DIALYTIC SEPARATION OF GASES BT COLLOID SEPTA. 425 



The same platinum was a third time charged with hydrogen ; but on this occasion the 

 platinum was placed in a tube of hard glass, and the tube connected with the air- 

 exhauster. The glass tube was heated by an oil-bath, and the platinum kept in vacuo 

 at a temperature of 220° C. for an hour. Not a bubble of gas was evolved. The glass 

 tube was afterwards heated by a small Bunsen burner, which was calculated to give a 

 degree of heat little short of visible redness, still no hydrogen came off. The tube 

 was now heated sufficiently to soften glass (500°). Gas began to come off, of which 1"8 

 cub. centim., containing 1"72 hydrogen, were collected in ten minutes. The glass tube 

 having cracked, the whole apparatus was allowed to cool, and the platinum transferred 

 to a porcelain tube. Further heated by a combustion-furnace for one hour, the plati- 

 num gave off 8-6 cub. centims. gas, of which 8-2 were hydrogen. The platinum there- 

 fore appears to have occluded altogether 3-79 vols, of hydrogen. 



The preceding experiment appeared to show a complete sealing up of the occluded 

 hydrogen at low temperatures, seeing that, although nearly four volumes of gas were 

 present, none escaped below a red heat. But to test the effect of time at the temperature 

 of the atmosphere, the platinum, again charged with hydrogen, was sealed up hermeti- 

 cally in a glass tube, which it nearly filled, and not opened for two months. The air in 

 the tube was was then transferred and examined. It did not exhibit any reduction of 

 volume under the electric spark or a pellet of spongy platinum. The air therefore 

 appeared to contain no hydrogen ; the latter had not diffused out, but, it is to be pre- 

 sumed, was retained by the platinum without loss. These experiments, although related 

 last, were the first performed in this inquiry. The included hydrogen was never entirely 

 extracted in an hour, and is probably understated. The gas always came off gradually, 

 more than one half of the whole in the first twenty or thirty minutes. The last results 

 may be stated as follows : — 



1 vol. hammered platinum occluded 2-28 vols, hydrogen. 



« 75 55 2"00 „ 



95 55 5« O' I J „ 



The high absorbing power of the hammered platinum, or rather the low absorbing 

 power of the fused metal, was ascribed to a mechanical difference between the two — to a 

 more open texture in the former, permitting more free access of hydrogen, liquefied as 

 it may be, to the interior of the metal. 



8. The extrication of occluded hydrogen from platinum had always required a degree 

 of temperature verging upon a red heat, even when aided by a vacuum ; and this remains 

 true of hydrogen originally absorbed at or near a red heat. But the fact appears to be 

 compatible with the absorption of the gas, under the pressure of the atmosphere, at a 

 considerably lower temperature. Thin platinum-foil was first deprived of a little natural 

 gas by ignition in vacuo in the porcelain tube. The foil was afterwards placed in a glass 

 tube and heated again in a stream of hydrogen, to a temperature not exceeding 230° C, 

 for three hours, by means of an oil-bath, and further allowed to cool slowly in an atmo- 



3n2 



