518 M. Deville on the Disassociation of Compound Bodies. 



took a mercury-bottle provided at the mouth with an iron tube, 

 and at the lower part, near the bottom of the bottle, with a 

 second tube at right angles to the axis of the bottle. This bottle 

 was heated in a blast-furnace in which a white heat could be 

 produced, the roof of the furnace being perforated to allow the 

 vertical tube to pass through. The bottle was filled either with 

 pure iron filings reduced by hydrogen, or with pure iron sponge. 

 When the temperature in the furnace had reached white heat, 

 hydrate of potash was passed into the bottle by means of the 

 vertical tube, and the horizontal tube was provided with a re- 

 ceiver. As soon as potassium had been obtained, the operation 

 was rapidly stopped, when he observed — 



1. That as soon as the supply of potash ceased to be very 

 rapid, much potash was volatilized. Potassium is only obtained 

 when large quantities of potash are added through the upper 

 tube. 



2. On sawing across the bottle, the iron in the part of the 

 bottle which had been hottest was found to be as bright as before 

 the experiment, and without a trace of oxide. In the lower part, 

 where there had been less heat, there was a magma of oxide of 

 iron and potash. 



3. If the whole bottle is kept at the temperature at which the 

 magma in the previous experiment is formed, no trace of potas- 

 sium, but only hydrogen and potash are formed. 



It may therefore be concluded that, in passing over the white- 

 hot iron, the hydrate of potash decomposes into its elements, and 

 potassium, hydrogen, and oxygen in their free state reach the 

 lower part of the apparatus. Only a portion of free oxygen is 

 absorbed by the iron, and hence potassium is formed, which is 

 carried away in the rapid current of hydrogen ; and the potash 

 mechanically prevents the oxide of iron from giving up its oxygen 

 to the potassium, or to the excess of hydi'ogen. 



The best proof that the iron on the hottest parts of the appa- 

 ratus only serves for the transmission of heat is the fact that, if 

 the whole bottle be kept at the somewhat lower temperature 

 which the lower part of the bottle has when potassium is pro- 

 duced, only hydrogen and potash vapour are formed. 



Deville finally mentions the great influence which velocity has 

 upon those chemical actions which depend on the decomposition 

 of compounds into their elements in the preparation of change- 

 able substances. Oxide of zinc volatilizes in a slow current of 

 hydrogen, but is reduced in a lapid stream. In the same man- 

 ner, the use of chalk in Deville's preparation of sodium consists 

 in the effect of the large quantity of carbonic oxide formed by the 

 action of the charcoal on the chalk. Experiment^; in this direc- 

 tion will furnish the elements of a chapter of cliiniical dynamics, in 



