TRANSACTIONS OF SECTION A. 533 



Robert Ball's observatory at Cambridge in continuation of the results which the 

 late Professor FitzGerald, Mr. W. E. Wilson, and I obtained in Mr. Wilson's 

 observatory in Westmeath. 



It is desirable that I should give a notion of the magnitudes of the voltages 

 developed in these cells by lights of various intensities. 



(1) An ordinary paraffin candle, held at a distance of 2 feet from a cell in 

 which the liquid was malonic ether, was found to give slightly more than "25 of 

 a volt. 



(2) For small intensities of the incident light the voltage will be proportional 

 to the square root of the intensity ; that is to say, the voltage varies inversely 

 as the distance of the luminous source from the cell. 



(8) A paraffin candle, at a distance of 8 feet from the cell, gives a voltage 

 almost exactly equal to that given by the light of Vega when this is concentrated 

 by a reflecting telescope whose aperture is 2 feet. 



(4) For strong lights the law that the square of the voltage developed in the 

 cell is proportional to the intensity of the incident light does not hold, as is shown 

 by the following observation recently made : — 



Room darkened and cell in the dark, except that the light of the paraffin lamp 

 of the electrometer was, to some extent, reflected from the walls of the room : this 

 very feeble light gave a deflection of 11 divisions on the scale. One candle held at 

 2 feet from the cell gave (not allowing for the above 11 divisions) 76 divisions. 

 Two candles held close together at 2 feet from cell gave (not allowing for the 

 11 divisions) 99 divisions. (One volt was represented by 275 divisions.) 



Let i be the energy incident on the sensitive plate when nothing but the 

 reflected lamplight falls on it ; let I be the incident energy due to one candle at 

 2 feet, and 1' that due to two candles at 2 feet ; then we have 



\ + i = hy. 76- 

 I' + ;. = 7^x99 - 

 These gi V9 



I'^99'-1P ■■.^, 

 i 76=- IP ' 



but I' should be 21, so that the law cannot hold. Indeed, diffused daylight itself 

 develops only about '5 volt in the cell. 



These cells are sensitive to all parts of the spectrum, the voltage developed 

 by the yellow being slightly greater than that due to the other parts. 



As to the nature of the action of light in a selenium cell, 1 may quote the 

 following interesting experiment which was made by Mr. Shelford Bidwell, and 

 communicated to me in a letter recently. 



Mr. Bidwell took a piece of platinum foil and coated it by electrolysis with a 

 very thin layer of selenium by making it the cathode in a solution of selenious 

 oxide, or of Se dissolved in HNO.,. The platinum foil, thus covered with red Se, 

 was gradually heated on a brass plate and thus brought into the well-known con- 

 dition in wliich it is sensitive to light. 



When this coated strip was immersed in a beaker of tap water in presence of a 

 clean platinum strip, there was little (if any) voltage in the dark; but when 

 difl'used daylight was allowed to fall on the coated strip a voltage of -lOl was 

 developed. (This was very much less than the voltage which would have been 

 developed in the cells which I have just described ; but the reasons for this are 

 obvious.) 



In this cell, as in all other forms of selenium cells, the selenium plate was to 

 the inactive plate as copper to zinc, and from this Mi". Bidwell concludes that, 

 just as Zn tends to combine with oxygen in H^O, so Se in the light tends to 

 combine with hydrogen and form HoSe. 



In order to test this, Mr. Bidwell took two small test tubes, and into each he 

 put some acetone and a strip of platinum coated with selenium ; each tube was 

 closed by a vaselined cork, and from each cork was suspended in the tube a clean 



1901. N N 



