The writer is unable to understand how this quantity of sewage will be produced, especially as the author states that, as far as the company is concerned, it is limited to the removal and disposal of the sewage, and is not required to provide for storm-water. He also states that: "Apart from that fact, however, the best system for a city like Monterrey, where rainfall for many months at a time is very scarce, is the strictly 'separate system'." The minimum velocities in the sewers, when running full, vary between 0.91 and 1.5 m. (from 3 to 5 ft.) per sec., and will be the same flowing half full. From the foregoing data it will be observed that: (1) The water supply is the only source from which sewage flow is anticipated; (2) The water supply is very liberally estimated at 200 liters (53 gal.) per capita daily; (3) For purposes of sewer design, the daily flow of sewage expected (all of which is derived from the water supply of 200 liters per capita) is estimated at 380 liters per capita, with a maximum rate of flow of 200% (or at the rate of 760 liters per capita), and with this quantity the sewers are designed to flow only half full; (4) The gradients are such that a velocity of from 3 to 5 ft. (0.91 to 1.5 m.) per sec. will be secured in the sewers flowing half full with the above quantity of flow per capita. The writer does not agree with this method of computation, as he feels sure that it will give sewers which are too large, with grades too steep for the best obtainable results. His experience, extending over more than twenty years in sewer design and hydraulic work, convinces him that the method pursued is wrong in principle. The principles involved in sewer design are first of all hydraulic. The quantity of flow, in the nature of things, cannot be forecasted accurately; success depends on getting the nearest possible approximation to average conditions. If 200 liters per capita per day is a liberal allowance, and 40,000,000 liters per day is a liberal expectation at this rate for double the present population, and the sewers are designed to flow half full only, why should this again be doubled? The design of a sewer system for a city such as Monterrey is, in fact, a very difficult problem, especially as the quantity of sewage will be very limited, flush-water will have to be used in considerable quantities, and water in that part of the world is precious at all times and often scarce. Under these circumstances, the size or shape of the pipes selected for the lateral sewers, should have been such as would more nearly agree with the requirements than does the 8-in. circular. A. P. Folwell, M. Am. Soc. C. E., writing of the 8-in. circular size, states:[10] [10] "Sewerage," by A. P. Folwell, M. Am, Soc. C. E.