Petra

Why this is extraordinary

Petra was the Nabataean capital, a trading city in a sandstone basin east of the Dead Sea rift. The World Heritage listing calls the rock-cut façades, among them the Khasneh, the Urn Tomb, the Palace Tomb, the Corinthian Tomb, and the Deir, an architectural ensemble of the first centuries BCE to CE. It also says the city was possible because of dams, a tunnel, channels, aqueducts, reservoirs, and cisterns that caught and steered seasonal water. The remains on the site run, in the same listing, from prehistoric times to the medieval periods. That long span is real. The great façades are the Nabataean part of it.

Al-Khazneh, the Treasury, is a tomb front cut in the cliff where the Siq opens. It carries no text that names who was buried there. An excavation of the court in front of it found older graves that the carving cut through. A handbook of the city’s history places the façade in the reigns of Obodas II or Aretas IV, from 30 BCE to 40 CE, because of those graves and the paving of the Siq. Ad Deir, the Monastery, is a larger front of the same family, higher on the mountain. No year is cut on it in the sources used here.

The order of the work is written in what was never finished. The regional authority describes a tomb on the east face of Al-Habis that was abandoned after about a quarter of the façade. The cuts run from the top downward. The tools named there are the hammer, the pick, and the chisel. From the Madbah slope, another cutting shows a court taken out of the cliff with doorways only outlined. A rock-cut face has no spare block. The unfinished tombs are the mistakes and the stops, left where they were made.

The stone is the contrast. Petra’s façades are sandstone, as at Hegra, where the same kingdom dated its inscribed tomb doors from the year 1 to 75 CE. The scratch order, and the difference between a quartz grain and a weak cement, are in the stone panel. Sheltered faces, the Treasury in its alcove among them, keep their detail. Exposed faces wear. That difference follows shelter and bedrock. It is not a second date.

The same cliffs have a sky reading, published from a 2013 survey of 50 Nabataean temples, shrines, royal tombs, and high places. Most of the bearings were taken in December 2011, with compasses corrected for the magnetic shift. The survey gives its average error as a quarter of a degree in azimuth, the compass bearing, and half a degree in the height of the horizon, which becomes about three quarters of a degree in declination. Declination is how far a rising or a setting sits from the celestial equator, so cliffs in different valleys can be compared. The directions bunch near the equinoxes, the days when night and day are about equal, at a peak around 0 degrees of declination, and near the solstices, the days when the sun's rising and setting stop and turn back. One solstice peak sits near 24 and a quarter degrees, and another near minus 25 and a quarter. A further peak near 8 degrees has no simple star or sun to match. Due-north bearings also pile up, which can happen because many horizons share the same northern declination even when nobody aimed.

At Ad Deir, the Monastery, winter-solstice sunset sends light through the door onto the inner platform, the motab, where the sacred stones would have stood. The light holds for about a week on either side of that day. From the same platform the sun sets against a rock the survey reads as a lion's head, and the lion is an animal of Al Uzza. Two thousand years ago the sun's northern rim at midwinter stood near 23 and a half degrees, so the effect would have been tighter then. Al Uzza, "the Most Powerful", is read as the evening star, the planet Venus. Dushara, the chief god, whose name means "He of Shara", the range east of the city, is read as a sky god, solar or lunar, and also as a god of the dead. One ancient report of Nabataean worship names the winter sun. An older Greek notice says they worshipped the sun on the roofs of their houses.

The Urn Tomb is the other measured door. Its gate is centred on equinox sunset over the mountain Umm al-Biyara, between two notches on the summit, and the same hall is argued to take summer-solstice sunset and winter-solstice sunset on further notches of that western skyline. On 21 December 2011 the last light of the winter solstice, watched from the court, reached the northeast corner of the inner hall. The Treasury, Al-Khazneh, looks out through the Siq at a slice of sky only about 18 degrees high, which cuts off a solar line. A lunar reading of that façade has been suggested, and it has not been shown. A southern peak near the star Canopus faded when the same bearings were set against a random spread of directions, so that one match is the weak one.

The catch is that a cliff already facing west will see some sunset, and a door cut for a tomb can inherit the way the rock happens to face. The water system and the unfinished tops do not need a solstice. The sky reading and the mason's reading use the same valley, and neither one has to lose for the other to stay.

One reading places the great façades in the Nabataean centuries without a year on the Treasury door, because the door cuts graves of that same age and matches a dated series at Hegra. A reading that puts the fronts much earlier would have to explain those graves, the unfinished tops, and the water system of the same centuries. Both readings are still live, and they use the same valley. The prehistoric and medieval remains are a different fact. They are the long life of the valley, and they are not asked to be the signature on the Treasury.