530 On the Dialytic Separation of Gases by Colloid Septa. 



hot and suddenly cooled, and was not altered in aspect or in 

 solubility in acids. The gas extricated by the air-exhauster 

 amounted to 



9*45 cub. centims. in 13 minutes. 



2-43 „ 5 „ 



8-05 „ 42 „ 



3-15 „ 60 .. 



23*08 „ in two hours. 



Of this gas 20*76 cub. centims. proved to be carbonic oxide. 

 Pure iron, then, is capable of taking up at a low red heat, and hold- 

 ing when cold, 4*15 volumes of carbonic oxide gas. This fact was 

 confirmed in various other experiments. It explains partly, if 

 not entirely, the abundance of carbonic oxide observed among 

 the natural gases of iron in experiments 1, 2, and 3. In the 

 course of its preparation wrought iron may be supposed to oc- 

 clude six or eight times its volume of carbonic oxide gas, which 

 is carried about ever after. How the qualities of iron are affected 

 by the presence of such a substance, no way metallic in its cha- 

 racters, locked up in so strange a way, but capable of reappearing, 

 under the influence of heat, at any time, with the elastic tension 

 of a gas, is a subject which metallurgists may find worthy of 

 investigation. , 



The relations of the metal iron to carbonic oxide gas appear to 

 be altogether peculiar. They cannot fail to have a bearing upon 

 the important process of acieration. The intervention of carbonic 

 oxide in the usual process of the cementation of iron with char- 

 coal, long recognized by accurate observers, may be said now to 

 be placed beyond all doubt by the recent beautiful research of 

 M. Margueritte*. Hitherto the decomposing action of the iron 

 upon carbonic oxide has been supposed to be exercised only at 

 the external surface of the metal. A surface-particle of the iron 

 has been supposed to assume one-half of the carbon belonging 

 to an equivalent of carbonic oxide (C 2 O 2 ), while the remaining 

 elements diffused away into the air as carbonic acid (CO 2 ), to 

 reacquire carbon from the charcoal placed near, and to become 

 capable of repeating the original action. It is now seen that 

 such a process need not be confined to the surface of the iron 

 bar, but may occur throughout the substance of the metal, in 

 consequence of the prior penetration of the metal by carbonic 

 oxide. The direct contact and action of carbon (in the form 

 either of diamond or charcoal) upon iron is allowed to produce 

 cast iron and not steel. It appears that the diffused action of 

 carbonic oxide is the proper means of distributing the carbon 

 throughout the mass of iron. The blistering of the bar appears 

 to testify to the necessary production and evolution of carbonic 

 * Annates de Chimie, &c. 4 ser. vol. vi. 1865. 





