1895.] The Electrical Measurement of Starlight. 145 



portions which can be metallically joined together outside the cell or 

 kept separate. If, while the portions are joined, one is exposed to, 

 and the other screened from, incident light, the E.M.F. indicated is 

 much less than it is when both are exposed, or when one alone is 

 exposed while the other is disconnected from it. (A description of 

 such tinfoil cell will be found in the ' Philosophical Magazine ' and 

 in the ' Proceedings of the Physical Society.') 



This fact now enables us to see the importance of preventing the 

 liquid from entering the glass tube AB, which contains the conducting 

 wire P, for it is clear that, when the light is incident at A, the liquid 

 which has crept into the tube round the aluminium wire will convey 

 a portion of the negative charge imparted to the liquid in the cell to 

 the wrong pole of the electrometer, and will thus diminish the effect 

 of the light. 



The capillary entrance of the liquid into the tube AB may, of course, 

 be prevented by sealing into the tube a platinum instead of an 

 aluminium wire, and coating the end of it with the selenium layer. 

 But, unfortunately, platinum is not so good a base for the selenium 

 as is aluminium, owing, almost certainly, to the fact that selenium 

 enters into chemical composition with platinum, while it does not do 

 so with aluminium, or with some other metals which, possibly, may 

 yet be used. 



The entrance of the liquid could also be prevented by using a 

 platinum wire instead of an aluminium one, and then coating the 

 end of the platinum wire at A with a deposit of aluminium ; but, 

 although this is doubtless possible, success in the attempt has not yet 

 been attained. 



Air-tightness is another essential condition of the constancy of 

 these photo-electric cells, for it is found that in cells which are not 

 quite air-tight the resistance of the cenanthol increases very much 

 after a few weeks, probably owing to the oxidation of the liquid by 

 the air ; and this great increase of resistance promotes sluggishness 

 in the response of the cell to the action of light. 



An examination of a seleno-aluminium cell with the various portions 

 of the spectrum of lime-light shows that the cell is sensitive to all the 

 rays from the end of the red, and below it, to beyond the violet, the 

 maximum E.M.F. being produced in the yellow : but the magnitude of 

 the E.M.F. does not vary very greatly until the violet is reached. In 

 this respect the seleno-aluminium cell differs from all other photo- 

 electric cells, for the sensitiveness of most of the latter is almost 

 wholly confined to the blue. It may be mentioned, however, that 

 a cell formed by immersing clean silver plates in a solution of cosine 

 gives electromotive forces of opposite signs for the red and the blue 

 rays. 



The energy incident on a photo-electric cell has been found to be 



