Return to Temple of Juturna

Aqua Virgo

"What shall I say of streams suspended on airy arches, where scarce the Rainbow-Goddess could raise her showery waters? You might rather call them mountains grown up to the sky: such a structure Greece would praise, as giant-wrought. Rivers diverted are lost sight of within thy walls: the lofty baths consume whole lakes."

Rutilius, A Voyage Home to Gaul (I.97-103)

Completed by Marcus Agrippa in 19 BC at his own expense, the Aqua Virgo was built to supply water to the Campus Martius (as well as to other marginally served regions, including the Transtiber on the other side of the river, where it crossed on the Pons Agrippae). Although its source was located to the east of Rome on the estate of Licinius Lucullus, the Virgo entered Rome from the north, the only aqueduct to do so. This wide circuitous route around the city walls, especially when its end-point is almost in a direct line with its source, avoided populated areas in the suburbium but also suggests that Agrippa, in spite of his power and the presumed support of Augustus (after whom he named the aqueduct, Dio, Roman History, LIV.11.7), may have had trouble securing private property on which to build and was obliged to route his aqueduct across public land (cf. Frontinus, who says that "most of the conduits had been laid out through the property of private persons," The Aqueducts of Rome, CXXIV.4). The sixth aqueduct to be built, it began at the eighth milestone outside the city along the Via Collatina (a minor road on the way to Collatia, the home of Lucretia, who had been raped by Tarquinius, Livy, History of Rome, I.57-58) but extended along a circuitous route for 14,105 paces or 21.5 kilometers (a bit more than thirteen miles), measuring a passus as two marching steps or five feet (Frontinus, X.5; Pliny, Natural History, XXXI.420).

The channel was in a closed conduit (specus) underground until it emerged from the Pincian Hill just below the Horti Luculliani, the gardens of Lucullus above the modern Piazza di Spagna (Frontinus, XIX.2). It then continued west around the Quirinal Hill for 700 paces on arches to the Campus Martius, where almost two-thirds of its water was directed (LXXXIV.2), terminating at the Saepta Julia (XXII.2; Ovid has it end near the Temple of Juturna), the old voting precinct reconstructed and dedicated by Agrippa in 26 BC (Dio, LIII.23.1). From there, at the castellum (a reservoir or distribution tank), water was routed through underground pipes. The aqueduct supplied the Thermae Agrippae, the first of the great Roman baths; the Stagnum immediately to the west, an artificial lake where Nero once set up brothels crowded with noblewomen and, on the opposite bank, prostitutes obscenely gesturing, while he floated on a barge in the middle (Tacitus, Annals, XV.37); and the Euripus, an ornamental canal fed directly by the aqueduct that extended all the way to the Tiber (20% of its water, in fact, was just for that purpose). Used for swimming (and possibly as a plunge pool for Agrippa's baths), the cold water of the channel was a source of repeated comment. In a park between it and the lake, Agrippa dedicated a statue of a fallen lion by Lysippus (Strabo, Geography, XIII.1.19).

An underground channel was less susceptible to damage and extremes of heat and cold (Frontinus, CXXI.1-3), and the water of the Virgo was celebrated for its purity (in spite of not having a settling tank, piscina, to slow the current and allow sediment to collect at the bottom, XXII.1; rather, a series of zigzags in the course slowed the water enough to precipitate sand and gravel). Augustus recommended it to those complaining about a scarcity of wine (Dio, LIV.11.7) and Pliny contends that the soft water was considered particularly fine for bathing, being cool to the touch. Statius, too, remarks on the Virgin of the swimming pool (Silvae, I.5.26). Martial muses about the Campus and its Virgin water, so brilliantly clear that one scarcely suspected it even was there, with only the polished marble over which it flowed to be seen (Epigrams, V.20, VI.42). The cold water also refreshed the athlete after exercise (XIV.163) and dampened desire if need be (XI.47). Seneca, too, comments on the temperature, ruefully remarking that, whereas he now enjoys a warm bath, he used to celebrate the new year by taking a cold plunge in the Virgo (Epistles, LXXXVII). Ovid is less austere and reminisces in exile on the pleasures of Rome: its porticoes (a favorite spot to meet women, cf. Tristia, III.285ff), the view from the grassy Campus, and the Virgineus liquor (Ex Ponto, I.8.38). The phrase is poetic and has been translated "the flowing streams of Virgo" but also "the damp Virgin Conduit," which causes one to wonder at the erotic ambiguity.

The Aqua Virgo was the third lowest aqueduct in terms of elevation (where the open flow of water is distributed through closed pipes, XVIII.7) and was designed to serve the Campus Martius, which itself was on a flood plain at a bend in the Tiber. The length of the aqueduct and its low terminus required that the gradient of the channel be excessively flat. Vitruvius (On Architecture, VIII.6.1) specifies that the fall of an aqueduct should not be less than one-half foot in every one hundred, a minimum slope of 0.5%. Writing almost a century earlier, Pliny (XXI.57) indicates that water actually can flow on a decline as little as one sicilicus (1/4 inch) in one-hundred feet, a slope of 0.02%, the figure given by Hodge for the Aqua Virgo. Rodgers states that it had a fall of 0.025%, a decline of only 25 millimeters every 100 meters (approximately 1 inch every 325 feet) and the lowest fall of all the aqueducts. Ashby calculates a fall of 1 in 3670 (or 0.027%) and Taylor, a fall in elevation of 3.6 meters over nineteen kilometers, a slope even less than Pliny's theoretical limit. Ashby provides the actual elevations of both the spring, which is some 24 meters above sea level, and the emergence of the conduit at the Spanish Steps, 18.37 meters—a drop of approximately 0.026%.

Any change in either the channel's length or elevation at its origin or terminus, of course, would alter the calculation. Too, no aqueduct was built on a constant gradient; some stretches would be steeper or more gentle than others. Water passing through a long tunnel, for example, which would be difficult to clean, might have a steeper gradient, and a faster flow of water to prevent sediment from settling. Typically, the gradient of a Roman aqueduct is between 0.15 and 0.30% (1.5 to 3.0 meters per kilometer). Other than topographical considerations, it may have been the purity of its water that accounts for the gentle decline of the Aqua Virgo, which was less affected by the calcareous deposits that afflicted other aqueducts.

Julius Frontinus, appointed curator aquarum by Nerva in AD 97, is the primary source for a discussion of Roman aqueducts and the Aqua Virgo. In his treatise De Aquaeductu Urbis Romae, he measures water in terms of the quinaria, the minimum value of which has been calculated to be a bit less than one-half liter per second. Water supplied by the Aqua Virgo was 2,504 quinariae (LXX.3). This is approximately 1.2 cubic meters per second or almost 104,000 cubic meters a day, all but two hundred quinariae, which were used outside the city, entering Rome (LXX.3, LXXXIV.1). This made Virgo the third largest aqueduct in terms of capacity, delivering an enormous volume of water to relatively few sites. Indeed, it had only eighteen castella or distribution tanks within Rome (LXXXIV.2), far fewer than would be expected and each receiving a far greater volume of water than tanks for any other aqueduct.

As to the name, Aqua Virgo, Frontinus relates that "this same Agrippa developed another source on the property of Lucullus and brought Virgo into Rome. The date on which this water first came forth in the City is recorded as the 9th of June. It was called Virgo, because a young girl indicated certain water veins to the soldiers who were hunting for water, and the diggers who were to pursue them summoned up an enormous quantity of water. A painting which represents this origin is displayed in a small shrine set up near the source" (I.10.1-4). Pliny, writing at about the same time, has the water trying not to co-mingle. Near the source of the spring was the stream of Hercules and "because the Virgin Water runs away from this it was so named" (XXXI.42).

Cassiodorus offered a later explanation. "Purest and most delightful of all streams glides along the Aqua Virgo, so named because no defilement ever stains it. For while all the others, after heavy rain show some contaminating mixture of earth, this alone by its ever pure stream would cheat us into believing that the sky was always blue above us. Ah! how express these things in words worthy of them?" (Variae, VII.6). More prosaically, it may be that, entering Rome on June 9, which was the feast day of Vesta (the Vestalia), the aqueduct was named for the Vestal Virgins. Or the date may be coincidental.

When Frontinus measured the Aqua Virgo at the seventh milestone, he found 2,504 quinariae, 1,852 more than officially recorded (652 quinariae, LXX.1-3). Typically, such a discrepancy would be due to fraudulent diversion or illegal taps on the conduit, but none were discovered (which is understandable, given that it bypassed most developed areas) and he reported that the aqueduct delivered its full volume of water. The difference, however, is striking and may reflect inaccurate calculations in the record, which is the explanation Frontinus offers for the discrepancy (LXXIV.2). Or there may have been difficulty in taking them, as the Virgo had no piscina or settling tank (XXII.1), and the water that filled the collection basin entered too slowly for a reliable measurement to be taken at the source. Instead, Frontinus was obliged to make his measurement near the seventh milestone, where the water had sufficient velocity (LXX.2-3).

Every 240 Roman feet, boundary stones (cippi) marked the course of the aqueduct, beginning at the Pincian Hill. Half a dozen have been found, in two different series inscribed both by Tiberius and Claudius. In AD 51/52, to commemorate the conquest of Britain, Claudius commissioned an arch where the Aqua Virgo crossed the Via Lata (Broad Street), now the Via del Corso. Although the Arch of Claudius had been destroyed by the eighth century, a section of the Aqua Virgo was discovered in 1887 at the Via del Nazareno, the travertine blocks cut in the heavily rusticated style favored by the antiquarian taste of the emperor.

Martial (IV.18) relates the poignant story of a boy who was killed by a falling piece of ice from the dripping arch, the weapon melting in the warm wound it had created. William Hay (1755) concludes the epigram with a particularly evocative line, "From cruel fortune can we more endure? / If waters stab, where can we be secure?"

In AD 537/538, the same year that Cassiodorus published his official correspondence, the Goths besieged Rome and sought to infiltrate the city through the dry conduit of the Aqua Virgo. It is a harrowing account.

"And the Goths not long after this wished to strike a blow at the fortifications of Rome. And first they sent some men by night into one of the aqueducts, from which they themselves had taken out the water at the beginning of this war. And with lamps and torches in their hands they explored the entrance into the city by this way. Now it happened that not far from the small Pincian Gate an arch of this aqueduct had a sort of crevice in it, and one of the guards saw the light through this and told his companions; but they said that he had seen a wolf passing by his post. For at that point it so happened that the structure of the aqueduct did not rise high above the ground, and they thought that the guard had imagined the wolf's eyes to be fire. So those barbarians who explored the aqueduct, upon reaching the middle of the city, where there was an upward passage built in olden times leading to the palace itself, came upon some masonry there which allowed them neither to advance beyond that point nor to use the ascent at all. This masonry had been put in by Belisarius as an act of precaution at the beginning of this siege, as has been set forth by me in the preceding narrative. So they decided first to remove one small stone from the wall and then to go back immediately, and when they returned to Vittigis, they displayed the stone and reported the whole situation. And while he was considering his scheme with the best of the Goths, the Romans who were on guard at the Pincian Gate recalled among themselves on the following day the suspicion of the wolf. But when the story was passed around and came to Belisarius, the general did not treat the matter carelessly, but immediately sent some of the notable men in the army, together with the guardsman Diogenes, down into the aqueduct and bade them investigate everything with all speed. And they found all along the aqueduct the lamps of the enemy and the ashes which had dropped from their torches, and after observing the masonry where the stone had been taken out by the Goths, they reported to Belisarius. For this reason he personally kept the aqueduct under close guard; and the Goths, perceiving it, desisted from this attempt."

(Procopius, The Gothic War, VI.9)

This was the first major disruption of the aqueduct. But the system continued to function, even if the public baths did not. There were further restorations by the papacy in the late eighth century, although the arcade beyond the Trevi fell into disuse, and again in 1453, when the simple medieval fountain into which the water flowed was altered, its three basins, each fed by a spout, being replaced by a single rectangular one. A dedicatory inscription, crowned by a heraldic shield with the papal tiara and crossed keys, proclaimed the pope's munificence. When water failed to reach the fountain, itself, there was a restoration of the entire aqueduct back to its original spring, which was completed in 1570. The magnificent baroque fountain above dates to 1762.

In the sixteenth century, a tank located where the aqueduct emerged from the Pincian Hill (near the Spanish Steps) distributed water to the Campus Martius. It was these conduits that gave the Via Condotti its name.

"Meanwhile Agrippa beautified the city at his own expense."

Dio, Roman History (XXVII.1)

This detail from the Model of Rome makes the topography a bit more clear. In the center foreground is the Pantheon, which Agrippa had built in 27 BC, during his third consulship (Dio, LIII.27.2). The Saepta Julia is immediately on the left and the Stagnum on the right. Directly behind the Pantheon are the Baths of Agrippa and behind them four temples from the Republic, the first of which is thought to be the Temple of Juturna. Abutting this sacred precinct is the portico of the Theater of Pompey, which can be seen in the upper right. The Theater of Balbus is at the top and, at the bottom left, the termination of the Aqua Virgo at the Saepta. Below the Stagnum is a park (nemus) where Nero celebrated his revels, "all the adjacent grove and surrounding buildings resounded with song, and shone brilliantly with lights" (Tacitus, Annals, XV.37). The late nineteenth-century Forma Urbis Romae by Lanciani is a reminder that much reconstruction is conjecture.

When Martial cataloged the pleasures of Rome (V.20), he mentions outings, conversation, and reading, and then the Campus Maximus—its shady porticoes, the Virgin water, and warm baths. It is places such as these that should be one's constant resort and daily occupation. But, as befitting a Field of Mars, the Campus Martius also provided exercise ground, hot baths, and the cold waters of the Virgo for the athlete (indeed, gelidissima Virgo, "very cold," Ovid, Art of Love, III.385). To Dio, the baths were the Laconian sudatorium or Spartan steam bath "because the Lacedaemonians had a greater reputation at that time than anybody else for stripping and exercising after anointing themselves with oil" (LIII.27.1; cf. Martial, VI.42, who speaks of Lacedemonian customs, the dry heat, and a plunge into the raw Virgo). Here Agrippa has provided the facilities for young men to exercise, take the dry vapors, swim in the cold water of the Virgo and, perhaps like its namesake, not have much time for women.

But there also was leisure and luxury. The first of the shady promenades in the Campus was the Portico of Pompey. Built behind the scaena of his theater, it became a favorite place for walks and, on a rainy day, offered shelter for the audience.

The Saepta Julia was a venue for gladiatorial games (e.g., Suetonius, Life of Augustus, XLIII.1; Dio, LV.8.5), and its surrounding arcade, the setting of luxury boutiques, where Martial envies (and satirizes) those who can shop there—such as Mamurra, who spent the day at the Saepta, "where golden Rome ostentatiously displays her riches," browsing its offerings of slave boys, wrapped table tops (of citronwood, no doubt, X.80, XII.66), ivory ornaments off the top shelf, a tortiseshell couch, Corinthian bronzes (which a true connoisseur could discern simply by the smell), statues (but none by Polyclitus), vases of rock crystal and murrine, emeralds and pearls, sardonyxes and jaspers. But, for all his browsing, he bought only two cups for a copper as" (XI.59).

Seneca did not approve of such excess: "Whenever you see the forum with its thronging multitude, and the polling-places [Saepta] filled with all the gathered concourse, and the great Circus where the largest part of the populace displays itself, you may be sure that just as many vices are gathered there as men" (On Anger, II.8.1).

This detail is from A Roman Art Lover (1870) by Lawrence Alma-Tadema. It is one of several pictures depicting Romans inspecting works of art, including A Sculpture Gallery in Rome at the Time of Augustus (1867), his first, which paired with The Collector of Pictures at the Time of Augustus. Other pairs were A Picture Gallery (1873) and A Sculptor Garden (1875), and A Sculpture Gallery in Rome and A Picture Gallery (1874), as well as A Roman Lover of Art (1868) and A Roman Art Lover (1868, 1870). The subject was a popular one: by depicting buyers as connoisseurs of art, Alma-Tadema connected them to their ancient counterparts and glorified the purchase of pictures.


The Fountain of the Tortoises (Fontana delle Tartarughe) is in the Piazza Mattei, which also is fed by the Acqua Vergine. In this detail, one of four bronze boys perched on a dolphin holds a turtle up to drink from the basin. The turtles, perhaps by Bernini, were added as part of a restoration in 1658. More recent work in 2006 to remove limestone deposits has made the fountain even more charming.


The Aqua Virgo (Acqua Vergine in Italian) is the only ancient aqueduct still functioning in Rome, although water from the original course is reserved for fountains in the Campus Martius. Wending its way into the city, the aqueduct now terminates at the Fontana di Trevi (above) at the foot of the Quirinal Hill. Two high reliefs depict its history, the one on the right above showing the young girl pointing out the spring to the Roman soldiers; the one on the left, Agrippa supervising construction. Salubrity and Abundance stand in niches, flanking the figure of Ocean driving a team of hippocamps—one restless, the other calm—and Tritons.

Take a drink, then, in memory of Agrippa, who as aedile in 33 BC also had built the Aqua Julia (which, together with the Aqua Virgo that complemented it, double Rome's supply of water) and effectively became Rome's first curator aquarum (Frontinus, IX.1-2, XCVIII.1), an office perpetuated by Augustus the year after his death in 11 BC (XCIX.2)—and to the little maid who first pointed out the spring to his legionaries.

"When they are brought home to the walls of the city a reservoir (castellum) is built, with a triple cistern attached to it to receive the water. In the reservoir are three pipes of equal sizes, and so connected that when the water overflows at the extremities, it is discharged into the middle one, in which are placed pipes for the supply of the fountains, in the second those for the supply of the baths, thus affording a yearly revenue to the people; in the third, those for the supply of private houses. This is to be so managed that the water for public use may never be deficient, for that cannot be diverted if the mains from the heads are rightly constructed. I have made this division in order that the rent which is collected from private individuals who are supplied with water, may be applied by collectors to the maintenance of the aqueduct."

Vitruvius, On Architecture (VI.1-2)

Frontinus speaks of a similar distribution (LXXVIII-LXXXVI). Water was furnished in the name of Caesar (to supply the imperial palace and other official buildings), to private individuals as a favor of the emperor, and for public use, which included military camps, public works (such as baths, circuses, and markets), fountains, and open basins.

References: Frontinus: De Aquaeductu Urbis Romae (2004) edited by R. H. Rodgers (now the standard text, having replaced the 1925 translation in the Loeb Classical Library); The Letters of Cassiodorus (1886) translated by Thomas Hodgkin; Ovid: The Poems of Exile (2005) translated by Peter Green; Procopius: History of the Wars (1919) translated by H. B. Dewing (Loeb Classical Library); Minor Latin Poets (Rutilius Namatianus: De Reditu Suo) (1934) translated by J. Wight Duff and Arnold M. Duff (Loeb Classical Library).

"Restoring the Ancient Water Supply System in Renaissance Rome: The Popes, the Civic Administration, and the Acqua Vergine" (2005) by David Karmon, The Waters of Rome, 3, 1-13; "Copia Aquarum: Frontinus' Measurements and the Perspective of Capacity" (1986) by R. H. Rodgers, Transactions of the American Philological Association, 116, 353-360; "The Aqua Virgo, Euripus and Pons Agrippae" (1979) by Robert B. Lloyd, American Journal of Archaeology, 83(2), 193-204; "Agrippa's Water Plan" (1982) by Harry B. Evans, American Journal of Archaeology, 86(3), 401-411, which is expanded upon in Water Distribution in Ancient Rome: The Evidence of Frontinus (1994) by Harry B. Evans (Rodgers' translation is included in its entirety); Public Needs and Private Pleasures: Water Distribution, the Tiber River and the Urban Development of Ancient Rome (2000) by Rabun Taylor; Empire of Pleasures: Luxury and Indulgence in the Roman World (2000) by Andrew Dalby; Roman Aqueducts and Water Supply (2002) by A. Trevor Hodge; Frontinus' Legacy: Essays on Frontinus' de aquis urbis Romae (2001) by Deane R. Blackman and A. Trevor Hodge; The Aqueducts of Ancient Rome (1935) by Thomas Ashby; The Building of the Roman Aqueducts (1973) by Esther Boise Van Deman.

See also the Arch of Claudius.

 Return to Top of Page