At the city end of the aqueduct an enormous reservoir was constructed called a castenuen aquarum, where the water was eleared by pass ing through several chambers, and from which it was then distributed over the city. These •astella were sometimes, as in the case of the Claudia and the Alexandrina, at Rome, impor tant artistic structures. Here there were sepa rate purifying and storing compartments for each class of structures supplied ; in the Re publican period there were only three—public fountains, baths, and private houses. Rut under the Empire the subdivision became much more elaborate. Certain very large single buildings, such as baths, had separate reservoirs, or tanks. The water was carried into private or public Buildings by lead pipes through an official bronze joint stamped with its exact capacity, and serv ing as a meter. The conservation and regulation of the water supply, the exact allowance to indi viduals, corporations, and public buildings, was secured by a very careful administration of the water-works. This care was not only applied in Rome itself ; but was coextensive with the entire line of aqueduct as it was tapped at intervals and used by towns, settlements, and private owners for drinking and irrigation. To assist the administration, a strip of land thirty feet wide was reserved along the entire course, as government property, and marked by boun dary stones at intervals of two hundred and forty feet. The administration was under the care of the censors, and then of the qmcstors and tediles; but under Augustus the bureau was better organized, and put in charge of a Curator Aquarum, with his two assistants, his clerks, his consulting engineer, and his various classes of offieials and of artisans comprising a familia of slaves: ushers, lictors, and criers, as well as pipe-layers, pavers, masons, levelers, meas urers, inspectors, reservoir keepers, etc. As usual with Roman buildings, the aqueducts were built by contract, and the use of unskilled labor made their cost relatively small. The Appia is said to have cost $675,000. The repairing of the Appia and Anio Vetus, and building of the Marcia in we. 144-140, cost only about $850,000. l'iler the more lavish Empire the Claudia and the Anio Novus cost about $4,000.000, hut none of the others were as expensive as these.
Among the Roman aqueducts. those of Rome itself possess the greatest interest, because of their number, length, and boldness of design and execution. Two of them, in fact. are still in use, and water from the very source that supplied one of them (Marcia) is now delivered to the city through a modern wate•-works system. Not only are they in remarkable preservation, but, most happily for engineers and arclneologists alike, they are described in sonic detail by a Roman engineer who was water commissioner of Rome in A.D. 97, named Sextus Julius Fron thins, in his Two Books on the Water Supply of Romc. This work was first made available to English readers in 1899, through a translation by Mr. Clemens Herschel, an American hy draulic engineer, who gives not only the Latin text, but also a photographic reproduction of the oldest Latin MS. in existence, in the library of the Benedictine monastery at Monte Cassino.
Besides all this, the book in question contains several chapters of comment by the translator, both on the aqueducts and the water supply of Rome in general. Sir. Hersehel concludes that the capacity of the ancient Roman aqueducts has been greatly overrated, and that, instead of the 400.000,000 gallons a day given by some writers, based on Frontinus's calculations, "thirty-eight million gallons one day with another" is "a fair estimate at \VI11,11 to set the water supply within the NVIIIIS of ancient Rome in A.D. 97, though the total ranged, no doubt, some 20,000,000 gallons per day either side of that mark from time to time. This would make about thirty-eight gal lons per day per inhabitant, which is still a very large figure when use alone, not waste, is taken into account: and when, further, we con sider that by far the greater part of the people undoubtedly used only such water as was carried to their homes in jars on the heads of slaves and other women.'• Still, Frontinus describes nine aqueducts in use in his day, the main facts regarding which may be summarized as follows: (1) Aqua Appia, built by and named after the censor, .Appins Claudius, in we'. 312. 1ts springs were between the sixth and seventh milestones from Rome, and its course was about 11 miles long. All but 300 feet was underground. The exact size of its channel is uncertain, but is given by several authors as about 2.5 feet wide and 5 feet high in the clear.
(2) Anio Ictus, constructed B.C. 272-270 by M. Curios Dentatns and Fulvins Flaccus. All but 1100 feet was underground. Remains may he traced both near Tivoli and near the Porta Maggiore. Its water is taken from the river Anio, about the twentieth milestone, three miles beyond Tivoli, and its course, which is very cir cuitous, is about 43 miles long. About 3.7 feet wide and eight feet high inside, of heavy masonry of peperino stone, plastered on the inside.
(3) Aqua Marcia, named after the prtor, Quintus Marcius Rex, we. 144-140, had its source in springs between Tivoli and Subiaco, near the thirty-sixth milestone from Rome, was over 62 miles long, carried into the city 195 feet above sea-level, so as to reach the top of the Capitol. Near its head it is 3.7 feet wide and S.3 feet high, and further on it is 3 X 5.7 feet. This and the two preceding aqueducts were built of rough-hewn dimension stone, 18x1Sx42 inches. or more. while the later ones, except Clau dia, were of concrete and brick. The greater part of Marcia was underground, but there were some long stretches on arches—over seven miles—some of \Villa are still standing, and bear parts of two and three other aqueducts (Anio Vetus, Claudia, and Anio Novus ) above them. This is especially the case near Tivoli, where there are superb viaducts and bridges alternating with tunnels. There are about six miles of arcades near Rome.
( 4 I .4 qua Ti pula, B.c. 125. lead ing from springs on the slopes of the Monti Albani, had at first an independent channel, on the arcades of the :\larcia, 6 feet above it, or 201 feet above sea-level. It was 2.7 feet wide, by 3.3 feet high, and commenced not far from the eleventh mile stone.