OF ARTS AND SCIENCES : MAY 24, 1870. 235 



hydrogen or carbonic oxide is passed through a red-hot iron tube, a no 

 inconsiderable portion of the gas escapes through the walls. The same 

 is true to a still greater degree when hydrogen is passed through a red- 

 hot tube of platinum, and Graham showed that through the walls of a 

 tube of palladium hydrogen gas passes, under the same conditions, al- 

 most as rapidly as water through a sieve. Moreover, our distinguished 

 Associate proved that this rapid transfer of gas through these dense 

 metallic septa was due, as in the case of the india-rubber, to an actual 

 chemical combination of its material with the metal, formed at the sur- 

 face, where the gas is in excess, and as rapidly decomposed on the op- 

 posite face of the septum. He not only recognized as belonging to this 

 class of phenomena the very great absorption of hydrogen by platinum 

 plate and sponge in the familiar experiment of the Doebereiner lamp, 

 but also showed that this gas is a definite constituent of meteoric iron, — 

 a fact of great interest from its bearing on the meteoric theory. 



We are thus led to Graham's last important discovery, which was the 

 justification of the theory we have been considering, and the crown- 

 ing of this long line of investigation. As may be anticipated from what 

 has been said, the most marked example of that order of chemical com- 

 pounds, to which the metallic transpiration of aeriform matter we have 

 been considering is due, is the compound of palladium with hydrogen. 

 Graham showed that when a plate of this metal is made the negative 

 pole in the electrolysis of water, it absorbs nearly one thousand times 

 its volume of hydrogen gas, — a quantity approximative^ equivalent to 

 one atom of hydrogen to each atom of palladium. He further showed 

 that the metal thus becomes so profoundly altered as to indicate that the 

 product of this union is a definite compound. Not only is the volume 

 of the metal increased, but its tenacity and conducting power for elec- 

 tricity are diminished, and it acquires a slight susceptibility to magnet- 

 ism, which the pure metal does not possess. The chemical qualities 

 of this product are also remarkable. It precipitates mercury from a 

 solution of its chloride, and in general acts as a strong reducing agent. 

 Exposed to the action of chlorine, bromine, or iodine, the hydrogen leaves 

 the palladium and enters into direct union with these elements. More- 

 over, although the compound is readily decomposed by heat, the gas can- 

 not be expelled from the metal by simple mechanical means. 



These facts recall the similar relations frequently observed between 

 the qualities of an alloy and those of the constituent metals, and suggest 

 the inference made by Graham, that palladium charged with hydrogen 



