Dialytic Separation of Gases by Colloid Septa. 517 



posed to hydrogen between 90° and 97° C. for three hours, and 

 then allowed to cool in the gas for one hour and a half. Now 

 placed in a glass tube, exhausted, and heated by a gas-name, the 

 palladium gave off gas in a continuous stream for twelve minutes, 

 when it ceased. The gas amounted to 85*56 cub. centims., of 

 which 96*8 per cent, was hydrogen; therm. 17°*5, barom. 764 

 tnillims. 



1 vol. palladium held 643*3 vols, hydrogen. 



By the care of my zealous assistant, Mr. W. C. Roberts, the 

 hydrogen employed in these experiments was purified to the 

 highest degree by passing it in succession through alcohol, water, 

 caustic potash, and tubes of 0*7 metre each, filled with broken 

 glass impregnated with nitrate of lead, sulphate of silver, and oil 

 of vitriol. The gas was inodorous, and burned with a barely 

 visible flame. 



No alteration was sensible in the metallic appearance of the 

 palladium-foil when charged with hydrogen, or when discharged. 

 The foil was much crumpled and rather friable after repeated use ; 

 but this may have arisen from frequent handling. 



3. Palladium appears to absorb hydrogen largely, even at natural 

 temperatures, provided that the metal has been recently ignited 

 in vacuo. The foil, without such preparation, was placed in a 

 bottle of pure hydrogen for several hours, but yielded nothing 

 when afterwards ignited in the Sprengel vacuum. The foil, 

 however, being immediately returned after cooling to a stoppered 

 bottle containing hydrogen, and left in the gas for a night, ab- 

 sorption now took place — air rushing in, on opening the stopper, 

 as into a partial vacuum; therm. 19°. When the palladium-foil 

 was afterwards transferred to a glass tube and connected with 

 the Sprengel pump, it was found difficult to obtain a vacuum 

 for some time, owing to hydrogen coming off at the temperature 

 of the atmosphere. But after a fair vacuum was produced 6*96 

 cub. centims. were collected, of which 6*78 proved to be hydro- 

 gen. Heat was then applied, and 42 cub. centims. came over in 

 five minutes, making altogether more than 50 cub. centims., or 

 376 volumes of gas. The absorption of hydrogen appears, then, 

 to be suspended at a low temperature, unless the condition of 

 the metal be favourable. The action of a plate of clean platinum 

 in determining the combustion of explosive gas is equally critical 

 at a low temperature. 



4. ^different specimen of palladium-foil, weighing 5*76 grms., 

 and having a volume of 0*485 cub. centini., was charged with 

 hvdrogen, and discharged, more than once. In the second ex- 

 periment, the foil was heated in hydrogen at 100° for three hours. 

 Distilled afterwards in a porcelain tube at a low red heat in the 



