from Hot Platinum in Phosphorus Vapour. 409 



At the end of an experiment of this kind, it was found 

 that the wire generally broke at some point, and on more 

 minute examination it presented a beaded appearance. This 

 was due to the formation o£ globules of a fusible substance 

 which collected at regular intervals on the wire. Evidently 

 a vigorous chemical action had occurred between the wire and 

 the surrounding vapour ; the alteration in the properties of 

 the wire caused by this prevented any very accurate 

 measurements being taken. 



It has been pointed out that this effect was first detected 

 owing to a wire which had been left cold inside the 

 apparatus gradually recovering its lost discharging power. 

 As was to be expected, this phenomenon was shown in a 

 higher degree if the pentoxide-bulb was heated. In one 

 case, after the bulb had been heated and the wire allowed to 

 remain cold for several days, the leaks at different times 

 were found to be those given in the following table : — 



Table II. 



Time wire heated [ 

 (minutes) j 







3 



6 



9 15 



25 



30 



36 



42 



46 



52 



53-5 

 29 



55-5 

 7T 



58 

 1-9 



61 



1-8 



Positive leak 



14286 



57 



36 28 



1 



17 



11-5 



6 



7 



5-2 









The effect appears to fall off with time in the same sort of 

 way that the positive leak from a new wire does. The 

 striking alteration which gradually takes place in the 

 properties of the wire, leads one to attribute the falling off in 

 the leak rather to an absorption of the vapour by the wire, 

 than to an emission of something previously absorbed. The 

 wire was not subjected to quite the same conditions during 

 the whole of the time represented by the last table. After 

 taking the readings corresponding to 30 and 46 minutes, the 

 temperature of the wire was raised, for the space of about 

 two minutes, from about 700° 0. to about 1200° C. It was 

 subsequently reduced to the previous temperature, and the 

 readings taken as usual. It will be noticed that in each case 

 the leak starts by having a lower value than what we should 

 have expected if the wire had not been heated ; the value 

 subsequently increases for the space of two or three minutes, 

 and finally decreases with time as before. These results 

 can be reconciled with the view that the effect is due to the 

 absorption of the vapour by the metal, if one supposes that the 

 rare of absorption is greater at high than at low temperatures. 



On examining the pentoxide-bulb which had been heated, 

 Phil. May. 8. 6. Vol. 9. No. 52. April 1905. 2 E 



