252 Mr. Shelford Bidwell on the 



or so a small current was indicated which, after slowly in- 

 creasing for about 15 minutes, became constant. While 

 magnesium was being burnt near the receiver, the strength of 

 the current was greatly intensified ; even diffused daylight 

 temporarily increased the deflexion to the extent of 150 scale- 

 divisions. The direction of the current was as before from 

 SetoPt. 



The plate was then enclosed in a receiver over sulphuric 

 acid. In 8 minutes the permanent current had completely 

 disappeared, and exposure to diffused daylight caused a 

 deflexion of only 2 divisions. In 20 minutes burning mag- 

 nesium at a distance of 6 inches failed to produce any effect 

 whatever. The Se had completely lost its power of generating 

 a current under the influence of light. 



The natural inference seems to be that Se (or Se containing 

 selenide) is slowly oxidized by water, and that the oxidation 

 is increased by the action of light*. Some points bearing on 

 the question will be further considered in the next section. 



§15. The Thermoelectric Quality of Annealed Selenium. 



Adams and Day believed they had obtained evidence that 

 the thermo-electric quality of crystalline Se was altered by 

 prolonged annealing. To use the old-fashioned terminology, 

 they thought that by the process of annealing Se was raised 

 from its place at the bottom of the thermo-electric series, which 

 the experiments of Matthiessen had assigned to it, to a position 

 somewhere above that of platinum. In other words, if a 

 Se-Pt junction were heated to a temperature slightly above 

 that of the other junctions in the circuit (the latter being at 

 the ordinary temperature of the air), then, if the Se were 

 crystallized but not thoroughly annealed, there would be a 

 current from Pt to Se through the heated junction, but if the 

 Se were well annealed the thermo-current would be in the 

 reverse direction — from Se to Pt. 



They applied heat at a junction by directing upon it a blast 

 of hot air from a glass tube terminating in a fine nozzle, the 

 tube being heated over one or more Bunsen burners. Where 

 it was possible, heat was also applied by holding the junction 



* Later experiments render this doubtful. With Se which does not 

 contain any considerable admixture of Pt 2 Se (and perhaps it should he 

 added, which has not been previously used as an anode) the photo- 

 electromotive force appears to be always in the opposite direction — from 

 Pt to Se. And this seems to be the case not only with water as the electro- 

 lyte, but also with the other badly conducting liquids, such as cenanthol, 

 used by Prof. Minchin.— June 28th. 



