So far, the air passed through the pipes was pure air from the room, and in this respect the conditions differed from those met with in practice in which soil-pipe air is used for branch waste ventilating.

To eliminate this difference the tests were next made on waste-pipes actually in use, and the ventilating current was then taken from the soil-pipe into which they emptied in the usual manner. The pipes containing the deposits were connected with the branch waste below the trap, and the deposits were moistened three times a day by means of a simple contrivance, enabling the moistening to be effected without removing the pipes or disturbing the deposits. The waste-pipe was ventilated above into a heated flue in such a manner as to draw the soil-pipe air. over the deposits at the rate of about seven feet a second.

After an exposure of one week to this current the deposits were dried and weighed as before. No. I was found to have lost 0.0575 grams, and No. II 0.0352 grams. This was equivalent to a loss of less than 1/70 of the entire weight of the deposit in No. I, and 1/100 in No. II.

Either of these amounts dissolved in water and spread uniformly over the surface of clean pipes of the size of those used, is found to be altogether imperceptible to the eye, and the complete purification of these pipes by ventilation under the most favorable circumstances would at this rate require from 70 to 100 weeks, or from 1 1/2 to 2 years, supposing there were no addition made to the deposit during the interval through use of the fixture.

Let us now make the tests under conditions less favorable to oxidation, namely, at the normal temperature of our houses, and with the deposit dry, as might be the case in pipes under fixtures temporarily in disuse.

We will continue to use a ventilating current of an average velocity of eight feet per second.

The weight of the soil in No. III, when dried, was 3.7070 grams. That of the soapy mixture in No. IV was 2.9085.

After two days' exposure to the air current the former was found to have gained in weight 0.1355 grams, and the latter 0.3860 grams. After the third day's exposure there was a loss of part of the weight gained during the first two days, namely, of 0.0820 in the soil deposit, and of 0.0060 in the soapy mixture. After the fourth day's exposure the deposits once more gained in weight. Undoubtedly, if the experiments had been continued for several weeks or months, an infinitesimal aggregate loss would have been perceived. Temporary gains and losses would result from changes in the moisture of the atmosphere.

Thus, we found in our tests that there was no appreciable variation in weight at all due to loss through oxidation, and that in three of the observations out of four there happened to be a gain due to absorption of matters in the air rather than a loss of weight.

Hence we see that only under the most favorable conditions, or when the deposits are spread out in an extremely thin film and left undisturbed for an indefinite length of time, can ventilation produce the slightest appreciable change in them.

II. - Rate of accumulation of deposits.

Our next task, then, is to determine how rapidly the deposits are apt to be formed in plumbing. To answer this no special experiments are needed. We have ample data in our every day's experience with waste-pipes in use.

Considering, first, the worst conditions, namely, those of the cold waste-pipe from the ordinary kitchen or pantry sink, we know that the accumulation of grease in these will be so rapid as to entirely clog up the pipes in a short time where special precautions are not taken to flush out the pipes from time to time with hot water or some solution of caustic alkali. In fact, the well-known and generally admitted tendency of the ventilating current in these cases is to congeal the grease and increase the clogging rather than to diminish it.

Consider next the case of an ordinary soil-pipe. We find that the tenacious soil will adhere stubbornly to the pipe in clots or masses wherever it strikes until it is washed away by a powerful fall of water, and that it is not equally distributed in a thin film all over the surface. Parts will be found which are never touched by the foul matter, and parts that are alternately fouled and then scoured clean again. Generally thick masses of deposit will be found in the cavities of the joints or in holes in the castings. In short, the deposits in soil-pipes are not slowly distributed favorably for oxidation, but are formed in lumps suddenly, and are either as suddenly removed by the flushing water, or are deposited in cavities which largely screen them from the influence of the ventilating current, and therefore, in this case also, the influence of aeration in removing the solid matter is comparatively very slight. The accumulations of heavy matter will continue in time to increase until they leave an opening only large enough to allow room for the ordinary water-flushing stream to pass.

Take next the waste-pipe from a lavatory. We find the solid deposits here of two kinds, one collecting in clots or masses in corners or unscoured areas, as of lumps of soap, hairs, lint, etc., and the other coating the pipes in thin films as of soap-suds. The former are deposited suddenly, and are either swept away by the water or are caught in the unscoured cavities and remain there, partially screened from the air current until other similar substances accumulate above them. The ventilating current, therefore, can have no appreciable effect in removing these masses of matter.

Where, on the other hand, the traps and waste-pipes are so constructed and flushed that no such masses can collect, the only kind of deposit that can form in the interval between the flushings will be of the second kind, namely, a thin film of matter like soap-suds.

It remains to be seen what effect a powerful water-flushing has on these deposits, and this brings us to our third consideration.