The Hill That Minted Money: Seventy Years of Digging Under Sardis

It is the summer of 2024, and a trowel is going through ash at the base of a wall that has not seen daylight in twenty-six centuries. The wall is Lydian, mudbrick on a stone socle, and the ash sitting against it is a fire that a Persian army set. Somewhere in that fire, weeks after it burned, someone dumped the bodies of two young men, face down, without ceremony, alongside the toppled brick. Beside the skull of one of them lies a coin the size of a lentil a silver coin one twenty-fourth the weight of a stater, 0.35 gramsA tiny silver fractional croeseid — 1/24 of a stater, weighing 0.35 g — was recovered beside a soldier’s skull in the Persian destruction layer at Sardis, in a stratified context that settled a long scholarly argument over when Croesus’s coinage began. Kroll, “The Coins of Sardis,” Archaeological Exploration of Sardis., worth almost nothing and, to a numismatist, worth almost everything: it is a coin of King Croesus, found in the exact layer of dirt that Cyrus the Great’s soldiers left behind when they took Sardis. The coin does not just tell you a man died here. It tells you when, and it tells you that the strange idea stamped into its face — that a piece of metal could be trusted at a glance, without being weighed or bitten or assayed — was already old news by the time the fire went out.

route 2 routes, 4 points

1. Route 1 (path): Sardis to Ephesus
2. Route 2 (path): Sardis to Gordion
3. Point: Sardis — the Lydian capital, on the Pactolus stream
4. Point: Bin Tepe — the royal tumulus cemetery
5. Point: Ephesus — Temple of Artemis, where the earliest electrum coins were excavated
6. Point: Gordion — the Phrygian capital, later absorbed into Lydia
7. Lydia — western Anatolia, modern Türkiye: 100

This is a chronicle about a hill, not a king. Sardis is a single mound of earth on the Hermus plain in western Türkiye, continuously fought over, built on, burned, and rebuilt for roughly three thousand years, and in July 2025 UNESCO’s World Heritage Committee finally put it on the list, under the name “Sardis and the Lydian Tumuli of Bin Tepe”Inscribed on 12 July 2025 during the 47th session of the World Heritage Committee in Paris, under Criterion (iii) — outstanding testimony to Lydian civilization. It became Türkiye’s 22nd UNESCO World Heritage cultural site. Türkiye Today, 2025; UNESCO World Heritage Centre, official listing No. 1731.. The listing did not happen because someone discovered Sardis. It happened because a joint Harvard–Cornell excavation has been digging into the same hill, one careful trench at a time, for sixty-seven years running by the time the committee voted¹1By the time ScienceDaily covered the story again the following June, the project’s own count had rounded up to “seventy years.” Sixty-seven, or seventy — the point is the same: no other classical excavation on Earth has run this long without a break, save two or three peer projects, and none has produced a stratigraphy this legible. ScienceDaily, June 2026 — and what they have found is not one story but a stack of them, each one buried by the next.

195819581958EXCAVATION REFOUNDEDHarvard–Cornell partnership

67 YRS67 YRS67 YRSOF CONTINUOUS DIGGINGto the July 2025 UNESCO vote

12 M12 M12 MDEEPEST TRENCHthrough river silt

108 HA108 HA108 HATHE WALLED LYDIAN CITYwalls up to 20 m thick

202520252025UNESCO INSCRIPTION12 July, Paris

Six layers, one hill

Sardis sits where the Tmolus mountains drop into the Hermus river plain, with a small stream called the Pactolus running off the mountain and past the citadel — a stream ancient writers insisted carried gold dust down from the hills, and which gave the city its epithet, “the golden”Sardis’s association with gold ran deep enough into antiquity that ancient writers used the epithet routinely; the Pactolus stream that runs past the site was believed to carry gold washed down from Mount Tmolus. Cahill, “The Archaeological Exploration of Sardis.”. Before any narrative can be told about this place, the ground itself has to be read, because the ground is where the story survives. Excavators dig down through it in shafts as deep as twelve meters — the plain is alluvial silt, so a Sardis trench is closer to a mineshaft than a hole²2Cornell’s own account of the dig describes the trenches as “essentially shafts,” some driven twelve meters into the silt before they reach anything Lydian at all — a function of just how much later building has piled up on top of the Iron Age city. Cornell Chronicle, 2025 — and what they pass through, in order, is a rough timeline of every power that has ever held this hill.

Every one of those six things is real, cited, and sitting under (or beside) the same hectare of Anatolian dirt. This chronicle follows them roughly in the order the trowel finds them — from the bottom up, the way stratigraphy actually works — and the thread that ties every layer together is the strangest and most durable thing this hill ever produced: the idea, first proven here, that a stamped piece of metal could be money.

The impurity problem

Coinage sounds like an obvious invention. It was not. For as long as people had traded gold and silver, they had to weigh it and test it every single time, because raw metal is never pure and buyers have every reason to lie. Herodotus, writing more than a century after the fact, gave the credit for solving this to one people in particular:

And of all men whom we know, the Lydians were the first to mint and use a coinage of gold and silver.

The raw material was electrumElectrum is a naturally occurring alloy of gold and silver (with trace copper), panned directly from Anatolian streams including the Pactolus. Its gold content varies unpredictably from one nugget to the next, which is precisely the problem the Lydians had to solve before it could function as reliable money. Kroll, “The Coins of Sardis.”, an alloy that came straight out of the Pactolus and its sister streams looking uniform and behaving nothing like it — the gold content of one nugget could be wildly different from the next³3Natural gold panned from the Pactolus runs roughly 17–30% silver by weight, meaning its gold fraction can range from about 70% to 83%. No two handfuls washed from the same stream are guaranteed to match, which is exactly the property that makes an untreated nugget useless as a fixed unit of value. Bryn Mawr Classical Review of Ramage and Craddock, 2001. The earliest hoard that lets archaeologists see this problem up close was dug up not at Sardis itself but eighty kilometers west, under the foundations of the Temple of Artemis at Ephesus, in 1904 and 1905: ninety-three electrum coins from the late seventh century BCranging from 4.7 grams down to a ninety-sixth of a stater, about 0.15 gramsThe Ephesus Artemision foundation deposit yielded 93 electrum coins spanning weights from 4.7 g down to roughly 0.15 g — a 1/96 stater fraction. About half the deposit weighed under 1.2 g, evidence of an already-sophisticated fractional denomination system in the earliest coinage known. Kroll, “The Coins of Sardis.”, some stamped with a lion’s head and Lydian letters — WALWET, almost certainly recording the name of King Alyattes, and KUKALIM, probably a treasury officialAlyattes, Croesus’s fatherThe lion-head electrum series is attributed to Alyattes, who reigned roughly 610–560 BC and was Croesus’s father; a hoard of 45 royal lion-head coins found at Gordion in 1963 shows the Lydian mint’s reach had extended into former Phrygian territory by the early sixth century BC. Kroll, “The Coins of Sardis.”.

Here is the part that makes the Lydian state look less like an honest inventor and more like a very early central bank. Natural electrum from the Pactolus averages about seventy-three percent gold. The lion-head coins were officially reckoned — traded, taxed, valued — as if they were that same natural, roughly seventy-three-percent alloy. But when modern metallurgists actually assayed the coins, they came out consistently lower: about fifty-four to fifty-five percent gold, with the rest silver and a trace of coppera built-in Lydian state profit of roughly fifteen to twenty percentLydian lion-head electrum coins were officially valued as natural (~73% gold) electrum but assayed at only ~54–55% gold, ~44% silver, and 1–2% copper — a deliberate overvaluation yielding the Lydian treasury an estimated 15–20% profit on every coin struck. Kroll, “The Coins of Sardis.”. The state was not just standardizing weight. It was quietly debasing its own coinage and pocketing the difference — the first known instance of a mint skimming seigniorage off the top, centuries before the word existed.

That workshop is the layer directly beneath the destruction the Persians left behind, and it is where this chronicle goes next.

The workshop by the river

In 1968 and 1969, Cornell archaeologist Andrew Ramage opened a trench in a suburban quarter northwest of the city walls, in a sector the expedition labels simply Pactolus NorthSector PNExcavated by A. Ramage of Cornell University in 1968–1969, outside the city walls, close to the Pactolus stream — a small-scale workshop set among what appear to be ordinary domestic buildings, not a palace industry walled off from the town. Greenewalt, “Gold and Silver Refining at Sardis.”. What came out of the ground was not a palace and not a mine — it was, in the excavators’ own words, “a fairly small-scale atelier surrounded by apparently domestic complexes,” sitting close to the water it neededthe only ancient gold refinery ever excavated anywhereRamage and Craddock’s King Croesus’s Gold (Archaeological Exploration of Sardis 11) describes the Pactolus North installation as the only ancient refinery of its kind yet excavated, and reconstructs its chemistry through parallel experiments at the British Museum. Bryn Mawr Classical Review, 2001.. Pottery in the fill dates the workshop to roughly 575–550 BC — spanning the reign of Alyattes and reaching into that of his son Croesus, who ruled Lydia from about 561 to 546 BCCroesusCroesus reigned Lydia circa 561–546 BC, the last independent Lydian king and, by legend, the wealthiest man of the ancient Mediterranean — the origin of the English idiom “rich as Croesus.” Croeseid, Wikipedia..

filter2 rows

Two furnaces, two metals: how Pactolus North actually split gold from silver, ca. 575–550 BC.

Process↕	Mechanism↕	Approx. temperature↕	Output↕
Cementation	Electrum foil sealed between layers of salt inside lidded clay pots, fired in a brick oven	600–800 °C	Near-pure gold — salt vapor converts the silver to chloride, which migrates into the pot wall and is recovered separately
Cupellation	Silver-lead alloy heated in a shallow porous bowl-hearth under an oxidizing draft from bellows	~1100 °C	Pure silver “buttons,” isolated from the lead and any base metal

Roughly two hundred of those bowl-hearths turned up in the excavation, along with furnace fragments, gold sheet, an iron blowpipe tip, bellows nozzles, and touchstones for testing purity by the streak they leavetouchstonesFine-grained dark stones used to test gold purity by comparing the color of a streak rubbed onto the stone against reference alloys of known composition — recovered among the workshop debris at Pactolus North alongside furnace fragments, tuyeres, and an iron blow-pipe tip. Greenewalt, “Gold and Silver Refining at Sardis.”. The cementation process recovered here is, per the excavators, the earliest known evidence for that specific technique anywhere in the archaeological record — which means the solution to Lydia’s coinage-purity problem was not borrowed from an older civilization. It was invented on this exact hill, by people trying to fix their own money.

1. causePactolus river gold arrives as electrum of wildly unpredictable purity

   Natural alloy panned from Anatolian streams runs 70–83% gold with no way to tell one nugget's fraction from another's by eye.

2. therefore
   effectLydian lion-head coins fake a consistent purity they don't actually have

   Struck and valued as if they were natural ~73% electrum, they assay at only ~54–55% gold — a hidden 15–20% state margin that works only until someone checks.

3. therefore
   effectSardis builds a refinery at Pactolus North to manufacture real purity instead of faking it

   Cementation ovens and roughly 200 bowl-hearths give Sardis, for the first time, the ability to produce gold and silver of guaranteed, testable fineness.

4. and so
   turning pointCroesus mints the Croeseid — the first bimetallic coinage of certified-pure gold and silver

   Two honest coins replace one dishonest alloy; the innovation outlives Croesus's own kingdom by centuries.

What the refinery bought Croesus was not just cleaner metal. It bought trust — the same commodity every monetary system since has been trying to manufacture. And the coin that trust produced did something almost nothing else from the ancient world managed: it kept its shape, nearly gram for gram, through a conquest.

An invention that survives being conquered by the people who conquered it is not a curiosity. It is evidence that the thing being copied was better than anything the conquerors already had. Which raises the obvious question this chronicle has been circling: how, exactly, did Sardis lose the war and win the argument?

The layer of ash and silver

Croesus went looking for that war. According to the tradition Herodotus preserved, he had already asked the oracle at Delphi whether he should attack the rising Persian empire under Cyrus, and had been told that if he crossed the border he would “destroy a great empire”The Delphic oracle’s famous, deliberately ambiguous answer to Croesus — he took it as a promise of victory. The empire it destroyed turned out to be his own. Croesus, Wikipedia. — the sort of prophecy that flatters the man asking it and betrays him regardless. Croesus marched east, fought Cyrus to an indecisive standstill at Pteria in Cappadocia, and pulled back toward Sardis for the winterCyrus caught the city unpreparedAfter the inconclusive engagement at Pteria, Cyrus pressed his advantage and moved against Sardis before Croesus had reorganized his defenses — the surprise itself is part of what the archaeological and historical record agree on. Cahill, “The Persian Sack of Sardis.”. Cyrus did not wait for spring. Herodotus records that Persian forces broke a Lydian cavalry charge by putting camels at the front of their own line — horses unused to the smell panicked and threw their riders — and that the city finally fell when a soldier named Hyroeades found an unguarded stretch of the acropolis escarpment and climbed it in the dark, letting the Persian army in behind the wallsHyroeadesHerodotus names the Persian soldier who scaled the supposedly unclimbable acropolis face at night, having reportedly watched a Lydian defender retrieve a dropped helmet along the same route by day. Once over the wall, the rest of the Persian force followed. Cahill, “The Persian Sack of Sardis.”.

…and thus Sardis was captured and the whole city was sacked.

Historians still argue about which year that sentence describes — the Nabonidus Chronicle points to 547 BC, while the traditional chronology built from Herodotus puts the fall a year later, in 546the exact year is still disputedCyrus’s conquest of Sardis is dated to 547 BC by some readings of the Babylonian Nabonidus Chronicle and to 546/545 BC by the traditional Herodotean chronology; the excavators themselves note the sources conflict and do not force a false precision onto the date. Cahill, “The Persian Sack of Sardis.”. What is not disputed is what the excavators actually found when they dug down to that year, whichever one it was: a siege that lasted fourteen daysHerodotus’s account, preserved in the excavators’ own history of the sack, records that the siege lasted fourteen days before a Persian soldier found an unguarded stretch of the acropolis escarpment and climbed it, letting the army in behind the walls. Cahill, “The Persian Sack of Sardis.”, followed by burning so total it turns up in every direction the trench has opened — fortification walls burned and then deliberately pulled down, houses collapsed under their own thatched roofs, and in places mudbrick fired hot enough to vitrify into a melted, glassy massburning identified along the western, northern, and eastern fortification stretches, at the gates, and in houses buried beneath the later Roman theaterThe destruction layer is not localized to one gate or one street — it is found wherever the expedition has dug down to 546 BC, across multiple, unconnected trenches around the city. Cahill, “The Persian Sack of Sardis.”.

The human cost of that fire came out of the ground in 2024, in the same trench where this chronicle began. Two skeletons, both men in their twenties, lay dumped face down in the rubble at the base of the wall, undecomposed enough to show they were buried within weeks of dying, not years. One had three separate skull wounds — two from a bladed weapon, one from something blunt, probably an axe — plus parry fractures on his forearm from raising it to block a blow, and a rib that had been broken and half-healed weeks before he died at alla pre-existing rib fracture roughly three to four weeks oldForensic analysis of the skeletons found in the 546 BC destruction layer showed one soldier had a rib injury that had begun healing three to four weeks before death — evidence he had already survived combat, likely at Pteria, before dying defending Sardis itself. Cahill, “The Persian Sack of Sardis.”. An excavated iron helmet, crushed and corroded, lay between the two bodies; it now sits in the Manisa Museum. Nicholas Cahill, who has directed the expedition since 2008, described what the bones themselves argue for:

Since these soldiers were thrown at the bottom of the city wall, they are not the military personnel of the victorious Persians, but the soldiers of the defeated Lydians.

Nobody buries a winning army’s dead like garbage. A hundred and thirty-six bronze and iron arrowheads turned up just inside the western gate alone136 arrowheadsThe concentration and mix of bronze and iron arrowheads recovered inside the western gate is read by the excavators as evidence of the diversity of forces on both sides of the fighting there. Cahill, “The Persian Sack of Sardis.”, along with a saber, a sickle apparently repurposed as a weapon, a spearhead, and four household iron spits that someone, in the last minutes, grabbed because there was nothing else left to fight with. And in among all of that wreckage sat the tiny silver coin this chronicle opened with — proof, in the stratigraphy itself, that Croesus’s refined, bimetallic currency was already circulating, already ordinary enough to be lost by a dying man, before the fire that ended his kingdom had even gone out.

The road that ended here

Cyrus did not raze what he took. He kept Sardis as the seat of the Persian satrapy of Sparda, and made it the western terminus of an imperial project every bit as ambitious, in its own register, as the coinage it inherited: the Royal Road, built under Darius I to connect the empire’s capital at Susa to its Aegean frontier.

route 1 route, 2 points

1. Route 1 (path): The Royal Road, Susa to Sardis
2. Point: Sardis — the road's western terminus
3. Point: Susa — the Persian capital, the road's eastern end
4. Sardis — satrapal capital of Sparda: 100
5. Susa — the Achaemenid capital: 60

The road ran roughly 2,700 kilometers and, per Herodotus and the classical geographers, was studded with a hundred and eleven relay stationsone hundred and eleven relay stationsAncient sources credit Darius I with establishing 111 stations along the roughly 2,699 km (1,677 mile) route from Susa to Sardis, each holding fresh horses and riders for the imperial post. Royal Road, Wikipedia. where mounted couriers changed horses without stopping. A message that would take an ordinary traveler ninety days to carry on foot could reach Sardis from Susa in ninenine days by relay, ninety on footThe Achaemenid postal relay system along the Royal Road is credited with cutting a roughly three-month overland journey down to nine days — the fastest long-distance communication achievable anywhere before the telegraph. Royal Road, Wikipedia.. Herodotus’s famous description of riders who let “neither snow, nor rain, nor heat, nor gloom of night” slow them down describes this exact system, and it is worth noticing what it means for Sardis specifically: the city Cyrus’s army burned became, within a generation, the fastest-connected node on the western edge of the largest empire the world had yet seen. Conquest did not end Sardis’s importance. It relocated the importance from a throne to a switchboard.

The coinage moved with the road. Persian-period Lydian mints kept striking the old lion-and-bull design for about three decades under Achaemenid rule, the artistry growing steadily more mechanical as production scaled upthe design hardened and became mechanicalKroll’s account of the post-conquest Lydian coinage describes the lion-and-bull imagery growing progressively stiffer and more standardized under Achaemenid mass production, before Darius I’s reform replaced it outright with the image of the Persian Great King. Kroll, “The Coins of Sardis.” — until Darius replaced the whole design around 515 BC with his own image, a crowned king with a bow, on the gold daric and silver siglos this chronicle already charted. A hoard buried at Old Smyrna around 500 BC, inside a small clay vessel, captures the moment of transition in miniature: four Persian silver sigloi sitting next to two of the last lion-and-bull coins and twenty local Greek fractional staters — three monetary systems, three political eras, in one buried jarfour Persian sigloi, two late Lydian lion-and-bull coins, and twenty Phokaian fractional statersThe Old Smyrna hoard (c. 500 BC) is a typical mixed western Anatolian assemblage of its period — local Greek small change alongside both the outgoing Lydian and incoming Persian higher-value coinages, evidence of how gradually one monetary regime replaced another on the ground. Kroll, “The Coins of Sardis.”.

The night twelve cities fell

Rome eventually absorbed Sardis peacefully, in 129 BC, as part of the new province of Asia, and for a while the city prospered as an ordinary, wealthy provincial capital. Then, in the winter of AD 17, in a single night, the ground under the whole region gave way.

Tacitus counted twelve named cities of Asia leveled in that earthquake, Sardis among them, and Pliny the Elder — writing within living memory of the event — called it, flatly, the greatest earthquake in human memorythe greatest earthquake within human memoryPliny the Elder’s Natural History (2.86.200) records the AD 17 earthquake in these terms; it is independently corroborated by Tacitus, Strabo, and Eusebius, an unusually thick source record for an ancient natural disaster. AD 17 Lydia earthquake, Wikipedia.. Tacitus’s own account of the imperial response survives, and it names Sardis specifically as the worst-hit and the most helped:

The calamity fell most fatally on the inhabitants of Sardis, and it attracted to them the largest share of sympathy. The emperor promised ten million sesterces, and remitted for five years all they paid to the exchequer or to the emperor’s purse.

Tiberius sent an ex-praetor named Marcus Ateius to assess the damage in person and administer the reliefMarcus AteiusTacitus names the commissioner Tiberius dispatched to survey and administer aid to the twelve stricken cities of Asia — direct imperial intervention rather than a distant decree. Tacitus, Annals 2.47, via Wikisource., and the affected cities later erected a colossal statue of Tiberius in the Roman Forum in gratitude, with two more figures added afterward for Kibyra and Ephesus. Some cities were so grateful they folded the emperor’s family name into their own; Philadelphia, hit in the same quake, briefly restyled itself NeocaesareaSardis itself briefly added “Caesarea” to its own nameSeveral of the twelve cities struck by the AD 17 earthquake adopted honorific “Caesarea” names after Tiberius's relief effort — a small but telling piece of civic gratitude carved directly into how a city referred to itself. AD 17 Lydia earthquake, Wikipedia.. Ten million sesterces rebuilt the streets and the walls. It did not fully restore what Sardis had been before the ground moved — the city never quite regained its old regional primacy — but it kept the hill inhabited, prosperous enough, three centuries later, to build one of the largest civic bathing complexes in Roman Asia, a marble bath-gymnasium whose ruins still stand today. One corner of that complex was about to become something Rome never intended.

A synagogue in the gymnasium

In 1962, four years into the Harvard–Cornell expedition’s second life, excavators broke into a hall inside the bath-gymnasium complex and found themselves standing in the largest ancient synagogue ever recovered from the Jewish diasporathe largest known ancient diaspora synagogueDiscovered in 1962 by the Archaeological Exploration of Sardis; the main assembly hall runs more than 50 meters long, roofed to roughly 14 meters, with a capacity approaching a thousand worshippers. “The Synagogue,” Archaeological Exploration of Sardis. The superlative ranking is independently corroborated by Harvard Gazette, 2025 (“the largest of the ancient Jewish diaspora uncovered to date”) and ScienceDaily (“the largest synagogue in the ancient world”).. The building had not been hidden or improvised. It occupied a prominent corner of the city’s premier civic bathing-and-athletics complex, and its main hall ran more than fifty meters long, roofed some fourteen meters up, big enough to seat close to a thousand people at once. Nearly fourteen hundred square meters of geometric mosaic covered the floor across seven structural bays, and the walls carried marble inlay panels — a technique called skoutlosisa Late Roman marble-veneer wall technique using thin, precisely cut panels set in geometric and figural patterns — used at Sardis for camels, birds, fish, and an arched frieze of urns and doves. — depicting camels, birds, and fish, some still bearing traces of the original red paint.

The community that built it did not hide its wealth or its names. More than eighty Greek and seven Hebrew inscriptions survive from the excavation, many recording donors by title — “citizen of Sardis,” holder of municipal office — embedded directly into the floor they helped pay foreighty Greek and seven Hebrew inscriptionsThe Sardis synagogue excavation recovered over eighty Greek-language donor inscriptions and seven in Hebrew, naming patrons who held civic titles alongside their religious ones — direct epigraphic evidence of integration rather than separation. “The Synagogue,” Archaeological Exploration of Sardis.. One donor inscription, set into the mosaic in the plainest possible language, still reads:

I with my wife Regina and our children… executed from the gifts of almighty God all the skoutlosis.

Some of the synagogue’s own decoration had been salvaged from a much older layer of the same hill: inscribed blocks bearing letters of King Antiochus III, dated to 213 BC, and a pair of archaic stone lions, originally carved for the Lydian-era sanctuary of Cybele that had stood nearby, quietly repurposed by their new owners — possibly reread, by then, as the Lions of Judahspolia from the earlier sanctuary of CybeleThe synagogue’s builders incorporated older architectural fragments from the Metroon, the Cybele sanctuary that had stood at Sardis since the Lydian period, including inscribed blocks recording letters of Antiochus III (213 BC) and a pair of archaic lion sculptures — reuse that layers Hellenistic, Lydian, and Jewish Sardis into a single wall. “The Synagogue,” Archaeological Exploration of Sardis.. The synagogue functioned into the early seventh century, when an earthquake and the region’s slow decline finally emptied it out; in 2021, a protective roof was finally built over the site, letting visitors walk across mosaic that had waited fourteen centuries in the open aira protective roof, finally built in 2021Excavated mosaics and marble at the Sardis synagogue were originally consolidated and reset in the 1960s; a permanent protective roof was completed in 2021, allowing public access to the site’s ancient floors year-round. “The Synagogue,” Archaeological Exploration of Sardis..

The acropolis that outlasted Rome

Go deeper: how the hill kept losing empires without disappearing

Sardis’s Byzantine centuries read like a slow retreat uphill. The lower city, exposed and increasingly hard to defend, was sacked by Sasanian Persian forces in 615 during a Byzantine–Sasanian war, and the settlement never fully recovered its earlier footprint afterward. What did keep functioning was the old Lydian acropolis itself — the same fortified summit that had once held Croesus’s palace — repurposed as a last defensible refuge while the town below thinned out. The Seljuk Turks took the city in 1078; the Byzantine general John Doukas retook it in 1097; the cycle of loss and reoccupation continued for another three centuries, until Timur’s forces are believed to have destroyed what remained in 1402. By the 1700s, European travelers found only two small hamlets where a Lydian capital of a hundred and eight hectares had once stood. Sardis, Wikipedia.

The acropolis is the last layer before the modern one: a fortress built on a fortress, on the same summit Croesus once ruled from, garrisoned and re-garrisoned by Byzantines, Seljuks, and Byzantines again, until there was nothing left worth garrisoning.

Seventy years, four directors

Today’s excavation runs three months a year with fifty to sixty international scholars and studentsHarvard Gazette, 2025, describes today’s Sardis team as averaging 50 to 60 scholars and students from around the world, joined by a similar number of local diggers, working three months each summer. plus a local Turkish workforce that now makes up more than half the teamover half the researchers are Turkish experts and studentsThe 2025-era Archaeological Exploration of Sardis draws its team from Harvard, Cornell, the University of Wisconsin–Madison, UC Berkeley, and Turkish institutions, with Turkish experts and students now composing more than half the working roster. Harvard Gazette; ScienceDaily., and Cornell doctoral student Leyla Uğurer, who grew up near the site, now surveys the rock-cut tombs her own region has always lived besideLeyla UğurerA Cornell doctoral student in history of art and archaeology, local to the Sardis region, who surveys the rock-cut tombs and excavates Roman-period contexts on the modern team. Cornell Chronicle, 2025.. The romance of a seventy-year dig obscures how much of it is a rescue operation running against a clock. Associate professor Benjamin Anderson, who documents the acropolis walls for the current team, described the threat to Bin Tepe’s tumuli in blunt terms:

[Looting has reached] industrial dimensions.

Treasure hunters now work the tumulus field with dynamite and bulldozers as well as shovels, and ordinary agriculture has erased still more of the field over the last two centuriesnumerous tumuli destroyed by farmingOf the roughly 149 tumuli believed to have originally stood at Bin Tepe, only about 115 survive; the rest have been leveled for farmland, on top of the ongoing losses to organized looting the current excavation team describes. Bin Tepe, Wikipedia; Cornell Chronicle, 2025.. It is a strange, sobering fact to sit next to the UNESCO plaque: the same summer the world agreed this hill was worth protecting forever, the archaeologists digging it were describing, in the same breath, how much of it is actively being destroyed.

Event 1 of 7: 1 Jan 1910, Butler's Princeton dig begins

011910opening

1 Jan 1910

Butler's Princeton dig begins

First modern excavation; finds the Temple of Artemis and over 1,000 Lydian tombs before war ends the project.

021958+48 yr

1 Jan 1958

The Harvard–Cornell partnership is founded

Hanfmann and Detweiler restart continuous excavation after a roughly 36-year gap, since Butler's campaign briefly resumed in 1922 before lapsing.

031962+4 yr

1 Jan 1962

The synagogue is uncovered

The largest known ancient diaspora synagogue, inside the Roman bath-gymnasium.

041968+6 yr

1 Jan 1968

The gold refinery is excavated

Andrew Ramage opens Sector PN at Pactolus North — the only ancient gold refinery ever found.

051976+8 yr

1 Jan 1976

Crawford Greenewalt Jr. becomes director

Leads the project for three decades, through the Lydian-era publications.

062008+32 yr

1 Jan 2008

Nicholas Cahill becomes director

Oversees the most detailed study yet of the 546 BC destruction layer.

072025+18 yr

12 Jul 2025

UNESCO inscribes Sardis and Bin Tepe

47th session of the World Heritage Committee, Paris — Türkiye's 22nd listed site.

The timeline above is really two timelines laid on top of each other, and that is the whole point of this chronicle. One counts empires: Lydian, Persian, Hellenistic, Roman, Byzantine, Seljuk, Ottoman, each one burying the last one’s rubble a little deeper into the hill. The other counts archaeologists: Butler, Hanfmann and Detweiler, Greenewalt, Cahill, each one digging a little further down into what the empires left. The empires all lost the hill eventually. The diggers, remarkably, have not — not to war, not to the roughly three and a half decades of war and its aftermath that stopped Butler’s team, not to a Cold War, not to seventy years of Turkish and American politics moving in and out of alignment. The project outlasted more governments than the coinage it studies outlasted kings.

And underneath both timelines sits the actual argument this chronicle has been making the whole way down: an idea invented on this hill, to fix a very local problem with river gold, turned out to be more durable than any empire that ever held the ground above it. Croesus lost his kingdom in a matter of weeks. His coin standard survived Cyrus, survived Darius, and — reduced now to a gram of silver in a museum drawer in Manisa — survived twenty-six more centuries of siege, earthquake, fire, and looting to end up, this time, cited correctly.