410 Dr. Richardson on the Discharge of Electricity 



it was found to contain one or two yellowish patches in the 

 part which had been converted into phosphoric acid by the 

 absorption of water. This led to the suspicion that the effect 

 was not due to phosphorus pentoxide, but to the lower oxides 

 or to phosphorus itself. A new bulb was therefore set up 

 containing the best phosphoric anhydride on the market. 

 This was found to give an effect which, though easily capable 

 of detection, was insignificant compared with that previously 

 obtained ; dearly showing that the effect is not due to P 2 5 

 itself. A blank bulb containing nothing at all was also 

 tried, and found to give no effect even when heated to 400° C. 

 The pure P 3 5 was heated to about 270° C. and, subliming 

 over the apparatus, caused the gas-pressure to rise to about 

 4*0 mm., but only produced a slight increase in the positive 

 leak. It was thought that possibly light as well as heat 

 would be able to drive off the stimulating vapour ; and the 

 effect of an arc-light on the original bulb was tried, but no 

 effect was obtained. 



An experiment was next tried in which a small quantity of 

 phosphorus itself was introduced into the apparatus. A piece 

 of yellow phosphorus was placed in a small bulb which was 

 sealed on to the apparatus beyond the pentoxide-bulb (see 

 figure) . It was found that the wire, when set up under these 

 conditions, exhibited a big initial leak. This result cannot, 

 however, be regarded as decisive ; since it is often found that 

 the wire possesses an abnormal emissibility for positive elec- 

 tricity after any great change has been made in the apparatus. 

 The author hoped to obtain an unobjectionable result by 

 heating the wire until as much as possible of this initial leak 

 had been destroyed, and then heating the lump of phosphorus. 

 It was thought that any increase in the leak thus produced 

 could only be due to the effect of the phosphorus. Unfor- 

 tunately, the action of the vapour from the cold phosphorus 

 on the hot wire was so great that the wire came to grief long- 

 before any readings could be taken. 



The above experiment made it evident that to be decisive 

 this test must satisfy tw T o conditions : — (1) It must be capable 

 of being made very rapidly ; and (2) the only cause tending 

 to give an increase of ionization at the moment of making the 

 test must be the presence of phosphorus vapour. To satisfy 

 these conditions, the apparatus was arranged so that the 

 platinum wire could be heated in a vacuum quite independent 

 of that containing the phosphorus vapour. In this way, the 

 initial leak could be got rid of and the wire brought to the 

 condition for making the test without the wire being rotted 

 by the action of the phosphorus vapour. Then, on turning a 



