The city in the clouds

What the site actually was, how it was built, and who came looking for it

A The site was never lost, and the people who lived beside it were unaware that it required discovering. When an American academic arrived in 1911, guided there by a local farmer for a small payment, he found families cultivating the terraces and their children playing among the walls. Their names appear in his own notebooks, and one of them had already written his on a wall in charcoal. The phrase that attached itself to the place — the lost city — describes the state of European knowledge rather than any property of the site.

B The expedition was not searching for it either. Its object was a different place entirely, the last refuge of a dynasty that had retreated from the Spanish, and the site now famous was reached on the way and initially treated as a secondary find. That other place was located decades later, in dense forest at a lower altitude, and attracts almost no visitors. Nor was it a city. The evidence now favours a royal estate: a country residence built for an emperor in the middle of the fifteenth century, occupied seasonally by a court of perhaps several hundred with a permanent staff, and never intended to hold a large population. The buildings include no marketplace and no fortification of the kind a defended settlement requires; the graves excavated there contain a mixture of people from across the empire, consistent with retainers rather than citizens; and its scale is that of a large estate rather than a town.

C The engineering is the aspect least appreciated by visitors, because most of it is invisible. The site sits on a steep ridge in a region receiving very heavy rainfall, and the greater part of the construction is underground. Beneath the visible terraces lie layers of gravel and stone chips that carry water away from the foundations, and a system of channels and drains distributes it around the site; something like half the total effort of construction was expended below ground. The terraces themselves are as much a drainage and stabilisation system as an agricultural one, and their surface area is far too small to have fed the population that built them.

D The masonry has attracted the most attention and the least accurate explanation. The finest walls are built of stones cut to fit one another exactly, without mortar, with joints that will not admit a blade. The frequent suggestion that this required unknown technology is unnecessary: the method has been reconstructed experimentally and involves patient repeated fitting, using harder hammerstones to dress the surface and abrasive sand to finish it, at a cost in labour that a state commanding conscripted work could afford. The absence of mortar is functional rather than decorative. In an earthquake the blocks move against one another and settle back, where a mortared wall would crack. The region is seismically active and the walls are still standing after five centuries, which is a stronger endorsement of the technique than any account of how it was achieved.

E The estate was abandoned within roughly a century of its construction, and the reasons are not recorded. The empire collapsed after the Spanish arrival, through a combination of civil war, conquest and epidemic disease that preceded the invaders inland; an estate maintained for a court that no longer existed had no purpose and no labour force. It is worth being clear that abandonment of this kind requires no catastrophe at the site itself. A country house is only a country house while somebody owns the country. The Spanish never reached the site, which is why it survived: everything comparable that they did reach was dismantled, its stone reused and its religious buildings replaced.

F The material removed after 1911 became a long diplomatic dispute. Thousands of items — mostly bones, ceramics and metalwork rather than treasure — were exported under a permit whose terms both governments later interpreted differently, and remained in a university museum for a century while Peru pressed for their return. They were repatriated after litigation and negotiation concluded around 2011, and the episode is now cited routinely in arguments about material held by museums elsewhere, in part because the eventual settlement was reached without a court ruling on ownership.

G The present threat is the number of people who come. Visitor numbers have risen to a level that the ridge, its paths and its drainage were never designed for, and the site has been subject to caps, timed entry and fixed routes for some years, with the limits repeatedly raised under commercial pressure. The dilemma is entirely ordinary and not therefore easy: the revenue funds the conservation, the visitors cause the damage the conservation addresses, and no one has yet found a defensible principle for deciding how many people a place can be shown to before showing it becomes the thing that destroys it.