﻿452 Prof. Threlfall on the Scattering 



of the light. The platinum sky behaved, in fact, in a quite 

 normal manner. The more carefully the directions of the 

 incident and observed light were adjusted to perpendicularity, 

 the more complete the extinction produced by rotating the 

 Kicol appeared to be. The fine particles remaining well sus- 

 pended for a considerable time (at least a week), several 

 friends were good enough to examine the polarization phe- 

 nomena for me, and all the observations agreed perfectly. 



Platinum particles therefore scatter light like ordinary non- 

 conducting particles within the limits of accuracy of the 

 observations. 



I was anxious to try the effect of a sky of iron particles, 

 and made the following experiments and observations on this 

 matter. 



The apparatus used was similar to that described in the 

 case of platinum, except that " pure " iron electrodes were 

 formed by binding the iron wire sold as " pure for chemical 

 analvsis " round the sealed-in ends of the platinum wires. 

 Five or six layers of fine iron wire were bound round the 

 platinum wire, and connected then to the actual electrodes — 

 two short lengths of Swedish charcoal-iron carefully cleaned. 

 Sparks were taken between the two electrodes, and were 

 practically confined in their effects to the charcoal-iron. As 

 a medium for catching and supporting the fine iron-dust, I 

 first tried a nearly dry sample of glycerine, prepared by 

 heating the strongest commercial glycerine in the presence of 

 strong sulphuric acid under reduced pressure. After sparking 

 for some hours a sky was formed, but the heat of the discharge 

 had been sufficient to obviously decompose some glycerine 

 vapour, and consequently the sky might have been due 

 to iron, oxide of iron, or carbon, or any mixture thereof. I 

 therefore tried to find some more suitable substance than 

 glycerine. I wished to exclude oxygen, if possible, from the 

 medium employed, and yet have a substance which should 

 have only a small vapour-tension. After a good many trials 

 I settled on a vaseline oil (?) known as "albocarbene oil." 

 The vapour-tension of this substance was at least 2 centim. 

 at 24° 0. even after heating and exposing in a vacuum several 

 times. Under the circumstances it hardly appeared worth 

 while to attempt to take up the oxygen it might hold by the 

 action of sodium, especially as I was not sure of the complete 

 action of the sodium ; while the certainty of its making the 

 oil dirty opened up the probability of other disadvantages. 

 The atmosphere employed was hvdrogen disengaged from 

 zinc and hydrochloric acid in presence of chromous chloride ; 

 it was washed by solution of sodium hydrate, and finally dried 



