Karnak · Egypt · 25.7188, 32.6573 · PYR record in the living atlas
The Egyptians called it Ipet-sut, the most select of places. No single pharaoh built it. Roughly thirty of them added to it over about two thousand years, from the Middle Kingdom into the Roman period, which means Karnak was an active construction site for longer than the Roman empire lasted. The earliest recognisable remains belong to the 11th Dynasty reign of Intef II, around 2050 BC, and the first formal temple was raised by Senusret I a century or so later. Both dates get quoted as the start, and they mean different things: one is the oldest material in the ground, the other is the first building anyone would call a temple. The precinct of Amun-Re at its heart covers around 25 hectares and is only one of four walled precincts on the site. The temple was entered from a quay that is now dry and sits several hundred metres from the Nile, because the river has moved since.
The stone arrived by water. The sandstone was quarried at Gebel el-Silsila, roughly 100 km upriver, and moved down by barge, then hauled on sledges and ramps; the Middle Kingdom elements are limestone from the Tura quarries near Memphis, which is a much longer haul in the other direction. Ten principal pylons define the two axes of the Amun precinct, east to west and north to south. Several of them were never finished, and the unfinished First Pylon, the one visitors walk through, is a live question: which dynasty commissioned it, and why the work stopped, are both unresolved.
A forest of stone
The Great Hypostyle Hall is the set piece: 134 columns in 16 rows across about 5,000 square metres, though published figures for the area vary and that one should be read as an estimate. Twelve of them, down the central aisle, stand around 21 metres tall with shafts over three metres in diameter and open papyrus capitals; the other 122 stand about 10 metres. The architraves on top are estimated at around 70 tonnes each. Egyptian builders had no arch capable of these spans, so the entire roof sat on that forest of stone, lit through clerestory grilles high above. The height difference between the central twelve and the rest is not decoration, it is what made the clerestory possible. Most of the roof is gone, which is why the hall now reads as a canyon open to the sky.
It was begun by Seti I and finished by Ramesses II in the 13th century BC, and you can see the handover in the technique: raised relief on Seti's northern half, and on the south a switch to the quicker sunk relief under Ramesses. On an adjoining wall Ramesses had his peace treaty with the Hittites carved, so the hall puts a diplomatic document and the columns being paid for at the same moment within a few paces of each other.
Recording it is a life's work. The Karnak Great Hypostyle Hall Project, run since 1992 by Peter Brand of the University of Memphis with the Université du Québec à Montréal, has documented the reliefs wall by wall. Part of its source material is the photographic record made by the Epigraphic Survey of the University of Chicago in the 1970s, and the reason that matters is bleak: the surfaces have decayed so fast since then that fifty-year-old photographs now preserve detail the stone no longer has.
The night eleven columns fell
In 1899 eleven of those columns collapsed one after another. The hall had stood for over three thousand years and then came down in an afternoon, undermined by rising groundwater and salt. Georges Legrain, appointed overseer of antiquities at Karnak in 1895, was handed the job of putting them back, laying proper foundations underneath first. He recorded the work complete in 1902, and comparable strengthening followed across the rest of the hall. The stone is genuine, and some of what holds it up is a French engineer's repair.
The pit full of gods
In 1903, digging in the courtyard before the Seventh Pylon, Legrain hit something else. Beneath the court lay nearly 800 stone statues and something in the order of 17,000 bronzes, spanning the Old Kingdom to the Ptolemaic period, over 350 of the stone pieces being block statues. It is the largest hoard of Egyptian statuary ever recorded, and clearing it took until 1907, fought the whole way by groundwater flooding the pit.
The Karnak Cachette was not a robbery or a disaster. It was housekeeping. Temples accumulated votive statues for centuries until there was no floor left, and at some point the priests buried the overflow respectfully inside the precinct rather than destroying it. The objects were dispersed to Cairo and to museums worldwide, and for most of a century nobody could say with confidence what had been in the hole; a catalogue project begun in 2006 has been reassembling the list. That is the untold half of Karnak: not the columns, but the paperwork of an institution that ran out of shelf space.
The heretic hidden in the walls
Akhenaten built for the Aten at Karnak before he abandoned Thebes, and he built in a standardised small block about 27 by 27 by 54 centimetres, known as talatat, light enough for one man to carry and therefore very fast to lay. When his successors dismantled those temples they did not discard the blocks. They used them as rubble fill inside their own pylons, which is why tens of thousands of talatat have come out of the Second, Ninth and Tenth Pylons.
The consequence is one of the strangest archives in Egypt. Each block carries a fragment of a carved scene, and because they were packed away almost immediately the paint and carving are often crisp. The Akhenaten Temple Project has spent decades photographing them and matching fragment to fragment to rebuild the wall scenes of a king his own dynasty tried to erase. His enemies preserved him by hiding him.
The axis and the midwinter sun
Karnak's main axis is oriented so that the sun rises along it at midwinter, and the length of the corridor means light travels far into the temple then and much less at any other time of year. This is usually credited to Norman Lockyer, whose Egyptian fieldwork in the 1890s launched the subject of temple alignment, and the credit is close but not exact. Lockyer read the main axis the other way along its length, as pointing to midsummer sunset out to the north-west. The midwinter sunrise reading is the same line taken from the inner sanctuary outward, and it was clarified by later workers, E. C. Krupp among them, on survey that corrects for the slow change in the obliquity of the ecliptic, which shifts where the solstice sun rises over thousands of years. The alignment itself is not in doubt. It is one of the few solar claims at a major ancient site that rests on published measurement rather than on repetition, and the Egyptian Ministry of Tourism and Antiquities marks the sunrise each year around 21 December.
An obelisk of Hatshepsut still stands in the temple, just under 30 metres of single-piece Aswan granite, the second tallest ancient obelisk standing anywhere and the tallest still standing in Egypt. Published heights for the Karnak obelisks vary between sources by a metre or so, and how much of them was originally sheathed in gold or electrum is still argued over, so treat both the figure and the finish as unsettled. Its twin lies broken nearby, which is its own honest lesson in how often these projects failed halfway.
Why this spot
The axis is well evidenced. Why the temple stands on this particular ground is not. In 2025 Ben Pennington, Angus Graham, Aurélia Masson-Berghoff and colleagues published a geoarchaeological study in Antiquity, built on tens of thousands of analysed pottery sherds from cores driven across the floodplain. Their reconstruction puts early Karnak on a natural island of high ground once ringed by Nile channels, the only elevated ground of that character anywhere in the Theban region. The authors raise, carefully and as a possible connection rather than a conclusion, that the site may have been chosen because it resembled the primeval mound of the Egyptian creation story, the first land rising out of the waters of chaos. That is a hypothesis under active discussion by the people best placed to test it. It is not a finding yet, and we are not going to report it as one.
What the study does establish is that several candidate reasons for the location now compete and nobody has ranked them: the solstice axis, the river frontage and its quay, the processional route running south to Luxor Temple, and the island itself. That avenue, incidentally, carried sphinxes of two kinds, human-headed and ram-headed, not one, which is worth saying because half the popular accounts mention only the first.
The complex was managing water from early on. The Sacred Lake is fed by groundwater rather than filled from the river, and underground drainage works run beneath the precinct, including a shaft now attributed to Amenhotep I. In 2025 François Larché re-examined, in Pražské egyptologické studie, the platform Henri Chevrier dug out in 1947 and concluded it can only be Byzantine or later, built to fill that shaft. This is the ordinary condition of the site rather than an oddity: the deep stratigraphy at Karnak lies under monuments nobody is going to dismantle to reach it, so the sequence of the earliest phases and the true extent of the Middle Kingdom buildings beneath the New Kingdom overbuilding both remain open.
After the pharaohs
Karnak was sacked by the Assyrian king Ashurbanipal in the 7th century BC, the event the Hebrew Bible records as the destruction of No-Amon. That did not end it. Building and restoration resumed under the Nubian 25th Dynasty and the Saite 26th, and the complex went on being used and rebuilt through the Ptolemaic and Roman periods, with Christian reuse after that. Excavations by the Supreme Council of Antiquities in front of the western facade between 2007 and 2013 turned up a Ptolemaic bath complex, a Roman bath facility, a residential quarter running from the late Ptolemaic period into the 4th century AD, an ancient dock, and structures associated with Taharqa. The building span is not two thousand years ending with the last pharaoh. It is well over two thousand, and the last phases are Roman and Christian.
Still being dug
Karnak is not a finished excavation. The Centre Franco-Égyptien d'Étude des Temples de Karnak, the joint French and Egyptian research unit that has worked the site for decades, announced in February 2025 that its team digging east of the open-air museum had found a ceramic pot packed with jewellery: gold rings, gilded amulets, statuettes and beads, including a gold triad of the three Theban gods Amun-Re, Mut and Khonsu. The deposit is dated to the early 26th Dynasty, roughly 664 to 525 BC. The pieces went on show at Luxor Museum that May. More recently the same centre reported that restoring the enclosure gate of Ramesses III turned up a stele of the emperor Tiberius, raised long after the last pharaoh.
One claim we will not repeat
Popular accounts sometimes present Karnak as a single cosmological scheme, laid out whole at the start and executed slowly over centuries. A century of stratigraphic work says the opposite. The complex grew opportunistically, with successive rulers building over, into and out of their predecessors' work for political reasons at least as much as theological ones. There is no documentary or physical evidence for a grand original plan, and no published measurement behind the claim. Absent that, it is not a finding, and we will not pass it on.
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. Karnak returned 44 golden-ratio pairs where chance predicts 41.3, giving p = 0.3318 (run on 1 October 2026).
That is a little more than chance would predict, and it is not significant. The control points the same way: the neutral control ratio 1.85 scored 1.22 times its expected count against phi's 1.07, 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. When the catalogue grows, the figures can move, so we rerun these tests and update this section rather than leave an old number standing.
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.
Explore it yourself
Open the Karnak record in the living atlas to walk the complex in 3D, see it among its neighbours in the Theban landscape, run the alignment and ley line engines on the real coordinates, and set the sky to the winter solstice of 1250 BC. The Karnak Temple site-emblem tee carries the site's emblem, the site name and the exact coordinates in the Pyramason Chronicle style.
Sources worth reading
Krupp, Echoes of the Ancient Skies (1983), for the winter-solstice axis and the correction to Lockyer. Pennington, Graham, Masson-Berghoff et al., "Conceptual origins and geomorphic evolution of the temple of Amun-Ra at Karnak (Luxor, Egypt)", Antiquity (2025), for the island and the primeval-mound hypothesis. Larché (2025) in Pražské egyptologické studie on the Chevrier platform and the Amenhotep I drainage shaft. Wilkinson, The Complete Temples of Ancient Egypt (Thames & Hudson, 2000). The IAO Portal to the Heritage of Astronomy entry for Karnak. The Karnak Great Hypostyle Hall Project for the reliefs, and the Karnak Cachette catalogue for the statue hoard.