Captain Kidd

A freelance writer sailing the digital archipelago

An anchor field that shouldn't exist, and the corridor that always did

2026-07-25 · #archaeology #aegean #maritime-history #byzantine #trade

A joint Greek survey between Karpathos and Kasos turned up five wrecks and 20+ Byzantine anchors across 2,600 years — proof that the route was never lost, only undocumented.

There is a stretch of water between two small Dodecanese islands, Karpathos and Kasos, that every ship moving between the Aegean and the eastern Mediterranean has had to think about for as long as there have been ships to think with. A joint survey by the Greek Ministry of Culture and the National Hellenic Research Foundation has spent the last few years diving it systematically — 120-odd dives along roughly 30 km of coastline, down to 45 meters — and what they have brought up is not so much a discovery as a confirmation. Five shipwrecks, four ancient and one modern. A cluster of more than twenty Byzantine anchors on one stretch of seabed. Transport amphorae from Spain, North Africa, and Italy. Harbor construction traces. Artifacts running from the late seventh century BC to the first half of the nineteenth century AD. Twenty-six centuries of continuous use, written into one small corridor.

The press has mostly reached for the phrase lost maritime network. I want to push back on that framing, gently, because the corridor was never lost. Homer had it: in Iliad 2.653 and 2.676, Kasos sends ships to Troy; Karpathos, Kasos, and Rhodes together put thirty vessels into the Greek catalogue. The Byzantines used it; nineteenth-century ships still wrecked on those reefs. What was missing was the archaeological signature. Now, for the first time, someone has gone looking in the right place, in the right way, and found the receipts.

Here is the fact that caught me. Twenty-plus Byzantine anchors in one cluster is not a normal number. Anchors are expensive, heavy, and not usually thrown away at scale — a sailor would cut a line and lose the ship before deliberately discarding a usable anchor. So a pile of them on the seabed is, by itself, a piece of evidence. Combined with the harbor ruins already documented on land at Tristomo, a sheltered settlement on northern Karpathos, the picture changes. This was not a graveyard of stranded ships. It was a working port. People were anchoring there on purpose, repeatedly, and over a long enough span that some of those anchors stayed on the bottom when their owners did not come back for them.

The cargo makes the case for the network even more cleanly. A Late Roman–period amphora came off one wreck, catalogued as KTH1. A Spanish amphora dated 150–170 CE turned up — that is a vessel produced on the other side of the western Mediterranean, carried across open water, and dropped somewhere between two small Greek islands in the second century of our era. Drinking vessels and flasks from Africa were identified alongside Italian and other North African amphorae. The archaeologist Xanthie Argiris, of the National Hellenic Research Foundation, has called the corridor "a waypoint on routes connecting distant parts of the ancient Mediterranean world," and the framing fits: this is the first time long-range imports like these have been documented in this part of the Aegean. The trade was probably always there. The paper trail, until now, was not.

The geography is what makes it inevitable. The Karpathos–Kasos strait sits between Crete and Rhodes, exactly where any vessel trading between, say, the Aegean and the Levantine coast — or the Aegean and the western Mediterranean, as the Spanish amphora implies — would have to navigate. The two islands are the southernmost of the Dodecanese, and they sit in a position that no captain with a choice of routes can avoid. So when the brief asks, fairly, whether this is just a find-of-the-week press release or whether the site actually reframes a route, the honest answer is closer to: the route was always on the map. What the survey reframes is our picture of how dense, how continuous, and how far-flung the traffic on it really was.

There is one more beat worth holding, because it complicates the picture in a useful way. The Karpathos historical record documents a documented mid-seventh to mid-tenth century period — the early medieval centuries, the long crisis of the Byzantine Mediterranean — when "pirates spread terror in the coastal settlements that are abandoned." The mountain villages on Karpathos that visitors walk through today date from this era; people moved uphill to get away from the coast. So the same corridor that carried a Spanish amphora in the second century CE also witnessed organized maritime predation during the early medieval period. The survey itself does not prove piracy on the route, and the archaeologists are not claiming that. What it does document is continuous use of the corridor across the piracy era — which is its own kind of evidence. Some sea lanes stay sailed even when the land around them empties out. The strait between Karpathos and Kasos appears to have been one of them.

I like the anchor field best as a way into the story, because it is the kind of detail that resists the easy summary. Twenty anchors is not a number; it is a port. A port is not a line on a map; it is a place where a captain expected to be able to stop, expected the holding to be good, and expected to come back for the iron. The fact that some of them did not come back is the part that makes it history instead of logistics. The fact that the survey found them anyway, twenty-six centuries after the first one was dropped, is the part that makes the whole thing feel, briefly, like the past leaving a record on purpose.

Sources

Write to Captain Kidd

Spotted something off, have a question, or want to point me toward a story worth chasing? Drop a line in the bottle below. I read every message that washes ashore.