M. Deville on the Disassociation of Compound Bodies. 517 



rounding water. This was confirmed by an experiment in which 

 high-pressure steam was passed through a platinum tube filled 

 with pieces of lime, and heated nearly to the melting-point of 

 platinum. In this case only a small quantity of detonating gas 

 was formed. 



Deville endeavoured to ascertain the temperature at which 

 this occurred. Keguault had observed that silver at its melting- 

 point decomposes water, which it again liberates in exhibiting 

 the phsenomenon of " spitting." Deville endeavoured to explain 

 this by assuming that the water, in passing through the heated 

 porcelain tube used in the experiment, decomposed into its ele- 

 ments, and on this depended the absorption of free oxygen, but 

 that silver had not a sufficient affinity for oxygen to decompose 

 water. He sought to confirm this explanation by using for the 

 experiment an oxide readily reduced by hydrogen at a low tem- 

 pei'ature, and which is consequently itself incapable of decompos- 

 ing water. A long and broad platinum tray, filled with pure fused 

 litharge, was placed in a porcelain tube which it filled to two- 

 thirds. The porcelain tube was heated to bright redness, and a 

 rapid current of steam passed through. Much oxide of lead 

 volatilized, and at the conclusion of the experiment there were 

 evidences that the oxide of lead in solidifying had disengaged 

 oxygen. On examining the tube, it was found that those parts 

 exposed to the strongest heat showed symptoms of glazing, from 

 the volatilized oxide of lead; further on there was pulverulent 

 oxide of lead, and finally a mirror of reduced lead. In passing 

 through the tube the water had been decomposed into its ele- 

 ments, the oxygen had been absorbed by the litharge, and the 

 hydrogen had reduced some oxide of lead, but only where the 

 decomposed water could be again formed. Deville estimated 

 this temperature at very near the melting-point of silver. 



Although water cannot exist at the melting-point of silver, 

 still by the combination of hydrogen and oxygen so much heat 

 is liberated that iridium can be melted. How is it that the 

 oxyhydrogen flame melts platinum, and that molten platinum 

 again decomposes water ? If we could compare the quantity of 

 platinum melted by a given weight of detonating gas with the 

 quantity of water decomposed into hydrogen and oxygen by the 

 platinum thus melted, a conclusion might be drawn in reference 

 to the condition of these gases at the moment of that disengage- 

 ment of heat at which water is no longer formed. 



Deville describes an experiment which seems to prove that 

 hydrate of soda, and especially hydrate of potash, easily decom- 

 pose into their elements at a temperature near that of the melt- 

 nig- point of wrought iron. The experiment is essentially that 

 of Gay-Lussac and Thenard, of the production of sodium. He 



