In the course of experiments on the electrical conductivity of selenium, a cylindrical bar of this substance in the vitreous state was enclosed in a glass tube which was attached to the exhaust-tube of a Sprengel pump. The bar of selenium was 45 millims. long and 3.5 millims. in diameter. Platinum wires were attached to the ends of it and passed through the sides of the glass tube. The tube was exhausted, and allowed to remain attached to the pump for four days. I t was now found that the selenium had acquired a conductivity greatly exceeding that of the pure element in its most highly conducting condition. The experiment was repeated with the intention of observing the time required to produce conductivity. In forty-two hours the needle of a highly sensitive galvanometer was slightly deflected when the selenium was placed in the circuit of ten Leclanché cells. The conductivity of the selenium increased rapidly for four days, when the experiment was unavoidably interrupted. On admitting air to the tube no change of conductivity was observed. The selenium was unaltered in appearance, even when examined microscopically. On breaking the bar it was found that the conducting-film was entirely superficial; it was not removed by rubbing forcibly with a cloth. Dilute nitric acid also failed to remove it. Bibulous paper moistened with solution of silver ammonio-nitrate was not stained by it (Merget, ‘Comptes Kendus,' vol. lxxiii. p. 1356). It therefore appears highly probable that the film does not consist of uncombined mercury. As it has not hitherto been known that mercury combines with selenium at ordinary temperatures, a bar of selenium was immersed in mercury and allowed to remain undisturbed for six months. At the end of this time it was found that the selenium was coated with a highly conducting film. I could not dectect any difference between this film and those produced in the Sprengel vacuum. An attempt was now made to estimate the quantity of mercury required to produce the observed conductivity. A bar of selenium 125 millims. long and 2 millims. in diameter, having platinum wires fused into each end, was enclosed in a glass tube, containing also a minute globule of mercury about 0.5 millim. in diameter. The tube was exhausted by means of the Sprengel pump, and then hermetically sealed and detached from the pump. In 92 hours the bar began to conduct, and the conductivity increased rapidly from day to day for four days. On the fifth day, no increase being observed, it was supposed that air had leaked into the tube; and on examining it a flaw, which would account for the leakage, was detected. The tube was therefore again attached to the pump, exhausted, and again sealed, the defective portion being removed. The conductivity of the bar again increased from day to day, and is still steadily but slowly increasing (eleven days after the second sealing of the tube). Although the bar of selenium now possesses a comparatively low resistance, I cannot detect the slightest alteration in the size of the minute globule of mercury which has supplied the material for the conducting-film, extending over a surface one thousand times greater than that of the globule. The granular modification produced by subjecting vitreous selenium to a temperature of 100°C. for three hours also acquires a great increase of conductivity when exposed to the vapour of mercury in the Sprengel vacuum.