Map of the aqueducts feeding western Rome from 800 BC to 1800 CE, from Waters of Rome.
Rome and its empire were founded upon the back of an extensive and innovative aqueduct system. These aqueducts enabled urbanization and supported concentrated populations and massive agricultural exploitation in the Roman world. For Rome, the network of aqueducts feeding the city is fairly well understood. Sextus Julius Frontinus, appointed curator aquarum in 95 CE, wrote a technical treatise describing the entire extant water supply system. Though several interpretive issues challenge our understanding of his De aquaeductu, his work is nevertheless invaluable to historians and archaeologists alike. Topographical and archaeological evidence supplement our understanding of the aqueducts of Rome. In the following sections, we focus on two of the aqueducts most closely connected to milling in Rome.
Photograph of the interior of Aqua Traiana running underneath the American Academy, from Wilson's site: http://users.ox.ac.uk/~corp0057/aquatraiana.html
During the reign of Trajan, construction likely began on the Aqua Traiana in 107/8 CE, and the aqueduct was officially introduced in 109 CE.[i] The Aqua Traiana is one of only two aqueducts, the other being the Alsietina, that are brought in from springs to the west of Rome. Rabun Taylor calls the Aqua Traiana the “most enigmatic of all Rome’s aqueducts,” and the exact impetus behind its construction has been a major source of continuing contention between scholars.[ii] Frontinus wrote his record about a decade prior to the construction of the Aqua Traiana, so little contemporary evidence is forthcoming on this subject. De Kleijn, amongst others, has proposed that Trajan built the aqueduct to supply his imperial baths on the Esquiline.[iii] A lead pipe stamped with the marking “Aqua Traiana” was found on the Esquiline, suggesting that the Aqua Traiana did, in fact, feed the baths there.[iv] Aqueducts are commonly associated with the building of imperial baths, but the conclusion that aqueducts were specifically planned and built primarily for the supply of baths may be a bit constrictive. Although the imperial baths may have been partial factors in Trajan’s intended use of the aqueduct, this explanation does not account for the full impact that the Aqua Traiana had in augmenting the general water supply for all of Rome, particularly in the region of the Transtiberim.
Until the construction of the Aqua Alsietina in 2 CE, which directly supplemented the water supply west of the Tiber, the Transtiberim region had relied primarily on natural springs and laci, or fountains, for water.[v] Frontinus also records that Agrippa extended the Aqua Virgo and Appia that came from the east into the Transtiberim, but the water output of these two extensions is hard to determine.[vi] These sources of water would have been sufficient to supply the lower levels of population density found in the Transtiberim in the early periods, but the Aqua Traiana’s introduction in 109 CE coincides with a large population surge in the early second century CE.[vii] As mentioned briefly above, the Aqua Traiana derives its source from previously untapped springs to the west of Rome, and its major distribution center or castellum currently lies under the casino of the Villa Spada, which indicates that the castellum would have been just inside the edge of the city walls.[viii] The location of this distribution center and the Traiana’s entry point to the north of the Transtiberim suggests that one of the aqueduct’s primary functions was the supply of water to the Transtiberim itself, since the imperial administration did not choose merely to augment pre-existing aqueducts coming from east of the Tiber with additional sources.[ix] Although the Aqua Traiana was primarily dispersed within the Transtiberim, the Traiana was eventually extended into all fourteen regions of the city and was known for its relatively high quality of drinking water.
Photo of the Villa Spada's casino and the accompanying 1639 inscription that led early excavators to believe that the Aqua Alsietina ran underneath the building. (From http://www.romeartlover.it/Vasi13.htm#Villa%20Spada)
In addition to supplying the imperial baths and augmenting the available water supply of the Transtiberim substantially, the Aqua Traiana has been proven to supply the water mills on the Janiculum, which are the major focus of our investigations. Historically, the Aqua Alsietina was thought to be the water source for the mills, since it terminates near the Janiculum mills. In 1912-1913, however, van Buren and Stevens identified the Aqua Traiana as the actual source for the Janiculum mills by dating pottery found in the construction trench of the aqueduct channel underlying the mills.[x] This pottery is dated to the first and second centuries CE, which corresponds to the 109 CE date for the Aqua Trainan but postdates the construction of the Alsietina in 2 CE.[xi] The Wilson excavations begun in 1998 confirmed van Buren and Stevens’ conclusions and proved that the Aqua Traiana fed the mills. Wilson’s team also uncovered the blocked intake of the south millrace and identified the intake’s location just inside the western edge of the mill where it diverged from the Aqua Traiana.[xii] Because of the evidence from the construction trench and Wilson’s excavations, most scholars now are in accordance that the mills derived their water from the Aqua Traiana.
Although this aqueduct was not specifically built for the purpose of feeding this mill, the Aqua Traiana’s connection with the Janiculum mills suggests a possible imperially-sanctioned industrial use of water, since it is unlikely that private businessmen would have received permission to tap the aqueduct without imperial consent. To see further information concerning the possible imperial oversight of the Janiculum mills, please see The State and the Mill.
[i] CIL VI 1260
[iii] De Kleijn.2001.27-28
[v] Taylor.1995.78-80, De Kleijn.2001.23
[vi] For a good diagram of these aqueducts, please see the “Waters of Rome” website. For further information about calculations of water volume of aqueducts, please see Taylor.1997 "Torrent or Trickle? The Aqua Alsietina, the Naumachia Augusti, and the Transtiberim"
[vii] Evans 1991.25-26
[viii] Taylor.1995.97, For information on the Villa Spada please see: “Villa Spada – A historical note” – Gearoid O’Broin
The Aqua Marcia
Standing remains of the Aqua Marcia. (From https://eee.uci.edu/clients/bjbecker/SpinningWeb/week2c.html)
The Aqua Marcia served the southernmost regions of Rome, including the Baths of Caracalla. Unlike what has been thought of for the Aqua Traiana, originally the Marcia was not built to support a large spectacle building, but rather to bring drinking water to the rapidly expanding population of Rome. At the behest of the praetor inter cives Q. Marcius Rex, the city of Rome funded the construction of the aqueduct, which lasted from 145 to 144 BCE. The construction of the aqueduct was so important that Rex’s term was extended one year to allow him to complete the aqueduct.
The original purpose of the Aqua Marcia was to provide drinking water to the area of the Palatine, along with the Aqua Julia. Several branches were added to the Marcia, both soon and long after the aqueduct’s original constructions. Eventually, the Marcia would stretch to the Aventine, Palatine, Capitoline, and Baths of Caracalla. The water of the Marcia was strictly controlled and its distribution stretched throughout the city. Frontinus 87.3-4 lists 51 castellae within the city that regulated the flow and distribution of the aqueduct.
Like Augustus, Nero, and Trajan before him, Caracalla designed a new adjustment to the Aqua Marcia specifically intended to support a new entertainment complex. In 212 CE, Caracalla began a new extension of the Marcia and a reservoir primarily for his new bath complex, even adding a new source to the aqueduct. This extension of the aqueduct was named Aqua Antoninina Iovia. Brick stamps found in the baths prove that construction on the bath and aqueduct projects began at the same time and were built contemporaneously. The baths were dedicated in 216 CE, suggesting that the aqueduct was most likely operational at that time. Despite the massive expenditure of resources, including water, on the Baths, Caracalla’s architects did include at least one “green” feature. The drained wastewater from the baths was used to power the water mills found in the southwestern exedra’s substructures, which suggests that the mills were an opportunistic addition to the bath complex that would make use of precious water resources already being funneled through the baths. This opportunistic use of entertainment-related wastewater for an industrial purpose supports the argument that the imperial administration did consider the conservation of resources and intended water from aqueducts to serve a variety of uses, not merely entertainment.
Map of the course of the Aqua Marcia outside Rome. (From https://eee.uci.edu/clients/bjbecker/SpinningWeb/week2c.html)
The Aqua Marcia ran 91.32 km in length from its source at an elevated portion of the River Anio to its endpoints within the city of Rome. After 212 CE, the aqueduct also drew from the Fons Antonininanus, whose exact location is unknown. Of these 91.32 km, 80.28 km are underground leading up to a settling tank on the city side of the seventh milestone on the Via Latina. From that settling tank, the water runs 0.78 km above ground on substructures and 9.58 km on arches. The water next runs underground on the Viminal Hill to re-emerge at the Porta Viminalis. This path gives the water sufficient height to permit flow into any region of Rome. Caracalla’s branch, the Aqua Antonininiana Iovia, passed over the Via Appia on the second century CE Arch of Drusus just before the Porta San Sebastiano. Coarelli believes that the Arch of Drusus would have acted as a monumental entry point into the city, further monumentalizing Rome’s water supply.
 De Kleijn 2001: 15-16. De Kleijn suggests that the year 146 BC would have seen the return to Rome of a large number of veterans, and thus a large spike in Rome’s population. This year coincides with the conclusion of the Third Punic War, and is shortly after the completion of the Fourth Macedonian War and First Numantine War.
 De Kleijn 2001: 15-16.
 Evans 1994: 131. The Anio Novus and the Claudianus were also intended to bring water to this area, further proving the well-known fact that the Palatine was a center of Roman activity throughout the city’s history.
 For a visual overview of the aqueducts of Rome, please visit the Aquae Urbis Romae page in our online resources section.
 CIL VI. 1245, Coarelli 2007: 327, De Kleijn 2001: 18.
 Coarelli 2007: 326-327.
 Wikander 1991: 142.
 The only study of the Aqua Marcia’s course, based in large part on the descriptions of Frontinus, was conducted by De Kleijn 2001: 15-16. The following account of its path is largely based upon her work.Click to add text, images, and other content