M. Deville on the Dissociation of Water, 537 



in a shorter impermeable porcelain tube, an annular space is 

 obtained between the tubes which may be filled with any gas. 

 For this purpose a glass tube is fitted in the cork at each end, 

 by one of which the gas is admitted, and by the other of which 

 it emerges. The porous tube is likewise fitted with tubes, by 

 which a different gas may enter the tube at one end and emerge 

 at the other. The apparatus being thus arranged, if a current 

 of carbonic acid is introduced into the annular space between 

 the two tubes, and a regulated current of hydrogen into the 

 interior of the porous tube, hydrogen gas may be lighted at the 

 tube fitting in the annular space where the carbonic acid would 

 be expected to emerge, while from the porous tube almost pure 

 carbonic acid emerges. Thus in virtue of endosmose these 

 gases have changed place ; the experiment is well suited for lec- 

 ture purposes. 



If the apparatus thus arranged be heated in a furnace from 

 1100° to 1300°C, it may be used to demonstrate the pheno- 

 menon of the spontaneous decomposition of water which Deville 

 has called dissociation*. In this case the annular space is 

 filled with coarse fragments of biscuit-porcelain. Instead of hy- 

 drogen, aqueous vapour is made to arrive in the interior of the 

 porous tube, while carbonic acid is passed into the annular space. 

 The gases are collected in long glass tubes placed in a bath 

 containing caustic potash, to stop all carbonic acid. When the 

 furnace is in full activity, a highly explosive gaseous mixture is 

 collected consisting of hydrogen and oxygen. Here the aqueous 

 vapour is spontaneously decomposed inside the porous tube; the 

 hydrogen traverses the porous diaphragm, which thus, like an 

 ordinary filter, separates it from the oxygen. A considerable 

 quantity of carbonic acid diffuses and mixes with the hydrogen. 



The detonating gas is not quite pure : the action of hydrogen 

 on carbonic acid produces a certain quantity of carbonic oxide. 

 Notwithstanding all possible precautions, too, a certain quantity 

 of hydrogen always escapes, and the detonating gas always con- 

 tains an excess of oxygen. It further contains some nitrogen, 

 which is introduced along with carbonic acid. 



The carbonic acid determines in the operation the separation 

 of the gases by endosmose, but it may also act mechanically. 

 Water alone, heated in a porcelain tube even to almost the fusing- 

 point of platinum, emerges entirely unchanged, and is not de- 

 composed to any appreciable extent. 



To explain these facts, Deville enters into the following con- 

 siderations. The temperature of the combustion of hydrogen 

 and oxygen does certainly not exceed 2500° C. At this point 

 the gases occupy a volume ten times that which they occupy at 

 * Phil. Mag. vol. xx. p. 448. 



