rates in different cities. It also shows in each city an increase of from 10% to 100% in consumption during each decade. Neither the increase of per capita consumption nor the difference in rates of increase in the various cities seems to follow any law, except that the former shows a constant advance. It might be expected that the per capita consumption would be greater in cities where there was considerable manufacturing or many well-kept lawns than where these conditions did not exist; and this is the general rule-with many exceptions, however. Also large cities usually have a higher rate than small ones; but this rule also has many exceptions. For each particular case the daily consumption should be obtained from the water-works record, or, if there are no records of consumption for that locality, a careful selection should be made of the per capita consumption of a city whose conditions closely resemble those of the place in question. From these the per capita rate will be obtained. In order to be on the safe side the present rate should be increased by at least 25% to allow for a probable increase in consumption, since the construction must serve not only the present population, but that of the next 30 or more years. If meters are used on a majority of the services a great reduction in the consumption can be effected-from 30% to 60% in most instances. Unless water-meters have become generally established and accepted, however, no allowance should be made for the reduction in sewage due to their use unless the average daily rate exceeds 100 gallons per capita. There is no reason for a daily rate exceeding this amount, and the present tendency is to meter supplies before they reach this point. An allowance of 100 gallons will be made in calculations in this work, as being a safe one for any but exceptional cases. The average daily consumption, however, " is not uniform throughout the year, but at times is greatly in excess of the average for the year and at other times falls below it. It may be 20% or 30% in excess during several consecutive weeks, 50% during several consecutive days, and not infrequently 100% in excess during several consecutive hours." (J. T. Fanning, "Water Supply Engineering.") Many waterworks engineers use 75% excess as an average. This gives for a maximum flow, on a basis of 100 gallons daily, a rate of 175 gallons per capita daily = .1215 gallons per minute = .00027 cubic feet per second. It must be most urgently insisted, however, that each case should be studied by itself in the light of all the data avail able. These figures are given as approximate averages only, to be used in designing when no local records exist. It should also be borne in mind that the consumption given is an average including that used in manufacturing and for all other purposes. These last constitute a very uncertain portion of the whole, but unless there were definite figures obtainable it would not be safe to reduce the average by more than 10% to obtain a rate for residences only. As the assumed maximum rate-175 gallons-was but a roughly estimated average, it may be used unchanged for residential districts; and where factories are to be provided for a study should be made of the processes employed in them in order that a close approximation may be made to the amount of sewage to be expected from each. The amount of house-sewage from buildings (other than waste water from factories and water-motors) which will reach any particular sewer will depend almost wholly upon the number of persons contributing to this amount. For a district or city this number may be obtained in two ways. -by estimating the ultimate number of residences and assigning a certain number of occupants to each, doing the same with factories, stores, and other. buildings; or by estimating the probable ultimate population per acre for different sections of the city. The former is the more accurate for built-up sections; the latter sufficiently so for undeveloped territory or that which will probably undergo a change in the character of its buildings. For use in calculating by the first method the following table adapted from the U. S. census of 1880 is given.* There are in each city certain districts in which the population is much more dense than is indicated by this table. One hundred persons in one dwelling is not an exceptional rate in certain portions of New York City. For an ordinary residence district six persons to each dwening is a * See also Table No. 31, page 430. sufficient average. water for manufacturing purposes the maximum hourly rate per capita of occupants is not nearly as great as in the case of residences; a maximum rate of 20 gallons per day will be sufficient allowance for ordinary cases, being contributed by water-closet flushes, urinals, and wash-basins. One person In factories and stores which do not use to each 50 square feet of floor-space may be taken as a maximum density for factories, and one to each 75 square feet for office-buildings. A method frequently used is that of adding a percentage of increase to the present population of each city or section. American towns under 50,000 population have been found as a general rule to double in size in about 15 or 20 years. Having ascertained for each case its past rate of increase and present population, these are taken as the basis for calculations. But this increase is far from uniform over the entire area of a town, differing in different sections; also after a section has reached a certain density of population it remains practically stationary, unless its character change-as from residential to business or manufacturing. The percentage of total growth of a town may be used, however, as a check upon the sum of the populations assumed for the various sections. The law of increase varies in different cities, but that followed in the past by the one under consideration having been obtained from the records can be projected into the future, it being assumed that this law will remain constant (see Art. 129). Considerable judgment must be used in locating divisionlines between sections and assigning to each its density of ultimate population. The most hilly sections will probably be least thickly, and those in the level bottom lands most thickly, populated. Further than this it would be unsafe to try to state any general law. The least population which should be assigned to any habitable section within city limits is 20 per acre. The per acre population in any residence section can be expressed by the equation in which the average length of a city block; For a section where the blocks are 400 ft. by 200 ft., streets 66 ft. wide, lots 50 ft. by 100 ft., and the population residential (o = 6), 43560 X 400 X 200 X 6 50 X 100[400 × 200 + 66(400 + 200 + 66)] = 34 ±. For a tenement district, each building on a lot 50 ft. by 100 ft. and containing on an average 80 occupants, P would equal 453, which is about P for the Tenth Ward, New York City. A block with lots 25 feet by 80 feet and with 6 occupants each represents fairly well the most dense residence |