Ship bridges over the Hellespont

from Wikipedia, the free encyclopedia

The ship bridges over the Hellespont were two floating bridges that the Persian King Xerxes I built in 480 BC. On the occasion of his campaign against Greece , he had it built in order to get with the largest army of the time from Asia via the Hellespont (today's Dardanelles ) to Thrace, which was also controlled by Persians at the time (in today's European part of Turkey, see Eastern Thrace ).

While the bridge construction reported by the Greek historian Herodotus in his histories is generally accepted as such, there are many doubts about the details he presented in detail.

The bridges in Herodotus' histories

Xerxes I. has the Hellespont flogged (illustration from 1909)

Herodotus describes that the campaign planned by Darius I was prepared for years. He mentions that Xerxes also ordered the provision of ships for the bridges and the manufacture of ropes from papyrus and white flax .

While Xerxes was marching with his army from Sardis to Abydos , at that time an important port city on the Hellespont , two bridges were built from there to the opposite coastal promontory at Sestos , one by the Phoenicians with ropes made of flax and one by the Egyptians with ropes made of papyrus. Herodotus gives the length of the bridges as seven stadia, which would correspond to about 1300 m.

However, both bridges were destroyed by a storm immediately after their completion. Xerxes was so angry that he ordered the Hellespont to be punished with 300 lashes, ankle shackles to be thrown into the sea and the two builders of the bridge to be beheaded.

Under new site managers, two bridges from Pentekonteren and Trier were built, the warships of that time, one of 360 ships and one further south of 314 ships, which were held in position by very large anchors. Three openings have been left free for smaller ships to pass through. The large ropes were laid over the ships and fastened. This time two flax ropes and four papyrus ropes were used for each bridge and tensioned with winches . The flat cables have a talent per cubit weighed and had thus been significantly heavier than the papyrus rush. Wooden planks were laid across the ropes and covered with brushwood and finally with pounded earth, so that a kind of road was created. Screen panels were attached on both sides so that the horses did not shy away from seeing the water.

The army with foot soldiers and riders crossed the Hellespont on the northern bridge, the train with pack animals and auxiliary personnel used the southern bridge. The crossing took a total of seven days and seven nights.

According to the story of Herodotus, Xerxes left the bridge standing: after the failure of his campaign, he feared that the Ionians or the Greeks might have destroyed the bridge. The Greeks, for their part, discussed whether they should destroy the bridge. However, when part of the Persian army retreated to the Hellespont, they found only the remains of the bridges destroyed by the storm. Xerxes managed to return to Asia by ship despite a storm.


While Herodotus hardly gives the location or any details for the ship bridge over the Bosporus built by Dareios I a good 30 years earlier , the wealth of details in the Hellespont bridges, which give a clear picture of the bridge with only superficial reading, is astonishing. On closer inspection, almost every detail of the bridges is the subject of questions, discussions and doubts, even if Herodotus' image of the bridges is also given without reflection on the Internet. His depiction can therefore be seen less as a sober, technically correct description of the bridges than as a decoration of the greatness of Xerxes I, which makes the Greek victories against him all the more significant.


The current on the surface in the direction of the Mediterranean Sea is given with an average of 1 ½ knots, but is very dependent on the wind. Also caused by the wind, the water level can rise up to 60 cm. At the bottom, an undercurrent flows in the opposite direction. Current eddies and shallows often occur especially at headlands .

The presently narrowest point of the Dardanellen between Çanakkale and Kilitbahir ( 40 ° 8 '40.9 "  N , 26 ° 23' 50.3"  O ) is about 1.4 km wide and up to 91 m deep, but also has the strongest current and is considered the most difficult part of the strait in seafaring. Çanakkale was only founded in Ottoman times on the gravel delta of a river that came from the mountains and often rushed in winter (which is now tamed by the Atikhisar dam). Around 2500 years ago, the gravel delta may not have advanced that far into the strait.

The Abydos said Herodotus lay on the Asian shore north of Çanakkale near the projecting westward in the strait Cape Nara (Nara Burnu, formerly Nagara Burnu) ( 40 ° 11 '46.6 "  N , 26 ° 24' 8"  O ). The water depth is shallow in the south and west of the cape, but drops to 103 m towards the center. The current runs in the middle of the strait at more than 2 kn to the southwest, but forms large countercurrents around Cape Nara.

Place of the two bridges

The location of the bridges between Abydos and the opposite bank at Sestos, indicated by Herodotus, is accepted by many historians. The British General Frederick Maurice , who visited the area in 1922, considered, for military reasons, only a beach further north as a suitable place for the construction of the floating bridges, which would then have been more than three kilometers long. The narrowest point in front of Çanakkale today, on the other hand, should not come into question, as the then still wild mountain river could form all rousing floods in a very short time.

Two bridges were necessary because the march of the huge army on the narrow streets of the Chersonese stretched very far and the entourage had to advance in parallel so as not to interrupt the supply of the troops with food, fodder and water.


Herodotus clearly writes that penteconters and triremes, i.e. only warships, were used as the floating bodies of the bridge. That seems to be universally accepted. But it makes little sense to use warships for a task that simpler and cheaper merchant ships would have been better suited to because of their wider profile, lower center of gravity and higher freeboard . In a trireme, the lower openings for the oars were only around 30 cm above the waterline and were therefore provided with leather cuffs, which is rather unsuitable for the pontoon of a floating bridge.

The ships should also all have the same height to ensure a level bridge deck. In addition, scaffolding in the ship was useful to compensate for differences in height and to carry the loads and divert them to the center of the ship so that the ship's walls, which are unsuitable for high loads, are not damaged. Even if the use of merchant ships is not discussed and therefore disregarded, the bridges could have consisted almost entirely of penteconters and, because of their greater height, triremes could only have been used on both sides of the passages.


On floating bridges on rivers, the ships are usually held in their position by anchors at the bow and stern, which is why Herodotus' description of the anchors initially seems credible. However, Herodotus did not even mention the depth of the Strait, although it was mentioned by the more recent historians, but apparently nowhere was discussed as a problem.

Anchor lines must be several times the length of the water to prevent the anchors from being pulled out or damage to the ship. The ships in the middle of the Hellespont would then have had to have anchor lines several hundred meters long at the bow and stern, so that for the allegedly 674 ships not only 1,348 heavy anchors, but also around three hundred kilometers of anchor lines would have been required. It is quite questionable whether such quantities could even be produced in the relatively short time at that time. On top of that, anchor lines that are so long cannot prevent the ships from swaying and colliding , especially if there are vortices and the anchor lines of the ships lying close to one another get entangled. After all, it should not be possible to place the anchors, whether they were stone baskets or even iron stick anchors, with their extra-long lines so that the ships lie exactly in one row.

If one proceeds from the anchoring of the ships described by Herodotus, it must be taken into account that every bridge with its anchor lines would have required a lateral strip up to 900 m wide. But then there would hardly have been enough space for two bridges on the bank near Abydos.

In addition, securing the ships with anchor lines as well as ropes extending from bank to bank would only be of theoretical advantage if the anchor and the ropes are coordinated in such a way that exactly the same tensile force acts on both, which is not possible in practice , especially not under the influence of changing winds, currents and counter currents. But then the entire train only bears on one of the two elements, the other contributes nothing to the safety.

It must therefore be assumed that the ships were only held in position by the long ropes and that anchors were only used temporarily in the shallower areas to hold the ships until the ropes were laid and attached to the ships.

Length of the bridges

The length of seven stadia or about 1300 m given by Herodotus is in any case too short.

At Abydos the direct distance between the banks is about 2000 m. However, even this distance of 2000 m cannot correspond to the length of the bridges. If these could not be held by anchors because of the great depth, the only option was to use ropes extending from bank to bank (regardless of whether one rope alone or several connected ropes had this length). Because of the current and the lateral wind loads, these ropes must have a certain slack so that the tensile force of the ropes on the fastenings on the bank does not increase to infinity. The ropes are therefore likely to have been around 5 to 10% longer than the distance between the banks - plus the lengths required for anchoring on the banks and, if necessary, on the ships. This results in required rope lengths of over 2200 m.

Assuming the width of a pentecontere to be 4 m, there would be about 3 m between the ships of the 2200 m long bridge made up of 314 ships, if one disregards the peculiarities of the culverts built with triremes. This seems like a reasonable value. For the bridge with 360 ships, this configuration would result in a bridge length of almost 2520 m, which also seems to be a sensible value for the northern bridge, which is not located directly on Cape Nagara.

Width of the bridges or the streets

Herodotus gives no indication of the width of the bridges or the roads leading over them. It is assumed that Greek roads were between 2.7 m and 3.6 m wide at that time, so that a width of 3.6 m can also be assumed for the bridges. This allows four men or two riders to march side by side. A greater width would have no positive effect, as the road at the end of the bridge would not be able to accommodate the incoming masses. In addition, wide swimming bridges are not recommended, as they swayed even more and scared the already nervous horses even more.


Herodotus mentions the order issued during the preparatory phase to manufacture ropes for the construction of bridges more casually than an order for larger quantities of commercially available goods. Only when describing the bridges rebuilt after the storm damage does he give a single concrete statement that the flax ropes, which are heavier than papyrus ropes, weighed 1 talent per cubit, which can roughly be translated as 26 kg / 46 cm. That would be 56.5 kg per running meter. With various conversion methods, this would result in rope diameters of 23 to 28 cm! Ropes with such a weight can no longer be handled, ropes with such a diameter can hardly be bent and therefore cannot be rolled up on cable drums - probably not known at the time - or made transportable in any other way. They could therefore only be attached to bollards several meters thick without breaking them. Herodotus seems to speak of continuous ropes extending from bank to bank. A single rope of 2200 m would have a weight of 124.3 tons and would therefore practically not be transportable today.

Because such ropes are not manageable and therefore have no practical application, it cannot be assumed that any ancient rope maker would have made them before. For this reason alone, the opinion sometimes expressed that the ropes were manufactured and delivered in manageable lengths and only spliced ​​together on site is not tenable.

The idea that the ropes were only made on site on the ships is therefore likely to fail due to practical considerations. If such ropes have never been made, it is more than unlikely that a completely unknown method of production on swaying ships was used to build bridges that were vital to the entire campaign, if those involved were aware that if they failed, they might be beheaded. In addition, the manufacture of ropes requires a certain tension on the strands and the finished rope. It can therefore be assumed that the ships were initially close together so that the tension could be achieved. With three or four ships this may be conceivable, but with a larger number of ships lying in the open water, inevitably swaying, this could quickly lead to considerable damage to the ships and serious disruptions in rope manufacture.

Tensioning such a rope with the help of winches, as described by Herodotus, is impossible.

It must therefore be assumed that the ships forming a wide arch were connected to one another with several ropes that were commercially available at the time. In the end, it is irrelevant whether a rope length only reached from one ship to the next or across several ships. It is also irrelevant whether it was sufficient to position the ships by fastening them to the bow and stern with just one rope. Otherwise, analogous to today's parallel wire ropes, several ropes could have been laid close to one another on suspension bridges, if one ensured that they were evenly loaded. In order to avoid confusion, these bundles of rope could also have been wrapped, which may then have given the appearance of an excessively thick and heavy rope.

The function of the ropes described by Herodotus also meets with considerable doubts. According to Herodotus, the ropes would not only have been used to position the ships, but also to support the wooden planks lying across the ropes . However, every seaman tries to prevent ropes from filing , chafing or rubbing against something in order to avoid their premature wear. The constant movement of the ships in the waves and under the marching army or baggage train, the heavy load of the soldiers and the earth on the planks and their pressure on the tense ropes would certainly have led to their early breakage. On top of that, no level road would have been possible with this construction. The ropes would have sagged under the high vertical loads between the ships (and without special supports also within the ships), so that the road would have been constantly up and down. On top of that, the earth would have collected quickly in the hollows and would have increased the local loads on the ropes considerably. There was also no need for this construction: with a distance of no more than three meters between the ships, the planks could have been attached directly to the ships (parallel to the ropes). This would have made a more solid foundation for the road and would have prevented the ropes from wearing out prematurely.

Bridge deck


The load-bearing bridge deck was formed by the wooden planks, the thickness of which must be assumed to be at least 10 cm. Since sawmills did not yet exist, the wooden planks must have been split and roughly hewn tree trunks. For one bridge around 800 solid cubic meters would have been required, for the other around 910 solid cubic meters, a total of 1710 solid cubic meters. With an average density of 0.5 t / m³, this corresponds to a total weight of 855 tons.


The point of covering the wooden planks with sticks is not apparent. Possibly the sticks should fix the earth.

Covering made of earth

A bridge consisting only of wooden planks was considered completely sufficient in modern times even without additional covering. In modern times, however, wooden roads were occasionally covered with earth to protect the boards from wear and tear, which was also pleasant for riders and chariots. So that the compacted earth does not immediately dissolve under the horse's hooves, it must have been at least 20 cm thick.

Load assumptions

With a 3.6 m wide bridge and ships with a width of 4 m and a gap to the next ship of 3 m, each ship has 3.6 × 7 = 25.2 m² of bridge area, the weight of which is 50 kg of wood and 360 kg per square meter Earth is to be assumed and is then about 410 kg. Each ship would have to carry 25.2 × 410 = 10,332 kg plus the weight of 4 × 7 = 28 people of 90 kg (with luggage) = 2,520 kg, a total of around 13 tons, which was probably an acceptable load for ships at that time .


In order to block the horses' view of the water, screens were installed on both sides of the bridges. Assuming that such a screen is only 2 m high, the area for one side of the 2,200 m long bridge alone is 4,400 m². Even with minimal wind strengths, the resulting wind loads would not have been manageable with the means available at the time. In modern swimming bridges, simple railings made of slats or ropes have proven to be perfectly adequate to keep the horses on the bridge.

Passage openings

The three openings for the passage of smaller ships were probably created by inserting higher triremes into the row of penteconters or merchant ships, and the ropes running through, similar to a bridge ramp, were slowly raised by trestles to the height required for the passage. Since the ships could easily lower their masts, a clearance of 2 m above the waterline should have been sufficient. If the load on the ropes increases due to the wind, the triremes would be pushed into the water a little, but this only lasts as long as the wind blows.

Storm damage

From Herodotus report that the storm completely destroyed the bridges, little concrete can be derived despite the apparently clear formulation. It is not clear whether and to what extent ships, ropes, planks etc. could be salvaged, repaired and reused. However, it cannot be derived from his report that all parts had to be newly procured. The preparations for the bridge construction took years, so replacement deliveries of ships, ropes, anchors and wooden planks would have taken at least months. The laying of the wooden planks and the earth layer alone must have originally taken a few days. Even if one assumes that everything could be repaired and that no subsequent deliveries were required, a repair time of several days must be expected. During this time, however, an extremely serious situation would have arisen for the army waiting on the bank, since the supplies of food, feed and water were not calculated for a longer stay.

In Herodotus' report it is noticeable that the first bridges are hardly mentioned and are already reported as destroyed, while the bridges built afterwards are described in detail without, however, saying a word about the duration of the repair. This can lead to the assumption that the bridges allegedly destroyed by the storm Herodotus only served as a hook in order to be able to immediately describe an outburst of anger of the great King Xerxes in numerous details, and even reproduce the king's speech in full.



  1. Herodotus 7:21 and 7.25
  2. Herodotus 7:34
  3. Herodotus 7:35
  4. a b Herodotus 7:36
  5. Herodotus 7:55, 56
  6. Herodotus 8.97
  7. Herodotus 8:10
  8. Herodotus 8,118
  9. A summary of the questions can be found in NGL Hammond, The construction of Xerxes' bridge over the Hellespont in the preface
  10. A direct pictorial representation of Herodotus' representation e.g. B. at EDSITEment !, The bridge over the Hellespont
  11. The Black Sea Pilot, p. 17
  12. GeoHack - MapTech, nautical maps
  13. The Black Sea Pilot, p. 30
  14. ^ The Black Sea Pilot, p. 30, Rhodius River
  15. GeoHack - MapTech, Nautical Maps
  16. ^ The Black Sea Pilot, p. 32
  17. Barker (p. 33) quotes the British General Frederick Maurice, The size of the army of Xerxes in the invasion of Greece, 480 BC
  18. Today's Atikhisar Dam, located in the mountains, is only 10 km as the crow flies.
  19. Barker, p. 41
  20. Barker, p. 32
  21. ^ Hoyer, Handbuch der Pontonnierwissenschaften, p. 393
  22. a b c Hammond, p. 98
  23. ^ Hoyer, Handbuch der Pontonnierwissenschaften, p. 403
  24. z. B. Barker, p. 30, Hammond, p. 93 in the location sketch
  25. See anchor # water depth . An anchor line should be ten times the length of the water, an anchor chain would still need five times the length
  26. Hammond (p. 98) quotes Robert Chapman, A treatise on ropemaking as practiced in private and public ropeyards ... (Philadelphia 1869), according to which an iron anchor for a ship comparable to the Pentekontere would have to weigh about 136 kg. That would result in a total weight of 1348 anchors x 136 kg / anchor = 183.328 tons of iron.
  27. Barker, p. 34
  28. Even if Hammond describes and sketches the iron stick anchors in detail, it can be ruled out that the iron smelting and processing of the time was able to produce iron anchors with a total weight of over 183 tons
  29. The different opinions on the length of a stadium cannot be discussed here.
  30. Hammond (p. 91) explains the difference to Herodotus' information with the fact that the water level was then 5 feet or 1.52 m lower than it is today. However, this does not explain why the coast then ran between the 20 m line of water depth on one side and the 30 m line on the other side (p. 93). Nor does he discuss that a lower water level would mean a higher flow velocity.
  31. See parallelogram of forces
  32. The Handbuch der Pontonierwissenschaften (p. 390) recommends that when using thick and long boards, the space between the ships should not be greater than 6 m for reasons of stability.
  33. Hammond, p. 95: 9 to 12 feet
  34. Hammond, p. 95
  35. Hoyer, Handbuch, p. 402
  36. The conversion takes place again without taking into account the geographically different units and the diversity of opinions of historians
  37. In today's rope trade, z. B. 200 m long ropes from Manila with a diameter of 60 mm and a weight of 2.49 kg / m or made of hemp with 40 mm and 0.56 kg / m are offered, whose breaking loads are 22 and 10 tons respectively.
  38. Hammond (p. 99) calculates with a cubit of 52.7 cm and a practical rule of thumb from Robert Chapman, A treatise on ropemaking as practiced in private and public ropeyards ... (Philadelphia 1869) a diameter of 23 cm; Barker (p. 34) uses simplified numbers to calculate a diameter of 25 cm over the area of ​​the circle; a comparison of the weight and the circular area of ​​a commercially available rope results in a mathematical diameter of over 28 cm.
  39. Hammond (p. 101) describes an attachment by means of an eye splice on a 45 cm thick bollard, without any discussion of how one can splice an eye on a 23 cm thick rope, how the rope should withstand such a tight bend or how the one at that time Iron smelting and processing could have made such powerful bollards.
  40. Hammond (p. 100) calculates a weight of 162,000 lbs or 73.636 tonnes for the 1,500 m long rope (which would result in a weight of 108 tonnes for the same 2,200 m long rope), but does not in any way apply to the same Weight related problems a.
  41. Even today, natural fiber ropes with such a diameter should not be made. This is probably why no attempt has ever been made to splice ropes with such a diameter that it is not even known whether the idea could be carried out in a practicable way.
  42. a b Hammond, p. 92 ff.
  43. "because - even with the help of an applied winch - one is never able to pull a rope that is 1000 and more Rhineland feet long enough , such as on the Rhine, the Vistula or the Danube " (Handbuch der Pontonnierwissenschaften p. 406 ), quite apart from the fact that it would have to be winches with huge drums.
  44. In the manual (p. 390) a thickness of 8-10 inches = 20-25 cm is naturally assumed, but for larger spaces
  45. 2200 mx 3.60 mx 0.10 m = 792 m³
  46. 2520 mx 3.60 mx 0.10 m = 907.2 m³
  47. Hoyer, Handbuch, p. 405
  48. Hammond (p. 100) quoted from Ira Osborn Baker: A treatise on roads and pavements (New York 1908) that such wooden plank roads were common in the wooded areas of the USA and Canada.
  49. Wood: 0.5 t / m³ x 0.10 m = 0.05 t or 50 kg; Earth: 1.8 t / m³ x 0.20 m = 0.36 t or 360 kg.
  50. Due to the inaccuracy of the assumptions, the weight of the brushwood covering and the screens, but also the ropes, can be neglected.
  51. Hammond (p. 100) uses 9 feet = 2.74 m more appropriately.
  52. For comparison: the Gorch Fock has a sail area of ​​2,037 m², the Kruzenshtern (formerly Padua ) has a sail area of 3,400 m².
  53. Hoyer, Handbuch p. 412; see. z. B. also the picture of a Swedish floating bridge
  54. Hammond, p. 92; Barker, p. 36