it is impractical to carry the decomposition to completion. It must be emphasized that the tank was by no means gas tight and gas was escaping continuously, through small cracks and along the edges. The quantity of gas produced-leaving aside loss from material in the tank was considerably below the quantities obtained in carefully controlled laboratory experiments. The possible amount of gas produced from material carried to complete digestion is at least 250 cubic feet per cubic foot dry organic material. In this experiment we obtained approximately 100 cubic feet per cubic foot organic matter. Since the possible gas production from material brought to complete digestion (under usual circumstances prevailing in good working tanks) is from 250 to 350 cubic feet per cubic foot organic matter, it would seem that during the period of gas measurements an equivalent of from 30% to 40% fresh solids added were digested (gassified). On this basis the tank would have taken care of considerably less fresh solids than it actually did. In a previous publication we have drawn attention to the fact that under certain conditions liquefaction can be induced. The laboratory experiments show that at a pH value of 7.3 liquefaction overbalances gassification.

The temperature of the tank contents, determined approximately half way between top and bottom of the tank, dropped gradually from 68°F in October to about 50°F in the middle of February. The temperature remained stationary and on February 22 hot water was pumped through a coil for the purpose of heating the sludge. No fresh solids were introduced and four days later the temperature began to rise. While the tank was being heated the superintendent kept a careful record of the amount of soft coal used. The experiment was continued for 30 days and from his records he calculated that it took one pound of coal per day to heat 1,000 pounds of material 13 degrees F., or, in other words, one pound of soft coal per day raised 1,000 pounds of sludge 0.4°F per day. The tank was well insulated and, moreover, our automatic records show that the air temperature was always lower than the temperature of the tank contents.

It has been mentioned above that the fresh solids were added every two or three days. This made storage of fresh solids in the sedimentation tank necessary and resulted in considerable acid production. The amount of hydrated lime added to the fresh solids to obtain pH values of 7.3 amounted to 71 grams per cubic foot fresh solids. According to our laboratory experiments, when fresh solids are added every day, the amount of lime should have been less than half of that used. If these same fresh solids had been stored for periods of 7 to 10 days before addition to the digestion tank the amount of lime necessary would be at least doubled again.

'N. J. Agr. Expt. Sta. Bull. 427.

The results obtained on the bacteriological part of this experiment showed similar fluctuations and an establishment of an equilibrium as has been previously reported from our laboratory experiments. The significant fact was that when conditions were good for digestion as indicated by low carbondioxide and high methane content, together with relative high alkalinity, the numbers of total bacteria were low, whereas, when the conditions for digestion were unfavorable, as judged by increasing carbondioxide and decreasing methane content of the gas and decreasing alkalinity, the numbers of total bacteria. showed an increased stimulation accompanied by a high average of lactose fermenting organisms. It has previously been concluded that conditions for favorable digestion are indicated by a low, uniform level of bacterial numbers.

The protozoan fauna was in general low and could be compared with the kind and numbers of animals found in the digestion compartment of an Imhoff tank. Only once during the half year of observation the protozoa count rose above 100,000 per cc. It will be remembered that our zoologist has pointed out frequently that high numbers of protozoa in the digestion chamber is a good indication of poor working conditions.

It is of interest to compare the general efficiency in the digestion of sludge of an Imhoff tank and in this covered separate sludge digestion tank. Roughly the covered separate sludge tank digested satisfactorily 100,000 cubic feet of fresh solids (95% moisture) during 180 days in winter time at an average temperature of 55.5°F. The quantity of fresh solids added was equivalent to 555 cubic feet, or 1,734 pounds per day on a dry basis. Using the graph of Mr. Downes' presented by him in the discussion of Mr. Eddy's paper on the "Behavior of Imhoff Tanks," we find that the 25,000 cubic feet capacity of the tanks was equivalent to 100 days accumulation of sludge in the various stages of digestion. The theoretical advantage of raising the slot of an Imhoff tank so as to increase the effective sludge storage space to 150 days has been shown by Downes to be insufficient for Plainfield. However, calculations made in our laboratory showed that this tank had received more fresh solids than it could handle properly. It seems, nevertheless, fair to assume that at least 100 days sludge storage capacity is required for good working Imhoff tanks and since the covered separate sludge tank had an equivalent to 100 days sludge storage, it might be concluded that this tank did as well as a good working Imhoff tank. However, this does not mean that the ultimate capacity of the separate sludge digestion tank was reached because (1) the fresh solids were added to the tank as the need for space arose rather than to the necessary additions to keep

"Trans. Amer. Soc. Civ. Eng. v. 58, p. 465-510.

'Public Works, v. 54, p. 363-365.

the relation between ripe sludge and fresh solids as nearly correct as possible, or, in other words, to hold the biological equilibrium at an optimum and (2) Mr. Downes' calculations for an Imhoff tank were based on summer temperature, whereas the results obtained with the separate sludge digestion tank were accomplished during the winter months, with an average of 10°F lower.

If we assume that the correction of the reaction of the fresh solids by the addition of lime compensated for the higher temperature in summer, we still find that the separate sludge digestion tank was as effective as a good working Imhoff tank. This assumption is very unfair to the separate sludge digestion tank, since a good working Imhoff tank does not receive more fresh solids than it can handle and reaction adjustment would be rarely necessary. The same holds for a separate sludge digestion tank provided the fresh solids are really fresh when added and the biological equilibrium is not upset.

It appears thus that with proper handling not more sludge capacity will be necessary for separate sludge digestion tanks than for Imhoff taaks.

Whether Imhoff tanks or separate sludge digestion tanks are to be used will depend upon their cost (construction and operating), on their flexibility and their ease of control. The reaction of tank contents is an important factor in digestion and appears to be controlled with a greater degree of accuracy in separate sludge tanks. The daily addition of definite quantities of fresh solids to ripe sludge. seems simpler in separate sludge digestion tanks than in Imhoff tanks. Submergence of scum by weighing down and the collection of gases will be equally feasible in both types of tanks, but it seems easier and certainly more economical to apply heat to the sludge to be digested in a separate sludge digestion tank than to the sludge of an Imhoff tank.

CONCLUSIONS

The following conclusions might be drawn:

1. Since it was possible with 1 pound of soft coal per 1,000 pounds of sludge to raise the temperature of the sludge 0.40°F per day it will cost probably less to maintain a higher temperature when heating is begun as soon as the temperature drops.

2. If really fresh solids are added every day the amount of lime necessary for adjustment will be small because the material in the tank in a proper biological balanced condition will maintain an optimum reaction for digestion.

3. A properly insulated separate digestion tank under good control is at least as efficient as an Imhoff tank, consequently sludge digestion capacity will be about equal for both types of tanks.

DISCUSSION

PRESIDENT HATTON: What is the pH value of your raw sewage? DR. RUDOLFS: The average pH values of the Plainfield sewage is 7.2-7.5 but the pH of the settled solids depends upon the time they have been in the tanks. If it is collected for one day it would be, ordinarily, 6.4; two days it would be 5.9; three days 5.6-5.3; four days or more it would run 5.3 to 5.1; this means that for every day's waiting we have to increase the amount of lime necessary to adjust the sludge to the optimum reaction.

MR. HAWLEY: What is the role played by protozoa in a digestion tank?

DR. RUDOLFS: The protozoa present in considerable numbers in a digestion tank are small flagellates, non-bacterial feeders, who imbibe their food by osmosis through their skin. We find that they are always present in large numbers when the tanks are acid. In the laboratory we have found that they increase rapidly when organic acids are added to the liquid of a digestion tank. Since they multiply most rapidly when a tank is acid and they produce carbondioxide the tank has a tendency to become more acid. The more acid the more protozoa of this kind and the more protozoa the more acid the tank becomes. Our zoologist has stated several times that the numbers of protozoa in a tank furnish a good index of the working of a tank; but what the exact role is of these protozoa I am not prepared to say.

THE PROBLEM OF CONCRETE DETERIORATION, MORE ESPECIALLY WITH REFERENCE TO ITS USE FOR

CONDUITS AND IN LINING IRON PIPES.

By John R. Baylis, Water Safety Division, Chicago, Ill. This article will deal more with the fundamental reasons. for changes taking place in Portland cement concrete after it has set than to specific instances of deterioration of the concrete in conduits and pipes. That a large percentage of the concrete constructed for such use deteriorates fairly rapid is well proven by the literature and from personal observations. There is a feeling amongst many engineers and other users of concrete that concrete is practically an indestructible product insofar as this relates to its exposure to water and the weather. So firm is this belief we see a great deal of concrete being manufactured with no thought or consideration given to its limitations.

When exposed to the weather without protection we know that iron rusts and that wood rots, yet it seems that many believe concrete will endure forever. It is true that some of the constituents of concrete are durable, however, one of the main constituents which gives cement its binding power will also give up calcium hydroxide to approximately the saturation point of limewater. A piece of burned. limestone (CaO) sealed in a perfectly tight container will remain unchanged indefinitely, but if submerged in water it will dissolve fairly rapid if there is quick replacement of the water. It does not seem to have occured to many that hydrated Portland cement will also give up calcium hydroxide to approximately the saturation point of calcium. hydroxide when submerged in water. It is true the rate of solution is not so fast as for lime, but the fact that it produces in water approximately the same concentration of calcium hydroxide as does lime, unless some of the calcium has been extracted or carbonated, is significant. Over 90 per cent of the calcium in Portland cement can be extracted by submerging in rain water or in practically any of the natural waters. Hydrated Portland cement is not even near nature's