688 Prof. 0. W. Richardson on the 



as the pressure is reduced to atmospheric. At atmospheric 

 pressure the current was less than 10~ 9 ampere. The first 

 explanation which occurred to me was that the actunl 

 pressures were not so low as those measured in the tube 

 experiment owing to the evolution of gas in the apparatus 

 and leakage through the iron tube under the influence of 

 heat. This explanation does not appear to hold, however, for 

 the measured pressures were as low as *005 mm. when the 

 pump was working continuously, and it took several minutes 

 for enough gas to accumulate after the pump was stopped to 

 show a pressure of O'l mm. ; whereas the maximum observed 

 by Garrett at 1005° C. was at about 1*6 mm. and would 

 probably be at a still higher pressure at the lower temperature 

 of these experiments. It is important also to remember in 

 this connexion that the experiments with the empty tube had 

 indicated that the gas which leaked through the hot steel 

 crucible was probably small in amount compared with that 

 evolved by the heated salt at these low temperatures, so that 

 the larger leak at lower pressures is not likely to have been 

 caused by the peculiar properties of such gases. On taking 

 down the apparatus after this experiment there was no 

 deposit in the inner electrode, but the apparatus had a faint 

 smell reminding one of that of soot. 



Beryllium Sulphate, BeS0 4 . — Experiments were made on 

 a specimen of this salt which was kindly furnished by 

 Mr. Parsons and which was supposed to be especially pure 

 from a chemical standpoint. The value of efm for the positive 

 ions was determined by Mr. Davisson and found to corre- 

 spond to potassium. The emission investigated is presumably 

 that from beryllium oxide since the sulphate loses tS0 3 and 

 is converted into the oxide at high temperatures. The 

 current-E.M.F. curves were similar to those given by 

 aluminium phosphate. 



The effect of pressure on the emissium from beryllium 

 sulphate is shown by the full curves in fig. 4. The actual 

 values plotted were obtained with -f 40 volts on the tube, 

 but precisely similar results were obtained with +160 volts. 

 The temperature rose gradually from 822° C. to 839° C. 

 during this experiment. The unit of current is 2*5 X 10~ 9 

 amp. The main features of the curves are a small drop in 

 the emission wiih increasing pressure at low pressures leading 

 to a minimum value in the neighbourhood of 1 mm., followed 

 by a gradually diminishing rate of increase to much higher 

 values at higher pressures. There is no indication of a 

 pressure at which the emission is a maximum under 25 milli- 

 metres. Experiments were not made at pressures higher 

 than this. 



