164 Proceedings of Royal Society of Edinburgh. [sess, 
the above-mentioned effervescence was presumably due to the 
escape of the latter gas. To test this, ammonium persulphate in 
considerable quantity was dissolved in strong ammonia solution ; 
a small flask was filled almost completely with the solution, some 
silver nitrate added, and an india-rubber stopper with delivery- 
tube fitted to the flask, so that the evolved gas might be collected 
in a vessel over water. The evolution of gas began at once and 
increased rapidly ; at the same time the temperature of the liquid 
rose, and soon the action became violent. Ultimately the stopper 
and fittings were driven out, and most of the liquid blown out of 
the flask. 
The first quantities of gas had been allowed to escape, after 
which sufficient for examination was secured before the unexpect- 
edly sudden termination of the experiment. The sample contained 
a mere trace of oxygen, the presence of which was almost certainly 
due to the method of collection. 
The quantity of silver salt employed in this experiment would 
amount to only a few centigrams, and it is therefore evident that 
the silver must oscillate very rapidly between the two stages of 
oxidation in order to cause such rapid decomposition. Apparently 
we have here an admirable example of a ‘catalytic action,’ in 
which the part played by the catalytic agent may be considered 
as definitely known. The final result is expressible by the simple 
equation — 
3(JS t H 4 ) 2 S 2 0 8 + 8NH 3 = 6(NH 4 ) 2 S0 4 + N 2 
leaving the silver compound entirely out of account, but there 
seems no reason to doubt that the action takes place in the manner 
and stages indicated. 
The experiment is one very well suited for class demonstration, 
and is exceedingly simple. Dissolve a considerable quantity of 
ammonium persulphate in concentrated ammonia solution, and 
place the solution in a tall beaker or jar. Add a small quantity of 
silver nitrate solution ; the evolution of nitrogen begins at once, 
and soon the temperature rises so high that large quantities of 
ammonia gas also escape, causing the liquid to boil over ; the result 
is not nearly so striking if dilute ammonia solution is employed. 
The decomposition of an ordinary aqueous solution of ammonium 
