It may be asked what amount of supply, in gallons, would the water represent, in order that a comparison can be made between the supply of towns at the present time, and that of Rome at the time at which Frontinus wrote. The answer is not easy. As said before, we have not sufficient data for an accurate determination; still we may attempt an approximate estimate. It is easy to compute the section of water given by the 24,805 quinaries. This being just upon 120 square feet, we can form some notion of the vast quantity, if we picture to ourselves a stream twenty feet wide by six feet deep, constantly pouring into Rome at a fall six times as rapid as that of the river Thames. It has been computed by a French engineer that, together with one or two additional aqueducts which were added between the time of Trajan and Aurelian, the supply of water to Rome was 332,306,624 gallons _daily_. If we assume the population to have been a million souls—and it is scarcely probable that there were more—we find that the rate was 332 gallons per diem for each person. In our own day we consider 40 gallons sufficient, and many think this excessive, including the use of water in manufactures, &c. No wonder, with the facilities for drainage which the rapid Tiber afforded, and with this plentiful supply of really good water, the great city was rendered, in spite of overcrowding and an unhealthy situation as regards its neighbourhood, comparatively free from those epidemics which are so fatal in all densely-populated towns. One part of the system which Frontinus seems to have been the first to adopt was this: he had 247 _castella_, or main reservoirs, from which the pipes were carried to the smaller cisterns and fountains, and he so arranged the supply of water to these that the several aqueducts could be interchanged. A net-work was established, so that when one channel, from whatever cause, failed, whether from requiring to be cleansed or repaired, another could supply its place. But this was not all. He tells us, cap. 87, that many of the parts of the city which were dependent on one source for their supply, were, by careful distribution of the water, able to avail themselves of a second in case of need; and he gives as an instance that the Marcian was made available in the Aventine, by being carried across from the Cœlian[160]. He also speaks of the improvements which were made in keeping the water pure. Hitherto a very slight shower had made many of the streams muddy, although at their source they were pure[161].