Dialytic Separation of Gases by Colloid Septa. 523 



for hydrogen — a result which is consistent with the crystalline 

 character of the substance. 



Copper. 



The power to occlude gases appears not be confined to palla- 

 dium and platinum among the metals. The exact experiments 

 of M. Dumas, by which the atomic weights of the leading ele- 

 ments were definitely settled, afford an indication of the absorp- 

 tion of hydrogen gas by spongy metallic copper reduced from 

 the oxide, sufficient to affect the weight of the metal to the extent 

 of about 3 parts in 100,000*. 



1. With the view of applying the method of extracting gas 

 followed in the treatment of the preceding metals, so much oxide 

 of copper was reduced by hydrogen as was calculated to yield 

 50 grms. of metallic copper. The reduced metal was again 

 heated to redness and slowly cooled in a stream of dry hydrogen. 

 After free exposure to the air for a few minutes, the metal was 

 now submitted, at a red heat, to the action of the Sprengel pump. 

 It then gave off in one hour 3*35 cubic centims. of gas, mea- 

 sured cold, which appeared to be pure hydrogen (the explosion 

 with oxygen indicated 3*4 hydrogen). Taking the specific gra- 

 vity of copper at 8*85, 50 grms. of that metal would be 5*65 

 cub. centims. in volume, and the result is that 



1 vol. reduced copper sponge occludes 0*6 vol. hydrogen. 



Hydrogen being about 12,000 times as light as copper (at 15°), 

 1 part of gas by weight has been taken up by 20,000 parts of 

 metal. 



2. The same weight and volume of fine copper, in the form of 

 wire thoroughly cleaned, was exposed to hydrogen at a red heat, 

 and then submitted to exhaustion for one hour. It gave 2*6 

 cub. centims. gas, of which 2 cub. centims. were hydrogen, and 

 the remaining 06 principally carbonic oxide. It may be repre- 

 sented that 



1 vol. wrought copper occludes 0*306 vol. hydrogen. 



Where a metal, such as wrought copper, may contain small 

 quantities of carbon and oxygen, an obvious cause will exist for 

 the production and evolution of carbonic oxide under the influ- 

 ence of heat. Gas so generated appears to be added to the oc- 

 cluded hydrogen when extricated in the last experiment. 



* Annates de Chimie et de Physique, 3 ser. vol. viii. p. 205. The ob- 

 servations of M. Melsens show that 240 grms. of copper may fix about 

 0*007 grm. of hydrogen, most being fixed when the oxide of copper is re- 

 duce 1 by hydrogen at a low temperature. In the subsequent oxidation of 

 the copper the gas does not come out suddenly, but in a gradual manner. 



