APPENDIX.

[A.]

METHODS OF ANALYSIS.

Collection of Samples.-The water was collected in glassstoppered bottles which have never been used except for this purpose. The water was taken, as a rule, at the depth of eighteen inches below the surface, and the bottles were thoroughly rinsed with the water to be collected before being finally filled.

Suspended Matters.-Various methods are in use for determining the amount of matter in suspension; none of them are perfectly free from objection. If there is considerable suspended matter it may be collected on a tared filter and weighed; this method is, however, quite inaccurate where small quantities are in question. The method of allowing the water to stand for some days or weeks and then collecting the sediment, is open to the very serious objection that even in the case of tolerably pure waters the chemical changes which immediately begin to take place alter very decidedly the amount and character of the suspended matters, an effect which is much more marked in the case of waters rendered impure by admixture with sewage or manufacturing refuse. The following method, as affording greater accuracy, has therefore been followed :-A portion of the water, well shaken up (100 or 250 c.c. according to its quality), is evaporated to dryness, and the weight of the residue determined; a quantity of the original water is then filtered through the best filter-paper, that which filters through at first being rejected, until the filter has been thoroughly washed with the water under examination. A portion of the water thus filtered is

evaporated to dryness, and the difference between the results obtained in the two cases may with sufficient accuracy be taken as the "suspended matter." In the foregoing tables the results of the analytical determinations are given just as obtained, and the suspended matter is not tabulated as such. When the amount of suspended matter was not large enough to estimate, the water was still passed through a filter in order that the results might be directly comparable with each other.

Solid Residue.-The amount of water taken for the determination of the solid residue was 100 c.c. or 250 c.c. according to the purity of the water. The evaporation of larger quantities tends to inaccuracy, as the operation is of necessity more prolonged, and in many cases the organic matter undergoes chemical change when subjected for a long time to the temperature of boiling-water. The residue was dried at 100° C. (2120 F.) and weighed. It was then ignited at as low a temperature as possible, was treated with distilled water saturated with carbonic acid, and dried again at 100 degrees. The difference between the two weights appears in the tables as organic and volatile matter.

This determination of organic matter is sufficiently accurate for practical purposes, except when the water contains a considerable amount of nitrates and chlorides.

Chlorine. The chlorine was determined volumetrically by means of a standard solution of nitrate of silver.

Ammonia and Albuminoid Ammonia.-The method of water analysis proposed by Messrs. Wancklyn, Chapman and Smith, of determining the amount of ammonia by distillation with carbonate of soda, and subsequently heating with a strongly alkaline solution of permanganate of potassium has been followed as affording the best and most available means at present at our command for obtaining indications of the amount of the nitrogenous organic matter in solution.

Oxygen. The oxygen dissolved in the waters has been determined in several instances,-generally by the use of a titrated solution of hydrosulphite of sodium, as suggested by Schützenberger and Risler (see Bulletin de la Societé Chimique, XX. (1873), p. 145.) This method supplies a want long felt, of some ready and convenient means of

determining free oxygen, and the operation can be performed, without difficulty in the open air, as for instance, in a rowboat or in other similar situations. For the determinations made during the summer, I am indebted to Mr. T. E. Pope. As, at that time, the complete details of Schützenberger's process, as finally adopted, had not been made public, we employed a method in some points modified from that originally suggested. To test the accuracy of the method employed, the following determinations were made, in which the results were compared with those obtained from the same water by the ordinary method of analyzing the gases dissolved by water (a Sprengel's mercury pump was used) :—

No. 105. Charles River water collected at Needham on September 5, was examined the same day, and found to contain

No. 106. Charles River, Newton Upper Falls. The water was collected on September 5, and examined on the following day. The results were—

No. 119. Inlet-chamber of filtering-gallery, Lowell, Mass. Collected September 10, examination made in the laboratory

No. 121. Merrimack River, Lowell, opposite inlet-cham

ber of filtering-gallery. Collected September 10, examined September 11.

The determinations recorded in December, January and February were made by Miss E. H. Swallow according to Schützenberger's method, as finally adopted by him. The following determinations were made to test the accuracy of the process:—

No. 1. Cochituate water, as drawn in the laboratory of the Institute of Technology. The temperature of the water was 7° C. The results obtained were

No. 2. Cochituate, drawn January 17. Temperature,

5° C. The results obtained were