Nemrut · Turkey · 37.9808, 38.7406 · PYR record in the living atlas
Antiochus I ruled Commagene, a buffer kingdom wedged between Rome and Parthia in the first century BC, and answered its smallness with altitude. On the summit of Mount Nemrut, 2,134 metres up, he raised a conical mound and flanked it with two terraces of enthroned colossi: Zeus, Apollo, Herakles, the goddess Commagene, and, seated in the row as their equal, Antiochus himself. The figures stood eight to nine metres tall with their thrones, though published estimates range from roughly seven to ten. The built sanctuary covers more than 26,000 square metres, and the terraces were cut into the living rock of the peak. The bodies are still on their thrones. The heads, each about a metre and a half to two metres tall, are on the gravel in front of them, lined up and looking out across Anatolia, and that accident of ruin has done more for the mountain's fame than anything the king had carved.
A god list that hedges its bets
None of the gods on the terraces has a single name. Zeus is Zeus-Oromasdes, folding in the Persian Ahura Mazda. Apollo is Apollo-Mithras-Helios-Hermes. Herakles is Artagnes-Herakles-Ares. Antiochus claimed descent on one side from the Achaemenid Darius and on the other from Alexander's successors, and the naming is that claim written into the theology rather than merely stated beside it.
He also had his own word for what he was building. He called it a hierothesion, a tomb sanctuary, a term that turns up in Commagenian inscriptions and almost nowhere else. Across the backs of the statues runs a Greek text more than 200 lines long, the nomos or sacred law. It is not poetry. It is an operating manual: a priest in Persian dress is to set a golden crown on the statues, burn incense at the altars, sacrifice animals, and lay on banquets with wine and music. Two festivals are specified every month, one for the king's birthday and one for his accession, with revenues assigned to fund them permanently.
Rome absorbed the kingdom into the province of Syria in 72 AD, under Vespasian, and the cult almost certainly ended there. Whether it ever properly began is a separate and open question. Historians of Hellenistic religion disagree over whether the rites Antiochus set out in such detail were substantively observed after his death, or whether the nomos is chiefly royal self-presentation with very little actual cultic life behind it. A very long inscription about festivals is evidence that a king wanted festivals. It is not by itself evidence that they happened.
East for sunrise, west for sunset
The two terraces are set outward from the tumulus, the east one facing sunrise and the west one sunset, and mainstream scholarship treats that as designed rather than accidental. It is consistent with the solar and fire-centred cult the nomos describes, which is the ordinary reason for accepting an orientation as intentional: the building agrees with the text carved on it. Both terraces also carry dexiosis stelae, reliefs of Antiochus shaking hands with an individual deity, which is the ruler cult in its plainest available form, the king greeting a god as an equal.
Past that, the alignment claims thin out quickly. Bearings have been proposed for the terrace walls, for processional routes and for sightlines between the statues, and none of them has been established by published, peer-reviewed measurement. One feature is worth labelling honestly while we are here: the stepped fire altar on the east terrace is a restoration rather than a survival. Theresa Goell rebuilt it in 1973.
A horoscope in stone, and a date nobody agrees on
Among the reliefs is the Lion Horoscope: a lion pacing right, its body scattered with nineteen eight-rayed stars, a crescent at its throat, and above its back three sixteen-rayed bodies labelled in Greek. They are named for gods rather than planets. Pyroeis of Herakles, the flaming one, is Mars. Stilbon of Apollo, the shining one, is Mercury. Phaethon of Zeus, the radiant one, is Jupiter. The lion is Leo and the star riding above the crescent is Regulus. It is among the oldest horoscopes carved in stone anywhere, which means the monument timestamps itself against a sky that anyone can recompute.
The recomputation is where it becomes interesting. Otto Neugebauer and H. B. van Hoesen worked the configuration and landed on 7 July 62 BC, allowing 6 July as an alternative, and that date has been repeated ever since, usually attached to Antiochus's confirmation on the throne under Roman favour. Their search was bounded, though: they looked inside a window of roughly 100 to 50 BC, because that is where the reign sits. Maurice Crijns, working with the Dutch project on the mountain and publishing in Brijder's 2014 volume, removed the window and found that the arrangement has a single clean match at 14 July 109 BC, which he suggests may commemorate Antiochus's birth alongside his father Mithradates I's accession. Both dates come out of the same sky and the same relief. The disagreement is about which assumption you make before you start searching, and saying so is more honest than pretending the stone names a year.
The mound is not a pile of rubble
Nearly every description of Nemrut, including one we have published ourselves, calls the tumulus a 50 metre heap of crushed stone, roughly 145 to 150 metres across depending on which source you take, and treats the missing tomb as a chamber buried somewhere inside that heap. Herman Brijder, who directed the Dutch archaeological work on the summit and edited the 2014 volume reporting it, describes the structure differently: the solid rock of the mountain top was shaped into the tumulus and then covered with a layer of hand-broken, fist-sized limestone pieces four to five metres thick. If that is right, the mound is not a pile with something concealed in it. It is a dressed peak, and a burial would have to be cut into bedrock.
How any of the stone got up here is also not understood. No ancient quarry has been definitively identified, and the quarrying, transport and setting of blocks weighing several tonnes each, some of them reported above nine, have not been satisfactorily explained in published scholarship. Building a fifty metre cairn at this altitude was a display of state capacity. Exactly what that capacity consisted of, nobody can currently say.
That reframes a century of failure. Karl Sester, a German engineer surveying for the Ottoman government, brought the summit to Western attention in 1881, and Karl Humann made the first documented survey in 1883. Systematic archaeological work began in 1953 with a joint American and German team, Theresa Goell working for the American School of Oriental Research alongside the epigrapher Friedrich Karl Dörner. Goell, whose crews uncovered the heads the site is now famous for, began excavating in 1954, ran two years of geophysical probing from 1963, was still working on the mountain in 1973, and never located the chamber or finished her report. She died in 1985, and her results were finally published in two volumes edited by Donald H. Sanders in 1996. She and the Lerici Foundation at one stage set off fifty one-kilogram dynamite charges trying to open a route in, damaging statues in the process. The Dutch-led International Nemrud Dağı Project, which worked here from around 2000 into the early 2010s and whose results Brijder edited into the 2014 volume that is now the standard reference, used non-destructive geophysics instead.
Saying that geophysics found nothing is not quite accurate, and the detail matters. In September 2012 a team reported detecting a pyramidal void beneath the tumulus, roughly six metres long, two and a half metres wide and three metres high, apparently containing a large object. Nothing has been excavated to test it, and the methodology and data have not been widely peer-reviewed, so what exists is an anomaly rather than a discovery. Popular accounts that turn that anomaly into an intact royal burial full of gold are going a very long way past what the survey supports.
Antiochus's inscription states that he lies in the mountain. Nobody has been able to test that sentence, and it is possible nobody can without dismantling the monument to do it.
Why the heads are on the ground
The standard explanation is earthquakes, and the East Anatolian Fault runs about five kilometres away, so it is not a foolish one. It does not fit the pattern, though. The west terrace collapsed far more completely than the east, and the two face each other on the same small summit, so a shock strong enough to flatten one should have flattened both. The precise cause and date of the collapse remain unknown, and the difference between the two terraces is itself unresolved; earthquakes, weathering and even attempts at ancient grave robbing using explosives have all been proposed.
Ahmet Turer, Yasemin Didem Aktas and Neriman Sahin Guchan modelled the problem for the Middle East Technical University programme and published in the International Journal of Architectural Heritage in 2012. Their conclusion is that snow loading on the west terrace is critical by itself, sufficient to slide and overturn the statues with no help from the ground. The west blocks are also thinner, around 20 centimetres against 40 on the east, and the foundations there had eroded. Wind piles snow on that side of the peak. On their reading, a single earthquake cannot heavily damage the statues; repeated shocks did cumulative work on the east while snow did the demolition in the west.
Aktas and Turer returned to the question in the Journal of Cultural Heritage in 2015 and weighted it differently, showing by simulation that blocks laid without mortar or metal clamps separate readily under lateral load and that amplified earthquake records bring the statues down completely. Their proposed remedy is telling: textile straps used as reversible confinement belts, so the courses can still shift and shed energy instead of being locked rigid. The two papers do not agree on how to split the credit between weather and seismicity. They do agree that "an earthquake knocked the heads off" is too simple, and that is the part worth carrying away.
The same programme has been cataloguing the wreckage since 2006: 768 architectural blocks recorded with their deterioration mapped, 400 stone cube samples tested in the laboratory, and more than 39 measured drawings, all of it aimed at working out what stood on that terrace before it came down.
What our engines found
We ran the same test here as everywhere else, with the settings fixed in advance and identical at every one of the fifty sites we cover: a 200 km search radius, at least 12 km between sampled sites, 15 sites, a 2 per cent tolerance, and 2,000 Monte Carlo runs against a null model resampled from real catalogued sites in the same region. The settings are frozen because the answer moves if you tune them, and tuning them after seeing the answer is exactly the trick we object to elsewhere. Nemrut returned 41 golden-ratio pairs where chance predicts 40.4, giving p = 0.4958.
That is almost exactly what chance would predict, and it is not significant. The control points the same way: the arbitrary ratio 1.75 scored 1.17 times its expected count against phi's 1.01, so nothing here singles out the golden ratio.
We publish the misses because a tool that only reports its hits is not measuring anything. Every figure above is reproducible from the same coordinates, the same settings and the same seed.
One limit worth stating plainly, because it is the one most often glossed over. This test compares positions between monuments. It says nothing about any building's own proportions, its height, base, slope or perimeter. That is a separate question, and at several sites a genuinely open one.
It says nothing about the sky either. Dating the Lion Horoscope is an astronomical question about one relief, and a regional test of distances between monuments has no purchase on it in either direction.
Explore it yourself
Open the Nemrut record in the living atlas to see the summit among its catalogued neighbours and set the sky engine to 7 July 62 BC and then to 14 July 109 BC, so you can compare the two candidate readings of the Lion Horoscope for yourself. The Nemrut site-emblem tee carries the fallen heads, the site name and the exact coordinates in the Pyramason Chronicle style.
Sources worth reading
H. Brijder (ed.), Nemrud Dağı: Recent Archaeological Research and Conservation Activities in the Tomb Sanctuary on Mount Nemrud (De Gruyter, 2014), the synthesis of the International Nemrud Dağı Project and the current standard reference; it contains Crijns on the horoscope, Schipperheijn and van Nijf on the epigraphy, and Jacobs on Commagenian religion and royal ideology. D. H. Sanders (ed.), Nemrud Dağı: The Hierothesion of Antiochus I of Commagene, 2 vols. (Eisenbrauns, 1996), Theresa Goell's work published eleven years after her death. F. K. Dörner, Nemrut Dağı'nın Zirvesinde Tanrıların Tahtları (Türk Tarih Kurumu, 1990). Turer, Aktas and Sahin Guchan in the International Journal of Architectural Heritage (2012), and Aktas and Turer in the Journal of Cultural Heritage (2015), for the collapse modelling.