Harbour defence-works using chains in a “limen kleistos“ could be used both to stop the enemy from entering the port and to trap the enemy once inside the port, as mentioned by Dio Cassius (Hist, 51, 9) at Paraetonium (Egypt):
“Gallus, it seems, caused chains to be stretched at night across the mouth of the harbour under water, and then took no measures openly to guard against his opponents but contemptuously allowed them to sail in with perfect immunity. When they were inside, however, he drew up the chains by means of machines, and encompassing their ships on all sides – from the land, from the houses, and from the sea – he burned some and sank others.” (translation Lacus Curtius).
Another story is also told by Dio Cassius (Hist, 12, Frag.) at the port of Hippo Diarrhytos:
“The natives put chains across the mouth of the harbour, and the invaders found themselves in an awkward situation, but escaped by cleverness and good fortune. They made a quick dash at the chains, and just as the beaks of the ships were about to catch in them, the members of the crews moved back to the stern, and so the prows were lightened and cleared the chains; and again, when all rushed into the prows, the sterns of the vessels were lifted high into the air. Thus, they effected their escape […]”. Note that as they “escaped”, they were trapped inside the port.
Ancient authors mention least 9 harbours with chains at the entrance:
- Syracusa, Sicily, in the 3rd c. BC (Frontinus, Strategemata, 1, 5),
- Byzantion-Bosphorion, in the 2nd c. AD (Zonaras, Constantin, 120, citing Dio Cassius),
- Byzantion-Kynegoi, in the 2nd c. AD (Zonaras, Constantin, 120, citing Dio Cassius),
- Chalkedon, near Istanbul, in the 1st c. BC (Appian, Mithridatic, 10, 71),
- Andriake, near Antalya, in the 1st c. BC (Appian, Civil wars, 4, 10, 82),
- Alexandria Portus Magnus (3 ports), in the 1st c. AD (Lucan, Pharsale, 10, 57),
- Paretonium, Marsa Matruh in Egypt, in the 1st c. BC (Dio Cassius, Hist., 51, 9),
- Carthage, in the 2nd c. BC (Appian, Libyca, 96) and in the 6th c. AD (Procopius, War against Vandals, 1, 20),
- Hippo Diarrhytos, Bizerte in Tunisia, in the 3rd c. BC (Dio Cassius, Hist., 12, fragments reported by Zonaras, 8, 16).
Chains strechting across a harbour entrance are mentioned by Vitruvius (Arch, 5, 12): “erect a tower on each side, wherefrom chains may be suspended across by means of machinery” (translation Lacus Curtius). A system closing a harbour entrance is also mentioned by Philo of Byzantion, in the 3rd c. BC, (Poliorcetica, 3, 29), but he does not explicitely mention chains and as the surviving text is incomplete, the translation is debated.
Archaeology has shown that chains were most probably also installed at the entrance of Phalasarna (Hadjidaki, 2019) and possibly at Myndos (Dumankaya, 2015). In addition, Halieis may have been a “limen kleistos” with doors that could be closed (Jameson, 1969, contradicted by Frank Frost, 1985). Many harbours (over 75) are known or suspected to be “kleistos” but it is not known if chains were used (Lehmann-Hartleben, 1923: 65-74, Mauro, 2020).
A very particular use of a chain was made by Polycrates when he symbolically linked Delos island to Rhenea island with a chain (Thucydides, Pelop. wars, 3, 104). Using Remmatia island located between Delos and Rheneia, the length of this chain must have been at least 425 m (250 + 175 m) and it must have been placed on the sea bed as it seems unlikely that it could be tense because of the large force this would require.
In order to install a chain (or doors) to close the entrance of a harbour, the width has to be limited. Except for Motya with an entrance width of 5 m (but this place is not considered any more as a military harbour), the smallest entrance widths known are at Lechaion and Phalasarna (around 10 m). Other narrow entrances range between 10 and 30 m (Naupaktos, Aegina, Halieis, Amathous, Andros, Knidos, Phaselis, Stratonos Pyrgos, Leuke Akte, Apollonia, Gummi, Carthage) and up to 75 m, as far as we can see from today’s remains (Kantharos, Munychia, Larymna, Thasos, Chalkedon, Elaia, Kos, Rhodos, Kydonia, Paphos, Seleucia Pieria). Remains seen on Google Earth seem to show some entrances around 100 m wide (Corcyra, Anaktorion, Oiniadae, Zea, Mytilene, Miletos, Kaunos, Kyrenia). Larger entrances may possibly also have had a closing chain (Myndos: 117 m acc. to Dumankaya, 2015; Golden Horn: 300 m acc. to Kastenellos, 2017).
Considering a unit weight of 25 kg per meter for a chain with 10 cm shackles, a length of 10 m (250 kg) is not a problem to be lifted by capstans. A 40 m-chain weights around one ton and can also be lifted, but a sag will be generated and it is interesting to know how much this sag is depending on the traction force.
The mathematical formulation was written down by Leibniz in 1691, after some discussions between people like Galileo, Bernoulli and Huygens, just to say here that it is not an easy matter. Anyhow, this formulation can now be used to compute the horizontal force required on a chain stretching between the lateral banks of a canal. If the chain is fastened 3 m above the water level with a sag of 2.5 m, it will hang at least half a meter above water, meaning that a trireme cannot pass over or under it.
On a canal 40 m wide and 3 m deep, the chain looks as follows according to Leibniz’s famous “catenary equation”:
The computation shows that the required chain length is 40.4 m, just a little more than the canal width. The required horizontal traction force is about 2 tons, i.e. about twice the mass of the chain, on each side of the canal. This should not be a problem with Roman capstans located on both lateral quays of the canal at 3 m above the water level.
A smaller traction force of 1 ton would induce a longer chain with a larger sag. With a sag of 5 m, the height of fastening the chain on the banks would increase from 3 m to 5.5 m, requiring towers on both lateral quays of the canal. A wider canal, e.g. 150 m, would require a heavier chain (3.75 tons) and more traction from both banks (14 tons) in order to have a 5 m sag.
Such towers are perhaps among those meant by Vitruvius (Arch., 10, 1) when he writes about “innumerable different machines, which it is unnecessary here to discuss, since they are so well known from our daily use of them, such as wheels generally, the blacksmith’s bellows, chariots, calêches, lathes, and other things which our habits constantly require.” (translation Lacus Curtius), which implies that machines were frequently installed in/on towers/lathes.
Our computations show that a chain can be stretched between both sides of a canal by means of a traction force not exceeding 10-15 tons, which may be considered feasible with Roman equipment like capstans and treadwheels.
It has been suggested that the chain closing a harbour entrance would need to be supported by floating pontoons (Diels, 1920). Although so-called “booms” have been used in the Middle Ages with wider entrances of around 300 m, our computations do not confirm the need for such an arrangement for entrances smaller than 100 to 150 m.
BLACKMAN, D. & RANKOV, B. et al., 2013, “Shipsheds of the Ancient Mediterranean”, Cambridge University Press, (617 p).
DIELS, H., 1920, “Antike Technik. Sieben Vorträge”, erweiterte Auflage, Teubner, Leipzig und Berlin, (243 p).
DUMANKAYA, O., 2015, “East-harbour mole at Myndos”, TINA, Sayi, N°3, (p 12-45).
FROST, F., 1985, “The “Harbour” at Halieis”, Harbour Archaeology, BAR International Series, 257, Oxford, (p 63-66).
HADJIDAKI, E., 2019, “Three Decades of Adventures with Honor Frost in Crete”, in: Blue, L. (ed.), 2019, In the Footsteps of Honor Frost. The life and legacy of a pioneer in maritime archaeology, Leiden: Sidestone Press (p 165-182).
JAMESON, M., 1969, “Excavations at Porto Cheli and Vicinity, Preliminary report, I: Halieis, 1962-1968”, Hesperia 38, (p 311-342).
KASTENELLOS, P., 2017, “The Chain Across The Golden Horn“, Apuleius Books.
LEHMANN-HARTLEBEN, K., 1923, “Die Antiken Hafenanlagen des Mittelmeeres”, (Klio, suppl. 14), Leipzig, (305 p).
MAURO, C., & GAMBASH, G., 2020, “The earliest “limenes kleistoi” a comparison between archaeological-geological data and the periplus of pseudo-skylax”, REA, T. 122, 2020, n°1, (p. 55-84).
 Closable harbour with a narrow entrance, Lehmann-Hartleben, 1923, p 65-74, Blackman, 2013, and Mauro, 2020.
 A chain of around 300 m seems to have been installed in the 5th c. near Yeralti mosque at the Golden Horn entrance.
List of “limenes kleistoi“
Lehmann-Hartleben was the first to provide a list of (42) “limenes kleistoi” in 1923. We now have over 75 of them (known “X” or suspected “X?”, some with a chain “X”) and with varying entrance-channel widths ranging from 5 m to 120 m.
|Ancient Name||Modern Name||LK||CH||Width|
|Centumcellae||Darsena Romana in the port of Civitavecchia||X?||?||?|
|Syracuse, Syrakus, « Small Port »||Porto Lachio, Syracuse||X||?||?|
|Motye, Motya||Mozia, on the isle of San Pantaleo||X?||?||5|
|Port of Ambrakia||Phidokastro||X||?||?|
|Anaktorion, Anactorium||near Nea Kamarina||X?||?||100?|
|Oiniadae||Katoxi, Trikardo, now inland||X?||?||100?|
|Nisa, Nisaea||Roman fort at Agios Nicolas, near Megara||X?||?||?|
|Aigina, Aegina||Roman naval base South of Kolonna hill||X?||?||30|
|Halai, Halieis, Halia||Portocheli||X?||door||20|
|port of the Pheacians, naval base of Alkinoos, Corcyra||Ormos Garitsa, Kokotou district on Corfu||X?||?||100?|
|Palaiopolis||Paleopolis, on the isle of Andros||X?||?||20?|
|Paros||Paros, Paroikia Bay||X||?||?|
|Samothraca||Paleopoli, on the isle of Samothrace||X?||?||?|
|Mytilene, naval base||on South side of Mytilini, on the isle of Lesbos||X||?||100?|
|Pythagoreion||Pythagoreio, on the isle of Samos||X||?||?|
|Kos, Cos||Naval base at Mandraki harbour, Kos, on the isle of Kos||X||?||75?|
|Rhodos||Naval base at Port Mandraki||X||?||50|
|Byzantion, Prosphorion||Marmaray Sirkeci railway station, in the Golden Horn||X||?||?|
|Byzantion, Kynegoi||Balat, Fener district, near Ferruh mosque, in the Golden Horn||X||?||?|
|Genesentis, Boona||Persembe in the bay of Vona||X||?||?|
|Chalkedon||Kadiköy in front of Istanbul, on R Kurbagalidere||X||X||50?|
|Cyzicos, home port of Classis Pontica fleet||on the isthmus of the peninsula of Erdek||X||?||?|
|Falasarna, Phalasarna||Falasarna, Phalasarna, on the isle of Crete||X||X?||10?|
|Cydonie, Kydonia||Khania, Chania, on the isle of Crete||X||?||75?|
|Salamis, Salamine of Cyprus||7 km North of Famagousta, on the isle of Cyprus||X||?||?|
|Kition||Larnaca, slipways at Bamboula, on the isle of Cyprus||X||?||?|
|Amathus, Amathonte||10 km East of Limassol, on the isle of Cyprus||X?||?||21|
|Nea-Paphos||Kato Paphos, on the isle of Cyprus||X||?||55|
|Soloi, Soli||Potamos tou Kambou, on the isle of Cyprus||X||?||?|
|Kyreneia||Kyrenia, with an undated chain closure, on the isle of Cyprus||X||X||100|
|Elaia, port of Pergamon, Pergamos||Naval base and commercial port at Kazikbaglar||X?||?||60?|
|Canae, Kanai, Kane Prom.||Karadag, near Bademli, island now connected to mainland||X||?||?|
|Elaia, port of Pergamon, Pergamos||Naval base and commercial port at Kazikbaglar||X?||?||60?|
|Palaia Smyrna||Bayrakli, Izmir||X||?||?|
|Klazomenai||Liman Tepe, near Urla Iskele||X?||?||?|
|Erythrai||Ildir, in front of Chios||X?||?||?|
|Hellenistic port of Ephesos||West side of Panayirdag hill, near Selcuk||X||?||?|
|Priene||Güllübahce on the silted up outlet of R Meander||X||?||?|
|Miletos main port, Dokimos Harbour||Milet, Lion Harbour||X?||?||100?|
|Myndus, Myndos||Gümüslük “East Harbour”, Dogu Limani||X?||?||117|
|Halicarnassus||Bodrum, small naval base inside modern marina?||X||?||?|
|Halicarnassus, Portus Secretus?||Bodrum, South of fort||X?||?||?|
|Knidos, Cnidus, naval base, ancient Triopion||Cnide West, former isle of Triopion now connected to mainland||X||?||25|
|Kaunos, Caunus||Sülüklü Gölü, near Dalyan||X||?||100?|
|Andriace, Andriake, port of Myra||Andraki, near Demre||X||?||?|
|Phaselis, Phaselide||near Tekirova||X?||?||18|
|Attaleia, port of Perge||Antalya||X?||?||?|
|Seleucia Pieria, home port of Classis Syriaca fleet||Cevlik, port of Antioch of Daphne, inner harbour at the toe of the hill||X||?||60|
|Tyre, “Sidonian port”||Sour, North port||X||?||?|
|Port of Aksaph, port of Megiddo||Tell abu Hawam, Haifa, near R Qishon||X?||?||?|
|Stratonos Pyrgos||Caesarea Maritima, pre-Herodian port||X||?||20?|
|Iotabe? Ezion Geber?||Geziret Faraun, Pharaoh’s island, Coral island||X?||?||?|
|Asabon||Jazirat al-Ghanam, Cape Musandam, Mussendom||X?||?||?|
|Ur, Uri, Sumerian Urim, North Port||Tell el-Muqayyar||X||?||?|
|Alexandria, Portus Magnus, home port of Classis Alexandrina fleet||Alexandria, includes 3 now submerged ports: near Palace (7 ha), near Antirhodos (16 ha), and inbetween (13 ha)||X||X||100?|
|Kibotos||port located inside the Port of Eunostos||X?||?||?|
|Leuce, Leuke Akte||Ras el-Kanayis, Ras Kanaïs||X?||?||15|
|Paraetonium||Marsa Matrouh, Matruh, Bates’ island, Geziret el-Yehudiyeh||X||X||?|
|Chersis, Xherson, Aphrodisias insula||el-Kerchi, Kersa, Chersa islets 15 km NW of Derna||X?||?||?|
|Apollonia, port of Cyrene||Susah, Soussa||X?||?||20|
|Ruspina||Monastir, islet La Tonnara (el-Ghedamsi islet) is probably just a quarry||X?||?||?|
|Hadrumete||Sousse, initial port near the North wall of city||X?||?||?|
|Carthago||Carthago, rectangular commercial port||X||X||21|
|Mandrace, Mandracium||Cothon of Carthage: circular naval base||X||?||?|
|Akra, Acra insula||excavated basin on the isle of Rachgoun in front of Takembrit||X?||?||?|