Histria · Constanța county, Romania
Settlers from Miletus built Histria on a tongue of rock on the Black Sea shore. From the ruins today you see Lake Sinoe, and the open sea is almost eight kilometres away. The sea level barely changed. New land, made of sand, grew in front of the city.
Scroll to go back to 630 BC. The dry land is the real terrain, measured from satellites. Water depths and the dates of the new land are reconstructed from the studies cited.
The city was founded in the second half of the 7th century BC: Pseudo-Skymnos gives 630, Eusebius 657/656. The name comes from Istros, the Greek name for the Danube. The red line shows the approximate shore of that time, after the scheme published by Bivolaru and colleagues in 2021, running south from the city to the cliffs at Vadu.
Where Lake Sinoe and the Chituc barrier lie today was open sea. The city had the sea to its east.
The acropolis, the highest and oldest part of the city, stands on an island of greenschist. A ridge of sand thrown up by the waves, about two kilometres wide, tied it to the mainland, dated by luminescence to 1350–850 BC. It was already there when the Greeks came.
Vasile Pârvan called it a "tombolo" in 1915 and the name stuck. Geomorphologists say it is wrong: none of the processes that build a tombolo are at work here.
Histria's harbour has never been found. Inscriptions mention it from the 3rd century BC to the 2nd century AD. Cores drilled in 2015–2017 point to an anchorage on the south side of the island, perhaps also a sheltered basin inside.
The ships in the picture sit where that model puts them. They are illustrative: we do not know what the ships that called here looked like.
In AD 100 the governor Laberius Maximus confirmed the boundaries of Histria's territory. In the 2nd century an inscription mentions the "remaking" of the harbour, perhaps dredging, perhaps a move. Either way, the sea still reached the city.
On the model, the shore around the city is still that of 630 BC.
In 238 the city was destroyed. The rebuilt town fitted inside a circuit of about six or seven hectares. This is the wall you see on the ground today.
In the 6th century Histria still had rich houses, and a ridge kept its way to open water. To the north, a Danube branch, the Dunavăț, had been building a new delta lobe for several centuries. The waves carried its sand south, straight towards Histria.
From about AD 650 the Saele–Chituc beach-ridge plain begins to form east and south of the city. The University of Bucharest team dated it by luminescence. At its fastest, roughly between 970 and 1230, the shore advanced into the sea by 10–15 metres a year, ten times faster than the old ridge had grown. The figure is how fast the shoreline moved, not how much sand was laid down.
How sand is dated. Sunlight wipes a signal from quartz grains. Once they are buried, natural radiation in the ground charges it up again. A lab measures how much has built up, which gives the time since the grain last saw the sun. A few samples from the ridges are dated, not every metre of shore.
In the same late 7th century the city was abandoned. The researchers warn that sand was not the only cause: many towns in Dobruja were abandoned in those years, a time of general instability.
Exactly when this land rose is disputed. The Bucharest team gives AD 650–1230. Giosan and colleagues, in 2006, give 450–1350 for the youngest set of ridges here.
The ochre hatching marks ground that, in a given year, existed by one dating but not by the other. As you scroll, the hatched band sweeps across the plain.
In front of the city today lies Lake Sinoe. When and how it formed is uncertain. One dating puts the closure around 550–650, when the front of the new Danube lobe reached Histria. Another puts it much later: the youngest ridges under the lake formed around 1170–1230, and the lake is dated to around 1250.
The mechanism is uncertain too. An older school says the Chituc barrier closed a gulf. The Bucharest team says the lake appeared when the ridge plain sank. The ground is still sinking, about 4 millimetres a year. While the closure is uncertain, the lake water is hatched.
In a straight line, from the acropolis to the nearest open sea is 7.6 kilometres on our model. The researchers say "8 km inland".
Now the process runs backwards. Dams on the Danube have cut the river's sediment from about 50 million tonnes a year to 25–35, and the barrier between the lagoons and the sea has begun to erode.
Miletus, the city that sent the settlers, fared no better: its gulf filled with the silt of the Maeander, and from the Middle Ages it had no direct way to the sea. Ostia, the port of Rome, now lies about 4 kilometres from the coast.
Pick a year and turn the coast to see it from any side.
630 BC: the sea is at the walls
At its fastest, in the 11th to 13th centuries, the shore at Histria advanced 10–15 metres a year. Type the year you were born to see what that would have meant for you.
The old ridge west of the city grew around 1380–780 BC at 1.3 metres a year, ten times more slowly.
| Question | One side says | The other says |
|---|---|---|
| When the land in front of the city grew | AD 650–1230 (Vespremeanu-Stroe and colleagues, luminescence) | AD 450–1350 for the youngest ridge set (Giosan and colleagues, 2006) |
| When Lake Sinoe closed | around 550–650, with the Dunavăț lobe | around 1250: the youngest ridges under the lake formed around 1170–1230 |
| How the lake formed | the Chituc barrier closed a gulf (Bleahu 1963, Panin) | the ridge plain sank (the Bucharest team) |
| Sea level | stable for 5,000 years, within +1.5 and −2 m (Giosan, Brückner, Bivolaru) | several metres lower in antiquity, the "Phanagorian regression" (Chepalyga, Panin) |
On the 3D model the sea stands at today's level in every year, following the first side of the last row. Histria's Greek layers now lie 4–5 metres deep. That needs no lower sea: slow sinking of the ground is enough.
The ruins stand on the shore of Lake Sinoe, next to the village of Istria, about 43 kilometres north of Constanța as the crow flies. On site you can see enclosure walls from every period, Archaic to Late Roman, four temples, altars, rich houses on the acropolis, two bath buildings and six early Christian basilicas. The museum beside the site holds pottery, coins, statuettes and inscriptions.
Opening hours and tickets change with the season. Check them on the Museum of National History and Archaeology Constanța page.
The terrain is the Copernicus GLO-30 model, one point every 30 metres, with no vertical exaggeration. Today's shore, the lakes and the walls come from OpenStreetMap.
The new land is not dated point by point. We placed the 630 BC shoreline from figure 3 of Bivolaru and colleagues (2021) on the ground and ordered the plain in front of it the way the dated ridges in the same figure run: older inland and to the north-east, younger seaward and to the south-west. The two published datings are stretched over that order. The figure is schematic, so we kept only its order, not its distances.
The area north of the city, the Lupilor barrier and Lake Zmeica, is drawn as it is today, in every year. The distance to the sea is measured in a straight line every 25 years, for both ends of the range.
All sources, with quotations and the status of each, are in the text version of this page.
The first version of this page, from September 2026, said the harbour silted up from Hellenistic times, "while Polybius watched". The dated ridges say otherwise: the land that cut the city off from the sea formed after AD 650 (after 450, by another dating), a ridge kept its access into the 6th century, and inscriptions mention the harbour into the 2nd century AD. Polybius wrote about the Black Sea and the Danube bar in general, not about Histria. The word "tombolo" is gone too.