44 M'Thenardon the [Jan. 



Let this experiment be repeated, after having so diluted tlie 

 peroxide, that it shall contain only seven or eight times its 

 volume of oxygen, the disengagement of the gas will not be per- 

 ceptible, even at 120°, but it becomes so soon afterwards, and 

 goes on increasing until it ceases. From this period, the liquor 

 contains no gas, and consequently will not effervesce with oxide 

 of manganese. 



All other circumstances being equal, peroxide of hydrogen 

 suffers no more alteration by exposure to light than in darkness. 

 In both cases, small bubbles are disengaged from time to time, 

 and it finishes at the exipiration of some months, even at common 

 temperatures, with being for the most part deoxidized. This 

 deoxidizement, which probably depends upon many causes, 

 appears to me to be principally produced by some particles of 

 matter which the peroxide retains. To preserve it as much as 

 possible, it must be surrounded with ice. 



When the peroxide is subjected to the action of the voltaic 

 pile in the same way that water usually is, similar results are in 

 both cases produced, excepting that with the peroxide, the dis- 

 engagement of oxygen gas is much greater. I ought to observe, 

 however, that I have not collected the gases to examine them. 



Of the Action of the Metals at Common Temperatures. 



In general the metals tend to decompose the peroxide of 

 hydrogen, and to restore it to the state of protoxide or water. I 

 know only four which do not sensibly possess this property ; 

 iron, tin, antimony, and tellurium. The most oxidizable are 

 oxidized, and at the same time produce a disengagement of 

 oxygen. The others, on the contrary, retain their metallic state, 

 KO that all the oxygen with which the water combines to become 

 peroxide is liberated. 



In order to effect the decomposition readily, it is indispensably 

 necessary that the metallic matter should be finely divided. Any 

 metal which in the state of fine powder readily disengages the 

 oxygen of the peroxide, effects it very slowly if the powder be 

 coarse, and still more so if it be in mass. 



The same phenomena occur even when the peroxide is diluted 

 with water, excepting that they are less distinct, and continue 

 longer. This will appear from the examination which I am 

 going to state with respect to the action of metals upon the 

 diluted peroxide. 



The experiments were all performed in the same way. The 

 liquid was first put with a small pipe into a little glass tube 

 closed at one end, after which the metal was introduced. The 

 <juantity of peroxide employed in each experiment amounted 

 only to a few drops ; when diluted with water, a larger quantity 

 was employed. The action was considered as complete, when 

 no more gas was evolved ; and this was rendered certain by the 

 addition of a small quantity of oxide of manganese. All the 



