30 THE POPULAR SCIENCE MONTHLY. 



and were called after their builders or projectors. The waters of the 

 Marcian, the most prized for their purity and coldness, were collected 

 from several springs. For the Anio Novus, which was unfailing as 

 well as the most abundant of the aqueducts, the river Anio was 

 arrested near its source by three gigantic walls at different levels, 

 and formed into as many lakes, one below the other. Over these 

 walls the waste-water fell in magnificent cascades, one of them 

 over 150 feet high. The object of the lakes was to clarify the 

 water; for the Anio, though usually a limpid stream, is liable to 

 become muddy after a heavy rain. The sources of the Anio Novus 

 and the Aqua Claudia are over 2,000 feet above the level of the city, 

 and those of the Marcia and Anio Vetus are not very much lower. 

 Descending from such a height and for distances varying in direct 

 lines from 30 to 43 miles, the water would naturally acquire great 

 velocity and tremendous force, which it was necessary to dimin- 

 ish, and that was done by making numerous angles in the con- 

 duits. The angles were made generally at every half-mile, and were 

 points at whi$h reservoirs (castella), or filtering-places {piscinoi), or 

 both, with accompanying air - shafts, were built. These were sur- 

 mounted by small towers. As an additional means of breaking the 

 force of the water, the bottoms of the conduits were given a succes- 

 sion of short undulations. The conduits, reservoirs, and filtering- 

 places, were lined with a cement called ojms signinum, which is so 

 compact that it will resist a hard tool. The art of making it has been 

 lost. The conduits, always covered, were carried on arcades only 

 where it was necessary to cross a valley or a plain above its level ; for 

 the rest of their way they ran in places upon the surface of the 

 ground, but mostly below it. Thus of the 58 miles of the Anio 

 Novus, 49 were underground. No two aqueducts were on the same 

 level, and so, where their courses converged, it was both possible 

 and convenient to carry one conduit upon another, because it was 

 forbidden by law to erect a building within a certain number of 

 feet on either side of an aqueduct; hence Ave find the Aquae Mar- 

 cia, Tepula, and Julia, carried from their ^Doint of convergence one 

 above the other on one arcade, and the Aqua Claudia and Anio 

 Novus on another. Each of the conduits was differently shaped, 

 some having arched, others angular roofs. Besides the small reser- 

 voirs referred to as occurring at the angles of the conduits, there 

 were larger ones at longer intervals. The ruins of one of these, 

 belonging to the Aqua Marcia, are still to be seen near Carciano. It 

 is a huge subterranean chamber divided by an arcade in the middle. 

 Between five and seven miles from Rome were the great filtering- 

 places to which most of the aqueducts converged. The waters, how- 

 ever, were not mingled, for each aqueduct had its separate chambers, 

 thougli it was always within the power of the attendants (aquarii) 

 to turn the water from one aqueduct into another at will. Of these 



