The samples at thirty and thirty-five feet had a strong odor of sulphuretted hydrogen.

The odor of the samples which contained no oxygen was not offensive, simply earthy.

DISSOLVED OXYGEN IN RESERVOIR No. 3, BOSTON WATER WORKS.

DISSOLVED OXYGEN IN RESERVOIR No. 4, BOSTON WATER Works.

The contrast in the condition of the water in these two reservoirs is very striking. Reservoir No. 3, in which the oxygen is exhausted at a depth of fourteen feet, receives a not inconsiderable amount of direct pollution from the towns of Marlborough and Southborough, while the drainage area of Reservoir No. 4, as has been already said, is very sparsely populated.

In connection with the determination of dissolved oxygen in the Ludlow reservoir of Springfield (which contains at all seasons of the year an excessively large number of organisms in suspension in the water) an attempt was made to ascertain whether the amount of dissolved oxygen decreased in the water in its passage through the pipes. It was found that the amount of oxygen was somewhat less after a passage of six miles in the pipes than in the reservoir, and still less after three miles additional passage in the pipes. But the differences are not very great, and it is possible that the loss of oxygen may in part be due to other causes than the decomposition of organic matter.

The reservoir is only ten or thirteen feet deep at the points where the samples were taken, and the comparatively large amounts of oxygen found in the bottom layers of the water indicate, probably, that the water in the reservoir was in circulation by means of the wind.

From Faucet at Indian Leap Hotel, after passing through six miles of pipe.

From Faucet at Winchester Park, after passing through three additional miles of

Some of the water on leaving the pipe passes into an open ditch five hundred feet in length, and then flows swiftly through a trough with a steep grade into Van Horn reservoir. A sample taken at the beginning of the ditch showed 77.49 per cent. of dissolved oxygen, and a sample at the end of the trough 91.28 per cent. These results are both somewhat too high, owing to the impossibility of collecting the samples without access of air; but the greater amount of dissolved oxygen in the water as it flows into the reservoir may be said to represent fairly the effect of aeration in the open ditch or trough.

Van Horn reservoir receives, in addition to the water of Ludlow reservoir, a certain amount of water from its own water-shed (about two hundred and fifty acres), which enters the reservoir mainly after filtering through the ground. Following is a series of determinations of dissolved oxygen in the water of this reservoir at different depths:

The water from the depth of seventeen feet had an earthy odor; that from the twenty and twenty-eight feet layers, a strong odor of sulphuretted hydrogen.

A consideration of the foregoing determinations of dissolved oxygen in the waters of ponds and reservoirs of different characters leads us to two important conclusions: —

First. That the stagnation of water is not in itself objectionable. Second. That when a water contains decomposable organic matter, or where water is stored in reservoirs having decomposable matter on its bottom and sides, the oxygen in the water will be quickly used up, and if not renewed the water will become foul.

Reservoir No. 4 of the Boston water supply is the only deep reservoir among those examined which contained some dissolved oxygen (15 per cent.) in the stagnant layer, and this was as late in the summer as August 20. It is not improbable that many natural unpolluted ponds, with sandy bottom, would show even a larger amount of oxygen in the stagnant layer than Reservoir No. 4. If there is nothing to use up the oxygen in solution in the water there is no reason why it should not remain indefinitely. There is always, however, some organic matter in all surface waters, and it is probable that the stagnant layer even in the best ponds is never fully saturated with oxygen. The ordinary process of decomposition of organic matter is one of oxidation and is not necessarily accompanied by offensive odors. In all surface waters which contain vegetable and animal life decomposition is constantly going on, and yet in the presence of an abundance of oxygen there is seldom any objectionable odor; but when the oxygen is exhausted putre

factive changes set in, and the products then formed are highly offensive. It is under these conditions that sulphuretted, carburetted and phosphoretted hydrogen are developed. These gases we find in water only when oxygen is absent.

In practical water-works management advantage may be taken of the fact of the formation of a stagnant layer in the bottom of deep ponds and reservoirs by drawing the water from near the surface in those cases in which the stagnant layer is foul; but the deterioration of the surface water cannot be prevented when the mingling of all the water takes place in the autumn. In deep storage reservoirs, which must be drawn very low in order to maintain the supply, it may become necessary to draw from the stagnant layer, and if this layer contains products of putrefaction bad water is supplied.

It is not uncommon to find iron in solution, in considerable amount, in the bottom layers of water in deep ponds, when the oxygen is exhausted. Insoluble iron oxide may, under these conditions, be reduced to the lower oxide, and go into solution. In such cases the water may be colorless when pumped up from the bottom but it quickly becomes turbid on exposure to the air, and iron oxide separates out, on standing, as a red deposit.

No stronger argument for vigilance in preventing the contamination of a water supply and for its storage in clean reservoirs could be made than is contained in the results of the foregoing investigation of the changes which organic matter undergoes in water, involving the exhaustion of the dissolved oxygen and the formation of putrefactive products. How far we may expect to remedy the evils consequent on the presence of decomposing organic matter in water by aeration will be discussed in the next chapter of this report.