278 PERMEABILITY OF HIGHLY- HEATED IRON BY GASES. 



hydrogen and heated in a furnace, permits this gas to escape so thoroughly as 

 to produce an almost perfect vacuum in the interior of the tube. These curious 

 experiments will serve to explnin many phenomena Vhich present themselves ia 

 meiallurgic operations, and which have never yet, I think, received satisfactory 

 explanation. I have the honour to submit to the Academy the result of some 

 researches I have made in this subject, and which it is my design to carry on 

 and complete. 



I caused some lengths of gun-barrels to be rolled flat, and then soldered the 

 two ends, so that I thus obtained long rectangles formed of two plates in con- 

 tact, soldered at the edges. On heating a lamina thus prepared to the high 

 temperature of a smelting furnace, it was soon observed that the portions not 

 soldered began to separate, and regained their cylindrical shape and original 

 Volume. This could doubtless only have been caused by the gases of the fur- 

 nace penetrating the mass of iron, and producing the distension of the portions 

 at first in contact. To this penetration of the gases, we may attribute the blis- 

 ters which frequently cover large pieces of cast-metal, especially those used 

 for blinding, at the instant when they are extracted from the welding furnace. 

 If one of these blisters is pierced on withdrawing the piece rough from the fur- 

 nace, a jet of combustible gas is seen to escape, having been doubtless accumu- 

 lated during the heating, in the cavities that occur in a piece which has been 

 incompletely wrought. 



It has b^en long observed that iron heated with coal-dust in the cementing- 

 boxes, was covered, after its change to steel, with a quantity of bubbles, more 

 or less numerous according to the nature of the metal employed, and it is easy 

 to convince ourselves,, by examination, that each of these bubbles corresponds 

 to a point where the junction of the metallic sponge has been imperfect, whether 

 owing to the presenu-e of some infusible matter, as lime or the ash of the com- 

 bustible used, or to the imperfection of the mechanical working. "We may 

 therefore suppose, after the experiments of Messrs. H. Deville and Troost, that 

 the gases contained in the cementing-boxes traverse the pores of the iron, and 

 accumulating: in hollows of the red-hot metal, form the bubbles of which we 

 are speaking. A rather simple experiment confirms this hypothesis. In fusing 

 together the iron-plates which commonly occur in commerce, and are not of 

 uniform texture, we always obtain the poule steel (as the steel covered with 

 blisters is called) ; whereas if we work with the perfectly homogeneous iron, 

 which ia obtained by exposing cast-steel for many hours to a high temperature, 

 it is then seen that the plates of this homogeneous iron return to the condition 

 of steel, but without a single blister on their surface. 



We may conclude from these experiments, that in order to procure steel with 

 a smooth surface, we should employ iron as homogeneous as possible, and hare 

 recourse to a rapid process of cementation. Also, to avoid in castings the pro- 

 duction of blisters, it is necessary to prevent the formation of hollows in the 

 rough material, for, as we have tried to demonstrate, these blisters are caused 

 by the gases of the furnace condensing in the cavities of the metal. 



