254 Mr. Shelford Bidwell on the 



The succeeding experiments indicate that at the anomalous 

 junction the electrical contact between the Pt and Se was 

 imperfect, and that the Se contained a certain amount of 

 platinous selenide. 



Exp. 35. — One part of platinous selenide having been 

 incorporated with six parts of Se, the mixture was fused and 

 crystallized. Thus was obtained a button of crystalline Se 

 containing about 15 per cent, of platinous selenide. The 

 button and a piece of Pt foil were partly immersed in water 

 and their dry ends connected to the galvanometer. A voltaic 

 current was indicated from the button to Pt through the 

 water. 



Exp. 36. — The button was dried and tested for a thermo- 

 current with Pt. Result, current from Pt to button through 

 the junction. The button, therefore, behaved thermo-electri- 

 cally like ordinary Se. 



Exp, 37. — By means of an adjustable spring a strip of Pt 

 foil was arranged so as to press lightly on the button. The 

 two were connected with the galvanometer. 



Touching the button near the junction with a hot wire 

 caused a current from Pt to button. 



Holding the tip of the finger very near the junction without 

 actual contact caused a current from button to Pt. 



Blowing air upon the junction from a hot glass tube pro- 

 duced a current from Pt to button when the air-blast was 

 derived from a foot-blower, and from button to Pt when air 

 from the lungs was blown through the hot tube. 



Thus all the thermo-effects of the " anomalous " junction 

 were completely imitated. Perhaps in the case of that junc- 

 tion good contact was prevented by a film of non-conducting 

 red Se, or possibly the wire electrode may have been simply 

 loose. 



These effects could not be produced when using a button 

 which consisted of Se annealed without contact with a metal 

 and containing no added selenide, the voltaic effect being 

 then similar in direction to the heat-effect. 



Summary. 

 1. The conductivity of crystalline Se appears to depend 

 principally upon the impurities which it contains in the form 

 of metallic selenides. It is suggested that the selenides 

 conduct electrolytically, and that the influence of light in 

 increasing the conductivity is to be attributed to its property 

 of facilitating the combination of selenium with metals in 

 contact with it. 



