210 PROFESSOR THOMAS GRAHAM'S SCIENTIFIC WORK. 



the process to tbat which he had employed in his India-rubber experi- 

 ments. The results he obtained Avere communicated to the Eoyal 

 Society, partly in the paper already referred to " On the absorption and 

 separation of gases by colloid septa," and partly m four supplementary 

 notices published in the proceedings of the society.* In carrying out 

 the investigation forming the subject of these several communications, 

 ]Mr. Gnvhaiu had the advantage of being admirably seconded by his 

 assistant, Mr. W. Chandler IJoberts, whose able and zealous co-opera- 

 tion he repeatedly acknowledged in the warmest terms. 



In the course of experiments made on the transmission of gases 

 through ignited metallic septa, a particular platinum tube, being ren- 

 dered vacuous, was found at all temperatures below redness to be quite 

 impermeable to hydrogen ; whereas, at a red heat, it transmitted 100 

 cubic centimeters of hydrogen in half an hour, the quantities of oxygen, 

 nitrogen, marsh gas, and carbonic gas, transmitted under the same con- 

 ditions, not amounting to .01 cubic centimeter each in half an hour. 

 It was ascertained farther that, with an ignited vacuous tube of 

 platinum surrounded by a current of ordinary coal-gas, (a variable 

 mixture of gases containing on the average about 45 i^er cent, of 

 marsh gas, 40 per cent, of hydrogen, and 15 per cent, of other gases 

 aud vapors,) a transmission of pure hydrogen alone took i^lace thi'ongh 

 the heated metal. This property of selective transmission, manifested 

 by platinum, was so far analogous to the property of selective trans- 

 mission manifested by India-rubber, that whereas a septum of Jiulia 

 rubber transmitted the nitrogen of the air in a much smaller ratio 

 than the oxygen, the septum of ignited platinum transmitted the 

 other constituents of coal-gas in an infinitely smaller ratio than the 

 hydrogen. Hence the knowledge of the absorption by India rubber of 

 the gases which it most freely transmitted, suggested to Mr. Graham an 

 inquiry as to the possible absorption of hydrogen gas by platinum. 

 Accordingly platinum, in different forms, was heated to redness, and 

 then allowed to cool slowly in a continuous current of hydrogen. 

 The metal so treated, and after its free exposure to the air, was i)laced 

 in a porcelain tube, which was next made vacuous by the Sprengel 

 pump. During the production and maintenance of the vacuum, no 

 hydrogen was extracted from the metal at ordinary temperatures ; 

 or even during an hour's exposure to the temperature of 220° ; or yet 

 at a heat falling just short of redness. But at a dull red-heat and 

 upward, a quantity of hydrogen gas was given oft' amounting in 

 volume, measured cold, to as much, in some cases, as 5.5 times the 

 volume of the platinum. Thus was opened out to Mr. Graham the 

 subject of his last, and probably greatest discovery, the occlusion of 

 gases by metals. Very many metals were examined in their relations 

 to different gases, but the most interesting results were those obtained 

 with platinum as above described ; and those obtained with silver, with 

 iron, and, above all, with palladium. 



* Royal Society Proceecliugs, xv, p. 502; xvi, p. 4"22; xvii, p. 212, p. 500. 



