Engineering

Rome's Water Machine

Eleven aqueducts, 800 kilometres of channels, a slope of centimetres per kilometre — and almost none of it on those famous arches.

In AD 97 the emperor Nerva handed Sextus Julius Frontinus — soldier, former governor of Britain, consummate administrator — the least glamorous great job in Rome: curator aquarum, commissioner of the waters. Frontinus took it seriously enough to write a book, De aquaeductu, a working manual of the city's water supply that survives to this day. In it he allowed himself one boast that has echoed for nineteen centuries: with such an array of indispensable structures carrying so much water, he wrote, compare if you will the idle Pyramids, or the useless though famous works of the Greeks. It was the engineer's creed, stated once and for all — usefulness as a form of splendour.

The system he inherited had been growing for four centuries. The first aqueduct, the Aqua Appia, was commissioned in 312 BC by the censor Appius Claudius Caecus — the same magistrate, in the same year, who began the Appian Way — and it crept a little over 16 kilometres into the city, almost every metre of it below ground. Ten more followed across half a millennium, ending with the Aqua Alexandrina in AD 226. At its height the network comprised eleven aqueducts and roughly 800 kilometres of channels, delivering — by modern estimates that vary widely — hundreds of thousands of cubic metres of water a day, a more generous ration per head than many modern cities manage. And here is the detail that surprises: of those 800 kilometres, only a small fraction ran on the arches that every postcard shows. The overwhelming majority was buried.

A fall measured in centimetres

An aqueduct was not a pressurised pipe but an artificial stream: a masonry channel, lined with waterproof cement, descending continuously from spring to city by gravity alone. That made the gradient everything. Drop too steeply and the current tears at the lining; too gently and the water stagnates. Roman surveyors — working with the chorobates, a long levelling bench with a water trough, and simple sighting instruments — held slopes that modern engineers still admire. The aqueduct serving Nîmes in southern Gaul, the one that strides across the Gardon valley on the Pont du Gard, falls only about 12 metres over its entire 50-kilometre course: an average of roughly 25 centimetres per kilometre, a slope invisible to the eye, held across hills and valleys with hand tools and patience. Where hills stood in the way they tunnelled; where valleys ran too deep to bridge they used inverted siphons — batteries of lead pipes plunging down one slope and climbing the other, as at Lyon.

Burying the channel was not timidity but sound engineering. Underground water stayed cool and free of algae; it was harder for thieves to tap or enemies to cut; and a trench cost far less than an arcade. The famous arches were reserved for the final approach, where the channel had to hold its height across the flat Campagna — which is why they march so theatrically towards the city walls. The grandest pair, the Aqua Claudia and the Anio Novus, completed under the emperor Claudius in AD 52, entered Rome stacked one channel above the other on a single arcade, crossing at what is now the Porta Maggiore.

Inside the walls the water fanned out through distribution tanks called castella to the baths, the street fountains and, for a licensed fee, the houses of the rich. The flow never stopped: Rome's taps had no off position, and the perpetual overflow sluiced the streets and flushed the sewers, including the great Cloaca Maxima. Frontinus's audit found the system leaking money as well as water — riddled with illegal taps and pipes wider than their licences — and he pursued the fraud with a general's thoroughness. One aqueduct outlasted them all: the Aqua Virgo, built by Agrippa in 19 BC to supply his baths, still flows today, feeding the Trevi Fountain.

The myth of the poisoned empire

A popular theory holds that all those lead pipes slowly poisoned Rome into decline. Offer it at a quiz, but flag it as disputed: Rome's hard water quickly lined the pipes with a crust of limescale, the water rarely lingered in the lead, and most of its journey ran through stone and concrete anyway. Romans certainly absorbed lead from other sources — sweetened wine above all — but few historians accept that plumbing felled the empire. The aqueducts' real ending came by war: when the Ostrogoths besieged Rome in AD 537 they cut the lines, and the city that had watered perhaps a million people shrank towards the bend of the Tiber, drinking river water again until Renaissance popes restored the ancient channels. Frontinus, it turns out, was underselling. The Pyramids merely outlasted their purpose; one of his aqueducts is still doing its job.

Quiz nuggets

  • Rome's first aqueduct, the Aqua Appia of 312 BC, was built by Appius Claudius Caecus, who began the Appian Way the same year; the eleventh and last was the Aqua Alexandrina of AD 226.
  • The network totalled roughly 800 km of channels, the great majority underground — only a small fraction ran on arches.
  • The Nîmes aqueduct, which crosses the Pont du Gard, falls about 12 metres in 50 km — roughly 25 cm per kilometre.
  • Frontinus became curator aquarum in AD 97 and wrote De aquaeductu, favourably comparing his waterworks with the "idle Pyramids".
  • The Aqua Virgo, built by Agrippa in 19 BC, still supplies the Trevi Fountain.

Written from public sources and not individually checked — worth confirming before you stake a pint on it.