1899-1900.] Dr Marshall on the Action of Silver Salts. 165 
persulphate in presence of silver salts is much slower than the 
above, and it appeared interesting to get some idea of the rate at 
which it took place as compared with that of a solution free from 
silver. So far it had been assumed that the products would simply 
be oxygen and ammonium hydrogen sulphate. 
Twelve grams of recrystallised, but not quite pure, ammonium 
persulphate were dissolved in water at 20°C., and, after the addition 
of 0-0125 gm. of silver nitrate (corresponding ultimately to a milli- 
gram equivalent per litre of solution), the solution was diluted to 
250 c.c. The solution was kept in a thermostat at the temperature 
stated. From time to time, 5 c.c. were withdrawn and titrated 
with fifth-normal alkali solution, using methyl orange as indicator. 
In the earlier titrations, when there still remained a good deal of 
undecomposed persulphate, the indicator became rapidly bleached, 
and in each of these cases it was found advisable to repeat the 
determination, adding the indicator only when the neutral point 
was nearly reached, as known from the first determination. 
Time. 
Vol. of 2N alkali 
for 5 c. c. of solution. 
Time. 
Yol. of -2N alkali 
for 5 c.c. of solution. 
Od. 5h. 
0-7 c.c. 
8d. 5|h. 
11-25 c.c. 
1 
H 
3*15 
9 
^4 
11-6 
2 
2 
5-45 
10 
6 
11-8 
3 
3 
7-2 
13 
Qh 
12-3 
4 
4 
8-5 
16 
6 
12-55 
5 
5 
9-5 
35 
2 
12-75 
6 
7 
10-3 
These results are plotted in the figure (see Plate), the curve 
showing the increase of acidity as the experiment progressed. 
It soon became evident that the reaction was not taking place 
in the way imagined. Allowing for the small quantity of sulphate 
in the sample of salt employed, the persulphate solution was slightly 
over 0*4 normal. Therefore it should ultimately have produced a 
slightly more than 0*4 normal acid solution, assuming the final 
decomposition to be expressible by the equation — 
2(NH 4 ) 2 S 2 0 8 + 2H 2 0 = 2(NH 4 ) 2 S0 4 + 3H 2 S0 4 + 0 2 . 
The quantity of alkali solution required for 5 c.c. of the liquid 
should therefore have approached a limit of slightly over 10 c.c. 
