It has been found by analysis of air containing large numbers of bacteria that showers greatly reduce their numbers, as it reduces the amount of dust in the air, and that prolonged rains may clear the air of dust and bacteria entirely. This accords with the action of the water and wet surfaces of sewers in filtering the air of germs, and if the sewers and house drains are long and wet enough in proportion to the amount of air passing through them, all dust and germs may be filtered from this air entirely.

*Frankland's "Bacteria in Daily Life."

Laws says "It is really remarkable to find that no organisms are given off from the walls of a sewer which has been empty and open to the air at both ends for such a lengthened period as twelve days. The sewage with which the sewer had been kept full for several periods of twenty-four hours would contain no less than three to four million organisms per cubic centimeter, and immense numbers of these must of necessity have been clinging to the walls of the sewer. * * * The velocity of the air current used in the above experiments was five feet and fifteen feet per second, respectively, the latter being far in excess of any current that would normally obtain in a sewer."

"Various experiments," says Roechling, "have been made with a view to ascertain how far germs can be carried away by air currents in pipes and sewers. Hesse, who first investigated this point, took a 2-inch glass tube about one yard long, the inside of which he had covered with a layer of nutritive gelatine, and sucked air through it at a slow rate. When examining the tube afterwards he found that a large number of bacteria had settled in its first fourth, that the number was somewhat less in the second fourth, and that it still further decreased in the third fourth, and that no bacteria at all had settled in the last fourth."

Ficker remarks that in his experiments in the Hygienic Institute at Breslau a current of air, with the velocity of several meters per second, was not able to lift up specific germs from half-moist soil, and that a current of the same strength was not capable of carrying away germs which had dried on several substances and adhered to them.

Author's Experiments and Suggestions for Improving the Air of Cities.

Various kinds of dust, from fine lint up to fine and coarse sawdust, were used successively in my own experiments. Dry powdered substances of different kinds, such as whiting cement and fine flour, were also used. Bacteria, being fine particles of organic matter, resemble these fine dusts in the manner in which they may be wafted about on air currents and are retained by water.

In no case could any of the particles which came in contact with the wet sides of the pipes be seen to be blown off again. If any escaped they were too few and small to be detected. The bellows were large and strong, capable of producing at will either a powerful draught through the pipes or a faint and almost imperceptible current. With the short pipes it was possible to drive some of the particles through in a strong draught because these did not have time to come at all in contact with the moistened surfaces, but with pipes twenty-five feet long and one and a half inch in inside diameter none could be forced through even under a pressure strong enough to put out the flame of a candle at the further end. Where bends were introduced the same result was obtained with shorter pipes. Fig. 51.

With gentler pressures still shorter pipes sufficed to entirely filter the air of all particles, however dry and finely powdered; and, in the very slowly moving air currents found in the average sewer, a very short travel is sufficient to arrest all dust or bacteria of every kind.

The same experiments were then tried with the pipes thoroughly dried on their inner surfaces. In these cases all of the dust was easily blown through all the pipes after a more or less prolonged application of the air pressure, the surfaces of bends arresting for some time some of the particles by back eddies, but all eventually passed through.

With jointed pipes, however, as might be expected, some little of the dust and powder would be permanently retained in the fine cracks of the joints.

The interiors of the pipes were next coated with sewage and moistened, and the experiments were repeated under these conditions with the same results obtained in the first experiments with pipes moistened with water alone. No particles once coming in contact with the moist surfaces could ever be detached again, however strong the air current, so far as could be observed.

Only the particles of dust which happened to travel from end to end in the middle of the air blast escaped from the pipes, and, as before said, this only occurred with short pipes and in currents stronger than those which prevail in ordinary sewers. As has been said, it is generally conceded that bacteria may rise through water in the center of bubbles bursting on the surface, or in droplets caused by agitation or spraying. But in well ventilated and properly constructed sewers such splashing and bubbling may be largely avoided. But should they occur, the very few bacteria thereby escaping into the air of the sewer could only travel a short distance, as we have shown, before they would again be caught and, if of the pathogenic kind, be quickly destroyed.

Finally, the pipes containing the particles of dust of various kinds imprisoned in the moisture and organic refuse along their inner surfaces were thoroughly dried by exposure for several days to the dry air of the laboratory, and then the air blasts were repeated to see whether the dried particles of matter caked against the walls of the pipes would give up the dust and bacteria blown against them when they were wet. None of the dust escaped, so far as could be seen. Under the most powerful blasts pieces of caked refuse would occasionally be torn off the surfaces to which they would normally adhere, but these were heavy and fell directly to the ground as soon as they were blown from the pipes. The dust particles still remained firmly imprisoned in these masses of scale and were therefore as powerless to do injury as if they had never been detached from the part of the pipe against which they originally fell. The masses containing the separate particles of dust or bacteria were simply transferred from one part of the pipe system to another. The reason for this is obvious. Wet sewage, containing many kinds of glutinous and fatty matters, forms, when it dries, a more or less pasty mass, which hardens into a tenacious cake, and in these experiments appeared to hold the dust or bacteria enmeshed in its mass as firmly as it did when it was in its moist state.