386 



Dr. A. LI. Hughes on the 



Phosphorus Pentoxide. 



A very thin transparent layer of P 2 5 was sublimed on 

 to the nickel plate. It was exposed to the light for two 

 hours, yet the leak, if it existed, was always less than *1. On 

 admitting atmospheric air at 1 cm. pressure and pumping 

 out immediately, the leak went up to 4. Air at 760 mm., 

 from the room, was admitted several times. The biggest 

 leak obtained after such treatment was 70. Similar effects 

 were obtained when the deposit of P 2 5 was thick enough to 

 be opaque. 



It is difficult to see why the ZnCl 2 and P 2 5 , which are 

 not photo-electric in themselves, become so when in combi- 

 nation with water. Water, according to Brillouin *, does 

 not emit electricity whon exposed to ultra-violet light. The 

 act of solution in water can hardly be regarded as causing 

 the liberation of a photo-electric element in any way analogous 

 to the surface decomposition of a mercury salt by ultra-violet 

 light. In tbe case of the phosphorus pentoxide and water, 

 not one of the elements involved has been shown to be photo- 

 electric when illuminated by the light from a mercury lamp. 



Water. 

 The arrangement shown in fig. 2 was devised to test the 



Fisr. 2. 



+ ^©o volts 



photo-electric properties of water. The water was contained 

 in a shallow dish (the lid of a circular tin). The two rings 

 of brass a and b were arranged as shown to prevent strav 

 light reaching the edges of the dish, since the metal has an 

 enormous photo-electric effect compared with compounds in 

 general. The rings a and b and the top-piece to which they 

 were soldered were covered with soot. Between the edges 

 of the rings a and b and the surface of the water, there was a 

 * Sir J. J. Thomson, ' Conduct, of Elee. through Gases,' p. 289. 



