﻿of Light by Metallic Particles. 451 



of platinum particles. Platinum is certainly one of the most 

 opaque, if not the most opaque metal ordinarily dealt with ; 

 and it seemed possible that very fine particles of platinum 

 might act as conductors even to electric disturbances reversed 

 with the frequency of blue light. 



A sky of platinum particles was easily produced as follows: — 

 A litre bottle with a wide neck was three quarters filled with 

 boiling distilled water, the water having boiled for an hour 

 before being poured into the bottle. The rubber bung used 

 to close the bottle was pierced by three holes — one allowing 

 the introduction of a thick glass tube carrying a sealed-in 

 electrode and passing into the water ; another carried a 

 second electrode of platinum reaching down to within about 

 half an inch of the water surface ; the third carried a tube 

 permitting communication to be made between the inside of 

 the bottle and the water-pump. On placing the bung in 

 position and working the pump, most of the air was rapidly 

 sucked from the bottle, the water began to boil, and after a 

 time the connexion was sealed-ofT. The bung and tubes were 

 well waxed, and the bottle was left to cool down. Next day 

 it was found that the water " hammered " in a satisfactory 

 manner when the bottle was shaken. The electrode, owing 

 to several causes, was now about 1^ centim. from the surface 

 of the water. An induction-coil worked by a transformer was 

 then caused to pass a current between the water surface and 

 the free electrode, the current being regulated so that at a 

 frequency of about 60 ~ it just failed to produce sufficient 

 electrolysis to free gas. The electrode was white hot. The 

 bottle was placed before an arc-lamp and studied by a Nicol 

 prism, as previously described, while the coil was actually at 

 work. It was noted that before the discharge began the water 

 was nearly optically clean, and looked exactly like a check- 

 sample of the boiled distilled water which had been preserved 

 for the purpose of comparison. After about half an hour's 

 discharge no appreciable increase in the scattered light could 

 be detected ; but on shaking the bottle, small brightly- 

 reflecting metallic scales became apparent. The discharge 

 produces considerable motion of the water-surface, and it 

 had been erroneously assumed that this would prevent the 

 formation of a platinum film on the water surface. The hint 

 was taken, however, and the bottle was shaken by hand for 

 about two hours while the discharge w T as taking place, during 

 the whole of which time the scattering of light by the water 

 particles continually increased, and was finally fairly brilliant. 

 At no time did the direction of maximum polarization appear 

 to vary from the perpendicular to the direction of propagation 



