During the past two years several reservoirs containing algae growths were treated with copper sulphate under my directions. The amount of copper sulphate was applied to meet the different conditions; the quantity used varying from I part in 10,000,000 to I part in 1,000,000 parts of water. In all cases twenty-four hours after the treatment, 1,000 c. cm. of the treated water being evaporated to 100 c. cm. were tested for copper. Fhe potassium ferro-cyanide test was used. When a solution of potassium ferro-cyanide—K Fe Cy (yellow prussate of potash) is added to a solution containing copper, a chocolate-colored precipitate of ferro-cyanide of copper, Cm Fe C will immediately form. Samples of water taken from reservoirs twenty-four hours after the copper treatment and tested by the evaporation method gave negative results. The reason for this is readily seen, as the copper is absorbed by the vegetable matter and settles to the bottom of the reservoir. As considerable time is consumed in evaporating 1.000 c. cm. of water, and, in some cases, it being impossible to discontinue the use of a reservoir for any great length of time, I have resorted to the use of Nessler tubes (same as used in the determination of color) for the copper test with good success. In making the test by this method, it is necessary to take a sample of the water before the copper sulphate is applied. This sample is used as a standard. When using this method for the copper test, a solution of potassium ferro-cyanide is added to the raw and treated waters after they are in the Nessler tubes. If any copper is present it can be easily detected by an experienced eye. As the potassium ferrocyanide solution is very yellow, it is absolutely necessary that the same quantity of the solution be added to both the raw and treated waters. The evanide solution being the heavier, it is only necessary to add it to the water in the Nessler tubes, as the inixing will readily occur; the comparisons of 1 he solution of potassium ferro-cyanide used was made by dissolving twenty gins, of the salt in 1 litre of distilled water; 2 c. cm. of this solution were used as an indicator for every 50 c. cm. of water examined. Table showing proportion of the colors can be made within two minutes after the solutions are mixed. Laboratory tests show that 1 part of copper sulphate in 10,000,000 parts of water can be easily detected by this method. copper sulphate in water, color and the intensity of the cyanide reaction:


•Paper read at the convention of the American Water Works association at West Baden, May, 1905.

*r-d, readily detected, tv-p, very pronounced.

No change could be observed in the turbidity of the samples after standing thirty-six hours. The water used during these experiments was taken from the laboratory tap, and contained the following in parts per T.000,000.

After the samples were taken, they were allowed to stand in the laboratory twenty-four hours before the copper sulphate was added. The reason for this was that the water would then be about summer temperature. The color of the water was reduced 22.9 per cent, when the copper sulphate was used at the rate of 1 part in When the copper sulphate was used stronger proportion~~ the color increasei. i ins increase was undoubtedly caused by the ferrous sulphate in the copper sulphate becoming oxi dised by the dissolved oxygen of the water. During these experiments a commercial copper sulphate was used which contained: Cu-2~.45 per cent-S04-37fi8 per cent.-H20-33.94—Fe-t r;t When the copper sulphate was applied at the rate of 1 part n i,ooo,ooo, the alkalinity (ery throcin metlu d) was reduced 3 parts p~r 1.000,000, while the hardness (soap method) was increased 0.36 parts per x,ooo,ooo. The increase in hardness was so slight that it can be disregarded. Any reduction in the alkalinity might be a very serious matter, especially if it were low in the water be treated, and the water were to be filtered by the mechanical process. If such a deficiency in the alkalinity should exist it can be easily overcome by the addition of a little lime or soda ash. Water containing copper sulphate (1 part in 1,000,000) 11!tercel through ordinary filter paper gave the copper reaction. Watcr containing copper sulphate (1 part in 1,000,000) to which sulphate of alumina had been added at the rate of 1 grain per gallon, did not give the copper reaction after being filtered through filter paper. The same solution, when tested without filtration, gave the copper reaction, which proves that the copper had been absorbed by the aluminum hydrate. This absorption is very important, as it shows that water, which has been treated with copper sulphate and is to be filtered by the mechanical filters, will yield a filtrate free from copper, if enough’ sulphate of alumina is used. In every instance where copper sulphate was used under my direction, only one treatment was necessary. Results of all treatments have been satisfactory. Since doing the original work which I reported in my paper I have made a series of experiments using sulphate of iron and lime as a coagulant and find that the ferrous hydrate precipitate will absorb the copper sulphate the same as the aluminum hydrate. I also find that the addition of lime will cause the copper to be precipitated. Waters, which have been treated with copper sulphate and are to be filtered by the American process, where ferrous sulphate and lime are used for a coagulant, will yield a filtrate free from copper. If the water is to be filtered by the English system, all that is necessary to prevent the copper appearing in the filtrate is to add a small quantity of lime. Potassium ferro-cyanide, when added to a solution of ferric salts will produce a dark blue precipitate of K (Fe Cy#)t~“Prussian blue.” When the iron is in the ferrous state it will give a light blue precipitate of Fe Kj (Fe Cy.). As nearly all waters contain iron in some form, the concentration of the water by evaporation is a serious bar to a successful cyanide tost for copper. When the water is concentrated, so that the iron is present at the rate of 3 parts per 1,000,000 and the copper sulphate 1 part in 500,000, the addition of the potassium ferroevanide solution will cause a blue precipitate, which will completely cover the chocolate or copper precipitate so that the test will be negative.

The installation of the new water system begun by Wilmington, Vt., in October last is almost completed. The supply is taken from springs on Haystack mountain, about three miles distant, and conveyed to a reservoir at an elevation above the town of about 200 feet. The supply is abundant and the pressure will be about 100 pounds.

No posts to display