A Wreck Found Where Empires Met

In 2014, underwater archaeologists from the Croatian Conservation Institute identified the remains of a Byzantine-period ship at Tatinica, off the island of Mljet in southern Croatia. The site lies on a sandy slope between rocks at a depth of roughly 32 to 40 metres, and the excavation campaign continued seasonally from 2015 to 2025 under the direction of Croatian specialists including Igor Miholjek and Pavle Dugonjić. 

This was not just another wreck in a sea full of them. Mljet lies on one of the principal maritime corridors of the Adriatic, between the Byzantine east and the Italian coast. National Geographic’s account of the excavation notes that the seabed around Mljet has produced more than two dozen wrecks from different periods, but Byzantine wrecks are exceptionally rare there; Miholjek had previously encountered only one other Byzantine wreck off the island, and that one was much later, from the eleventh century. 

The rarity matters because the ship belongs to a period that historians once described, too readily, as a low point in Adriatic exchange. Roff Smith’s National Geographic feature captured the older assumption clearly: some historians had seen the seventh- and eighth-century Adriatic as an impoverished backwater, almost empty of shipping. More recent scholarship has complicated that picture. Richard Hodges, drawing on the archaeology of Butrint, argues for a “measured transformation” of the Adriatic rather than a simple collapse, while Francesco Borri has explicitly suggested that around 700 the region experienced an adjustment in its relationship with Constantinople, not a complete rupture in seaborne communication. 

That broader scholarly revision gives the Mljet wreck its real weight. Vesna Zmaić Kralj has argued that one of Byzantium’s important merchant routes ran along the eastern Adriatic, linking Constantinople and the wider Mediterranean to Ravenna, Comacchio, and Venice. At the same time, she notes that Byzantine amphora finds are scarce in areas associated with the Croatian duchy and tend to appear in elite contexts, which already hints at a world in which exchange survived unevenly and often at a high social level. The Mljet wreck fits that pattern uncannily well. 

What Lay on the Seabed

The finds are striking because they combine the ordinary material of seafaring with an assemblage that is anything but ordinary. According to the official Croatian Conservation Institute summary, the ship carried several types of Byzantine amphorae, rich trade goods, commercial equipment, several types of bronze scales, a large scale and a weight shaped as the goddess Athena, lead ingots, bronze and ceramic vessels, gaming counters and an ivory die, as well as parts of the ship’s wooden structure. National Geographic adds millstones, crockery, fishing tackle, and game pieces for backgammon or draughts. 

The vessel itself survives only in fragments. Even so, the excavators were able to identify waterlogged remains of the keel, ribs, and planking, enough to judge it a medium-sized cargo vessel around 60 feet long with a carrying capacity of about 60 tons. Around a hundred amphorae and other ceramic containers were recorded, though that total appears to be far lighter than the cargo one would expect on a ship of that size. That imbalance is one reason the wreck has prompted so much discussion. 

The excavation was painfully slow. The wreck lies deep enough, and on difficult enough ground, that divers could only work for a little over half an hour a day, and only once daily, using oxygen-rich gas mixtures to extend safe bottom time. Money was tight, so work could be done only for a few weeks each summer. That is why a discovery made in 2014 matured into a ten-year excavation story, with interpretation advancing season by season rather than all at once. 

The date of the wreck rests on several strands of evidence. The official Croatian Conservation Institute statement places the coin finds at the end of the seventh century or the very beginning of the eighth. National Geographic reports that the combination of coins, amphora types, and workmanship has allowed researchers to narrow the sinking to the first half of the eighth century, while Reuters summarised the current consensus more broadly as seventh or eighth century. In other words, the chronology is narrow enough to be historically significant, but not yet so precise that every interpretation can be fixed to a single decade. 

Gold, rank, and the diplomatic question

Gold solidus of Heraclius and Heraclius Constantine – Solidus of Heraclius and Heraclius Constantine, 629–631. Metropolitan Museum of Art, CC0.

The finds that transformed the site from an important wreck into a major historical problem were the gold objects. During the third diving season, Dugonjić uncovered an elaborate gold buckle decorated with birds pecking grapes from a cornucopia. Over subsequent seasons, the team recovered four gold belt sets, a gem-set gold buckle decorated with pearls, emeralds, and rubies, a gold signet ring probably depicting Emperor Heraclius, and fourteen gold solidi from the Constantinople mint. Reuters reported that the total recovered gold weighs more than 600 grams, which Miholjek described as the largest quantity yet found on a Mediterranean shipwreck. 

What do those objects mean? At the simplest level, they indicate rank. Pavle Dugonjić told Reuters that the signet ring is not the sort of thing “just anyone” would have worn, and National Geographic’s reporting follows the same line: this was probably not an ordinary merchantman crewed only by traders and sailors. The site instead suggests the presence of a wealthy, well-connected passenger, or perhaps a small party attached to him. 

“Golden belt sets, even the signet ring of Heraclius, are precisely the sorts of gifts a Byzantine emperor would present.”
— Peter Sarris, quoted in National Geographic, 2026. 

That observation by Peter Sarris points towards the most discussed interpretation of the wreck: a diplomatic voyage bearing imperial gifts. National Geographic reports that some of the belt sets seem to have been old even when the ship sailed, spanning much of the seventh century, which makes the cargo harder to read as a simple batch of newly made luxury goods. Falko Daim has also noted strong parallels between one of the Mljet sets and one from an Avar grave, suggesting a shared workshop and a wider political-gift culture across the region. Even the presence of scales, weights, crucibles, and mercury raises further possibilities: were some items being transported for remelting, refashioning, or valuation as part of a mission rather than straightforward trade? 

That said, the diplomatic reading remains a hypothesis, not a verdict. Miholjek has also considered the simpler possibility that the gold belonged to a wealthy traveller aboard a cargo ship. The material evidence does not yet force a choice between those explanations. What it does rule out, or at least weaken sharply, is the image of a routine small-scale coaster carrying nothing but staple goods between neighbouring ports. 

Why the Mljet wreck matters

The wider historical setting makes the wreck more than a local curiosity. National Geographic situates the sinking in the aftermath of plague, famine, long war with Persia, and the rapid expansion of Islamic rule into the Middle East, North Africa, and beyond. By around 700, Byzantium had lost some of its richest provinces, and the old certainties of late Roman Mediterranean exchange had been badly shaken. Written evidence for the Adriatic in this period is sparse, and, as Justin Leidwanger observed, much of the surviving text was written by clergy inclined to see the age in bleak terms. 

That is exactly why archaeology matters here. A wreck does not preserve rhetoric; it preserves a journey interrupted. The amphorae, coins, gaming pieces, metalworking tools, and regalia all belonged to one voyage, one moment, one practical movement through the sea. Leidwanger put the point starkly when discussing the gold belt fittings: until the Mljet wreck, such objects were known almost entirely from graves, not from transport contexts. The site therefore lets historians see elite artefacts not as static burial possessions but as things in motion. 

“For me, it’s the most important shipwreck of its period.”
— Justin Leidwanger, quoted by Reuters, 2026. 

That may sound like excavation-season exuberance, but it rests on a serious historical argument. Preliminary study of the amphorae suggests links not just with the Adriatic but with southern Italy and northern Africa. If that reading holds, then the wreck potentially bears on one of the hardest questions in seventh- and eighth-century Mediterranean history: how far Byzantine maritime contact across the post-conquest sea still endured. This does not mean that older narratives of contraction were wholly wrong. It means that contraction and connectivity may have existed together, unevenly, and that elite travel or diplomacy could survive where bulk commerce had thinned. 

The likely route also matters. The wreck’s position on the south-eastern approach to the harbour of Polače suggests that the ship was moving north up the Adriatic when it sank. Miholjek thinks the captain was trying to reach the shelter of Polače, perhaps driven by the bora, the violent katabatic wind that can descend suddenly from the mountains and make navigation dangerous. Meteorological research on bora events describes gusts exceeding 40 metres per second and notes their hazard to coastal shipping; National Geographic’s description of the wind’s effect around Mljet fits that wider scientific picture. 

What cannot yet be known

The most responsible way to write about the Mljet wreck is to keep its unanswered questions visible. We still do not know the ship’s home port. National Geographic suggests Constantinople as a probable starting point, and Ravenna as a possible destination, but both remain informed reconstructions rather than demonstrated facts. Nor do we know whether the gold was intended as a diplomatic gift, personal property, portable wealth, or stock for refashioning. 

We also need caution about the cargo itself. Amphorae point towards a commercial function, but the number found so far seems small for a vessel of this size. That could mean the ship’s main purpose was something other than transporting bulk staples. It could also reflect what has and has not survived, what remains buried, or what once shifted downslope before excavation began. Underwater archaeology is exacting, but it is rarely complete. 

What the wreck has already done is more than enough. It offers hard evidence that the early medieval Adriatic was not a dead zone. It was dangerous, selective, and politically charged, but still used. The people aboard this ship carried scales, amphorae, coins, gaming pieces, and high-status goldwork through waters that later writers left largely unrecorded. That does not abolish the crisis of the seventh and eighth centuries. It shows, instead, that even in crisis, the sea went on binding places together.