CHEMICAL SCIENCE. 221 



its vertical branch being 80 centimeters long. For the space of eight 

 or ten hours, a current of hydrogen was driven through the apparatus, 

 which was maintained at a high temperature, so as to exhaust the ac- 

 tion of the hydrogen on the sides of the iron tube, and to drive away 

 all the atmospheric air, as well as the moisture contained in the tube, 

 or likely to be produced there. This done, the communication 

 between the iron tube and the hydrogen apparatus was cut off by 

 melting down the glass tube by the, aid of the blowpipe. No sooner 

 was this rlli-eted, (han the mercury, no longer kept down by the stream 

 of hydrogen, yielded to the pressure of the air, and rose in the verti- 

 cal glass tube to the height of 740 millimeters, or very nearly the 

 usual barometrical hight. This would not have happened, had there 

 not been a nearly complete vacuum in the tube the instant the supply 

 of hydrogen was cut off. But what had become of the hydrogen 

 supplied before ? There is but one explanation possible, viz : that, 

 notwithstanding the pressure of the atmosphere, the hydrogen had 

 passed through the pores of the steel tube. 



Additional Facts. M. Cailletet has detailed to the French Acad- 

 emy some very curious additional facts bearing on the experiments 

 of M. Deville. He says : " I passed portions of a gun-barrel through 

 rollers till they were flattened. The ends were then welded. Thus, 

 rectangular pieces of iron were obtained, formed of two plates in con- 

 tact, and sealed at their extremities. On strongly heating one of 

 these pieces in a furnace, the non-welded portions separated, and 

 resumed the cylindrical form and their original volume. It could not, 

 therefore, be doubted that the gases of the furnace had penetrated the 

 mass of iron and distended its walls." To a similar action, M. Caille- 

 tet ascribes the cavities in large masses of forged iron ; and he states, 

 that in the process of cementation, acier poule or steel with vesicles, 

 is constantly produced ; but if soft, perfectly homogeneous iron, such 

 as can be obtained by keeping melted steel for several hours at a 

 high temperature, be employed, it is reconstructed into steel without 

 blisters. M. Deville remarks upon this communication, that it is 

 "very interesting and conclusive, ami he adverts to the discharge of 

 gas from molten matter, often observed in metallurgical operations. 

 These gases, he considers, penetrate the walls of the crucibles by 

 endosmose, and give rise to bubbles in the metals. 1 ' 



Action of Porcelain and Lac as at hiyk temperatures on Gases. Pos- 

 sible action of the Moon. M. Deville, in commenting further upon the 

 above facts, stated that his brother and M. Troost have shown that if 

 hydrogen traverses without diiHculty the walls of a porcelain tube at 

 a high temperature, it docs not do so when the tube begins to soften or 

 vitriiy . The gas is then absorbed by the vitrified surface, from whence 

 it escapes, leaving it porous. He connects these facts with the ap- 

 pearance of certain lavas. He says the lavas of Vesuvius, whatever 

 the rate of their cooling, arc always crystalline, and that they disengage 

 aqueous vapor, chlorides, sulphides, etc., as the crytallization proceeds, 

 just as oxygen escapes from silver that takes the rocky form, or air 

 escapes from freezing water. The ciystallizatioiv-of lavas he states to 

 be accompanied by increase of density and evolution of heat. He 

 traces a resemblance between the Campi Phlegnci and the surface of 

 the moon, and considers that the latter may have behaved like erup- 

 19* 



