10 Dr. A. R. Leeds on the Determination 
After two days' exposure, no precipitation of reduced silver 
occurred in the neutral solution of argentic nitrate, nor in that 
of the ammonio-argentic oxide, nor in the ammonio-argentic 
chloride, nor in the sodio-argentic hyposulphite. 
Having thus established the fact that no reduction of silver 
occurs in case organic matter is rigidly excluded, even after 
prolonged exposure to sunlight, the next point of inquiry was 
to determine which of these five solutions was most affected 
by such organic matters as are ordinarily present in potable 
water. The inquiry as to the relative energy of action of 
different kinds of organic matter in general did not neces- 
sarily form a part of the present investigation. 
In the study of this point, a sample of potable water was 
made use of from the river Schuylkill, the water-supply of the 
city of Philadelphia. This drinking-water had been affected 
by a sudden cachexy in the month of January, becoming 
most offensive to taste and smell ; and the authorities of the 
city had requested me to investigate the causes of its non- 
potability. The scope of the inquiry being thus narrowed to 
the finding of specific causes, I was necessitated to tax the 
capabilities of the methods at present possessed by water-ana- 
lysts, and was furthermore led to test this new one. Now on 
adding, to the five silver solutions prepared as in the first 
series of experiments, 100 cubic centim. of this Schuylkill 
water, it was found, after 5 hours' exposure to sunlight, that 
the ammonio-argentic oxide was slightly changed, the potassio- 
argentic cyanide was unaffected. Of the remaining three, 
the sodio-argentic hyposulphite threw down the heaviest pre- 
cipitate, the ammonio-argentic chloride a precipitate nearly 
as great, and the neutral solution of argentic nitrate the 
least. 
From these experiments it would appear that the solution 
of argentic hyposulphite in excess of sodium hyposulphite is 
the reagent best adapted for use in this actinic method. But 
subsequent experiments showed that the neutral solution of 
argentic nitrate fulfilled all needed requirements, and that the 
labour of preparing a special reagent was uncalled for. The 
method finally adopted is as follows : — 
250 cubic centim. of the natural water is treated with 10 
cubic centim. of decinormal silver solution in the tall glass- 
stoppered cylinders or bottles. The waters become turbid and 
frequently coloured ; but after a moderate interval, usually 
less than two days, the turbidity entirely disappears and the 
entire precipitate collects at the bottom of the vessel. In case 
sufficient silver has been employed ^and the amount recom- 
mended is ordinarily a large excess), this clarification indicates 
