-45 157 1-68 177 190 213 231 



do. =about -5 mm. ... 1 1*25 1 -44 154 161 190 245 3'47 5 



do. =5-5 mm 1 1-29 1*53 171 1*80 183 1'83 2-5 15 



The three lower rows give the thermionic currents (positive 

 emission) at the respective pressures when the voltage applied 

 is that at the top of each column. The unit of current is 

 about 10 ~ 7 amp. Saturation was never reached, strictly 

 speaking, in these experiments, as the current always in- 

 creases slowly with increasing voltage. The mode of variation 

 does not differ much at different pressures for voltages 

 under 200, so that it seemed reasonable to suppose that a 

 comparison of the thermionic currents at a constant voltage 

 in the range between 40 volts and 200 volts would be a fair 

 way of testing the thermionic emission at different pressures. 

 In each test of the pressure variation, experiments were 

 made both with 40 and also with 1.60 volts applied potential 

 difference, in order to make sure that the applied voltage was 

 not an important factor in the results. With the empty 

 platinum tube the pressure of the air had very little effect 

 on the positive emission, either with 40 or 160 volts applied 

 potential difference. A specific test at 945° C. with 160 volts 

 gave a current of 21*7 divisions (1 div. = l*25 x 10~ 8 amp.) at 

 0*01 mm. pressure, and 21*8 divisions at 12*0 mm. pressure. 

 The results obtained with the lower voltage and with changes 

 to intermediate pressures were similar, except that slightly 

 larger increases were usually observed. The increases might 

 always have been accounted for as experimental errors, but 

 it is significant that out of 6 sets of observations every one 

 showed a slight increase with increased pressure. This result 

 is consistent with the author's earlier experiments* on the 

 ionization from pure platinum in air, according to which 

 the effect of the air on the pure platinum itself should have 

 called for a current of one-half to one division in the case of 

 these observations. It is clear that in these experiments the 

 emission from the empty platinum tube was greater than 

 that which would have arisen from pure platinum at the 

 * Phil. Trans. A. vol. ccvii. p. 40 (1906). 



