Fig. 1 - Location map.
Carte de localisation.
1During antiquity, a number of major engineering works, requiring high technical and geodetic skills, was undertaken. These included projects to build the Suez canal in Egypt and the Corinth canal in Greece (Fig. 1), both eventually completed during the nineteenth century. The reasons for the failure of the projects were not a lack of funding, changes in the political situation, or unresolved technical problems. Historical sources suggest that the plans for such impressive works were mostly abandoned because geodetic studies revealed a 10 m difference in sea-level elevation at their inlets. In light of this, it was hypothesized that excavation of both canals would lead to flooding of the low-lying coasts, while the expected continuous flow of water from one inlet to the other would hinder ship navigation. This postponed the construction of the canals until the nineteenth century; since this time the canals operate without locks, confirming the absence of significant sea-level differences (Lewis, 2001).
2Such gross errors in leveling appear surprising, given that engineers in ancient Greece and Egypt were notorious for their mastery of surveying techniques. These skills are for instance epitomized first by the ~1000 m long tunnel of Eupalinus (Samos Island, Aegean Sea), successfully excavated in ~540 BC (Kienast, 1995), and second, by qanats. The latter, widespread in the Middle East and not uncommon in the Mediterranean, are underground sub-horizontal water channels up to tens of kilometers long, excavated at the bottom of deep wells (English, 1998; Stiros, 2006). Stiros (2006) has shown that ancient surveyors were able to obtain high accuracies in leveling. This makes geodetic errors unlikely for the Suez and Corinth canals. In light of this, we re-evaluate the available information concerning the early history of these two canals, and try to shed some light on the real causes of their abandonment in antiquity, as well as on the misconceptions for the threat of coastal flooding following their excavation
3The present-day Suez canal, completed in 1869, represents a system of artificial water channels and lakes along the isthmus connecting Asia and Africa, and separating the Red Sea from the Mediterranean. The isthmus is about 160 km long, of low elevation and flat-topped (Fig. 1a) and has been formed by Late Quaternary tectonic uplift (cf. Jacksonet al., 1988; Gvirtzman, 1994).
4The construction history of the Suez canal is unclear, but has been summarized by various authors (e.g. Marlowe, 1964; Lewis, 2001). Plans for a navigable channel in this region are very old, with ancient reports and archaeological remains indicating that an early canal had been excavated in the isthmus 2-3,000 years ago.
5The first historical report is that of Aristotle (Meteorologika 353b), the famous fourth century BC philosopher, who reported that Pharaoh Seti (or Sesostris, around 1300 BC) and the Persian king Darius, who occupied Egypt in circa 500BC, tried to excavate a canal similar to the present-day one. Their projects were, however, abandoned due to fears over freshwater contamination by saline water caused by flow of saline water due to altitudinal differences between the Red Sea and the Mediterranean. Interestingly, a channel mentioned in an inscription at Karnak, dating from Pharaoh Seti’s reign, is usually taken as evidence of an early Suez canal.
6Around 100 AD, the Greek historians Diodorus Siculus (Historical Library 1, 33.9-10), Strabo (Geography XVII 1.25) and the Roman historian Pliny (Natural History VI 33), probably citing older information, reported efforts to excavate a canal by Pharaoh Sesostris or later by Pharaoh Necho II (~600 BC), and then later by Darius around 500 BC, were abandoned due to the risk of flooding of Lower Egypt. Another possibility is astrologers warned that the channel would open the country up to foreign invasion. Pliny also noted that some feared contamination of freshwater reserves by saline Red Sea water (Pliny, Natural History VI 33). For these reasons only the part of the canal from the Nile to the Bitter Lakes, 34miles long, was completed (Fig. 2). Inscriptions on a number of granite stelae state that “I ordered the canal to be dug up from the River called Pirava (the Nile), which follows in Egypt to the sea that comes out of Persia (The Red Sea)” and suggest that the canal to the Red Sea was completed by Darius. In reality, it appears that the connection between Bitter Lakes and the Red Sea was made over land, using camels and horses.
7This interpretation is supported by the fact that a canal was excavated in the area during the Ptolemaic (Hellenistic) period ~270 BC, with the construction of a lock to counteract the expected flow of water. The town of Arsinoe was founded at the canal’s inlet, near present-day Suez (Fig. 2).
8The ancient canal was repaired by the Roman emperor Trajan ~100 AD, and was extended to meet the main branch of the Nile near Cairo (Fig. 2). It remained navigable until the 3rd century AD, while repairs were made to the southern branch between the Bitter Lakes and Suez. The exit of this canal was not at Arsinoe, but a few kilometres to the west. The Roman canal soon silted up, but was again re-opened in the seventh century AD by Amr ibn al-Aas, the Arab conqueror of Egypt, and came to be known as the “Canal of the Commander of the Believers to the God.” Amr ibn al-Aas planned its extension to the Mediterranean, but his project was dropped on military grounds (the fear that it would facilitate invasions), and for these reasons the canal was closed in 776 AD.
9Shortly after the opening of marine routes around Africa in the fifteenth century, the Venetians drafted the first modern plans for a Suez canal to shorten the distance to India; work was resumed by Napoleon around 1800. Yet again, however, his engineers estimated a 10 m elevation difference between the Red Sea and the Mediterranean. The result was debated by Fourier and Laplace, famous mathematicians and physicians of the period, but nonetheless served as a catalyst to the abandonment of Napoleon’s project. Interestingly, parts of the ancient canal were used as an aqueduct bringing fresh water for the workers and the engineering works during the construction of the modern canal (Marlowe, 1964; Lewis, 2001).
Fig. 2 - Location map of the Suez canal area.
Carte de localisation de la région du canal de Suez.
Names in italics denote ancient sites. Dashed lines with numbers in circles indicate partially excavated sections of the ancient canal during 1: Pharaonic and Persian times (up to 500 BC); 2: Hellenistic times (~270 BC); 3: Roman times (~100 AD).
Les noms en italique correspondent aux sites antiques. Les pointillés avec des cercles localisent les secteurs creusés durant : 1. Périodes pharaonique et perse (jusque vers 500 avant J.-C.) ; 2. Période hellénistique (vers 270 avant J.-C.) ; 3. Période romaine (vers 100 après J.-C.)
10The available historical, archaeological and geomorphological data allow us to suggest that the geomorphology (a low-altitude, flat area) and the geology (soft Quaternary and Holocene sediments) of the Suez isthmus facilitated early plans for an irrigation channel between the eastern (Pelusiac) branch of the Nile near the town of Bubastris, to the swampy area of Lake Timsah and the Great Bitter Lake (Fig. 2). This was a sizeable task for, as can be seen in the satellite photo (Fig. 3), the channel was to be excavated along stable, low-elevation ground following an E-W trending abandoned branch of the Nile River (Wadi Tumilat) flowing into the Red Sea. This channel closely resembled the present-day canal, and its main, if not its sole function was irrigation. It is likely that it was completed around 600 BC (during the reign of Pharaoh Necho), and was probably the only channel constructed during Pharaonic times, given the need for irrigation and flood control.
11After the mid first millennium BC, when the need for a marine passage between the Persian Gulf and the Mediterranean became necessary, a second navigable channel was excavated from the Bitter Lakes to the Red Sea. This broadly followed the path of the southern segment of the modern canal from the Bitter Lakes in the north, to the Suez, on the Red Sea coast (Fig. 2). Plans for an extension of this channel to the Mediterranean were, however, never realized.
12Persian period inscriptions commemorating the construction of a navigable canal (see above) seem to indicate a waterway completed under Darius’ reign. However, it is more likely that ~500 BC Darius rendered the northern part of the waterway navigable from Bubastris to the Bitter Lakes only, and that the southern section from the Bitter Lakes to the Red Sea was completed later, around 270 BC by Ptolemy. The foundation of Arsinoe at its inlet, near present-day Suez (Fig. 2), supports this conclusion. Because of the silting up of the Nile’s eastern (Pelusiac) branch, it is likely that during Roman times (Trajan’s reign, ~100 AD), it became necessary to excavate a N-S extension canal from Bubastris to the main branch of the Nile at the town of Babylon, modern Cairo (Fig. 2). The path of the Suez section was also modified, reaching the sea at the town of Clysma, some kilometers west of Arsinoe.
13The elucidated history of the ancient canal is a reasonable one, though variations on this version are presently debated by historians and archaeologists. What is important, however, is that the construction of irrigation channels was welcomed by everyone. Nonetheless, the plans for a navigable canal were delayed by detractors (mostly astrologers and geodesists) for many millennia!
Fig. 3 - Satellite view of the Suez area.
Image satellite présentant la zone de Suez.
Grey zone marked represents Wadi Tumilat, a cultivated, low-elevation corridor through which the first irrigation canal was excavated around Lake Timsah.
La dépression du corridor de Wadi Tumilat (zone grise) correspond à la première zone de creusement du canal dans la région du lac Timsah.
14Modern science does not permit us to evaluate the conclusions of astrologers in the Pharanoic court, who were apparently opposed to the construction of the navigable channel. However, it does allow us to evaluate the results of ancient surveyors who are supposed to have measured a 10 m altitudinal offset between the two inlets of the planned canal.
15Measurements for Darius’ canal were probably made by his engineers, who first introduced qanats to Egypt. His engineers, well-trained in leveling long distances in arid areas, were definitely able to obtain accuracies of 1-2 m over distances of 100 km on flat terrains. Stiros (2006), based on the analysis of qanats, concluded that even using primitive instruments ancient surveyors were able to measure elevation changes over long distances with standard errors σΔh defined by the formula(1)
16σΔh = σo√S
17where S is the length of the leveling route in kilometers and σo =0.45- 0.03m/√km. For a distance of 160 km (i.e. equal to the length of the isthmus), a standard error of 0.5- 5 m can be estimated. Such impressive accuracies, obtained after repeated and time-consuming measurements, obviously exclude the possibility of leveling errors of 10 m along the Suez isthmus, especially considering the survey was made over totally flat and uniform terrain.
18On the contrary, the survey by Napoleon’s engineer J. Lepréduring during the nineteenth century, although based clearly on higher accuracy leveling instruments (i.e. instruments with an alidade and bubble levels), led to an error of nearly 10 m (29 feet). This difference can be explained by the fact that the French engineer was not acquainted with refraction errors, a major source of error in arid climates (Bomford, 1971; Vaniceket al., 1980; Strange, 1981; Stiros, 2006). It appears that his measurements were severely affected by systematic errors. The correct elevation difference was finally computed in 1846-1847, as a result of surveys made by the French Society for the Study of the Suez Canal.
19The estimation of sea-level differences on an isthmus is based not only on leveling measurements, but also on the identification of the necessary reference levels. Such levels in the modern hydrographic literature are known as Mean Sea Level (MSL), Mean Low Water (MLW), Low Water Line (LWL, used as a reference level by port authorities), High Water Level (HWL) or Mean High Water (MHW). The above levels can be identified on the basis of either tide-gauge observations, or biological and morphological data (Parker, 2003; Leartherman, 2003). Although tide-gauges are known to have been used to monitor the level of the Nile (the “nilometer” has been in use since at least the eighth century BC) and are described by Strabo and other authors, we can reasonably assume that ancient geodesists and engineers have probably relied on non-instrumental evidence to identify reference levels (“datums”) for their calculations.
20Modern data reveal that the difference in mean sea level between the two canal inlets is practically nil. There is, however, a significant difference in the tidal ranges: ~10 cm in the Mediterranean and up to 2.7 m in the Red Sea (Bourdon, 1925). Even if it is assumed that ancient engineers based their measurements on the Mean High Tide Level, i.e. levels that can be easily identified on the basis of geomorphological and biological observations (Pirazzoli, 1996; Morhangeet al., 1998; Stiros and Pirazzoli, 2004), this only accounts for ~1 m of error.
21Since the two errors (i.e. leveling and sea level) are not correlated, according to the law of error propagation (e.g. Bomford, 1971), the total error σ can be calculated from the formula (2)
22σ2= σ2leveling+σ2sea level
23This leads to a standard error of around 5m. In light of this, a 10 m elevation difference seems too large even if both the datum (i.e. deriving from the identification of the mean sea level) and leveling errors are taken into account.
24Corinth isthmus, 6.2 km long and more than 80 m high, comprises a horst of faulted and uplifted Pliocene and Quaternary sediments at the eastern edge of the Gulf of Corinth graben (Freyberg, 1973; Mariolakos and Stiros, 1987).
25The isthmus linked the Greek mainland to the Peloponnese, and served as a physical barrier to marine traffic. Significantly, it isolated the Gulf of Corinth from the flourishing regions of the Aegean, in addition to southern Italy and Sicily (Fig. 1). Due to its strategic position next to the isthmus, the town of Corinth flourished in ancient Greek and Roman times. The city was served by two harbours: (1) the eastern harbour of Kenchreai in the Saronic Gulf; and (2) the harbour of Lechaion in the Gulf of Corinth (Fig. 4). The latter was the first artificial harbour to be constructed, probably around 600 BC (Stiros et al., 1996). During the mid first millennium BC, a paved ramp (“Diolkos”) was constructed along the isthmus, permitting ships to be transported from one side to the other. This ramp followed the path of the present day canal (Fig. 4), and some of its remains are still visible today (Verdelis, 1956).
26Diolkos made Corinth very wealthy, but apparently could not meet the needs of commerce and communication. For this reason plans to excavate a canal were conceived early in antiquity. The Xerxes canal was excavated at the NW edge of the Aegean for military purposes during the Persian Wars (~480 BC) and attests to the potential for major engineering projects in ancient Greece (Isserlin et al., 2003; for location see Fig. 1).
27The major advantage of the Corinth isthmus is that it was built of very coherent Pliocene and Quaternary sediments, mostly marls, which permit excavation of subvertical walls up to 80 m high. These favorable engineering conditions were known to the ancient Greeks and the first plans for a canal were probably made as early as 600 BC, at a time when ancient Corinth flourished.
28Historical accounts suggest that Dimitrios Poliorkitis, the most prominent successor of Alexander the Great, first tried to open the Corinth canal ~300 BC. He was advised by Egyptian experts who stated that, due to the higher elevation of the Corinthian Gulf, water flow would drown islands in the Saronic Gulf (Aegean Sea). Interestingly, Strabo also refers to a debate between the ancient geodetist Eratosthenes, who proposed an alternative concept to Archimedes’ work on fluids (Strabo, Geography, 1 3.11, 14), to suggest that uneven sea levels could exist even over relatively short distances.
29Roman emperors Julius Cesar and Caligula are thought to have planned the excavation of the canal. The most important works were, however, undertaken by Nero, in 67 AD, using captives from the Jewish War. Logistic difficulties meant that the project was eventually abandoned (Gerster, 1884), although partial remains of this early excavation are still visible. Two later reports refer to the efforts of Nero. The first, by pseudo-Lucian (pseudo-Lucian,Nero IV) written ~200 AD cites Strabo’s explanations for the causes of the interruption of the works. The second, by Philostratus, a Greek writer around 220 AD repeats the same scenario, but cites political instability as the cause for abandonment of the project (Philostratus, Life of Appolonius of Tyana, IV 24). The final excavation was started in 1881 and was completed in 1893.
Fig. 4 - Satellite view of the Corinth area, Greece.
Image satellite de la région de Corinthe.
The Canal excavated at the end of the 19th century is shown. A minor current from the Gulf of Corinth (to the west) to the Saronic Gulf (to the east) is observed, but is too small to represent a threat for flooding of the Saronic Gulf, as was argued in antiquity.
On voit clairement le canal creusé au XIXe siècle. Un léger courant marin s’écoule du golfe de Corinthe à l’ouest en direction du golfe Saronique à l’est. Durant l’Antiquité, certains auteurs pensèrent que ce courant aurait pu submerger les rives du golfe Saronique.
30The Corinth isthmus is 6.2 km long and ~80 m high in its central part, with a smooth relief. On the basis of equation (1), the possibility of an important elevation error in between the two exits of the canal is very small.
31As is the case with the Suez isthmus, there is no significant difference in sea level, but significant differences in tidal level do exist on either side of the Corinth canal, with tide gauges recording a mean tide of 0.30 m in the Gulf of Corinth and 0.08 m in the Saronic Gulf (Zoi-Morou, 1981). Even if it is assumed that ancient engineers used the high tide mark and not the mean sea level as their datum, the difference is only around half a meter. Such a discrepancy is obviously too small to drown the Saronic (Aegean Sea) coasts, although it can produce a constant current. A permanent, slow current is indeed observed in the modern canal.
32Debate over sea level differences between the two inlets of the Suez and the Corinth canal, both leading to the abandonment of these projects in antiquity, present a number of similarities: (1) Measurements by Egyptian engineers are assumed to have led to significant sea-level offsets between the two inlets of the planned canals. It was hypothesized that such differences would lead to the drowning of “low-lying” coastal areas. (2) Significant differences are observed in tidal levels either side of the two isthmuses, and such differences do produce weak currents. (3) Although important in the 1800 survey, the leveling errors cannot explain the reported elevation differences in sea level measured in antiquity.
33Ancient engineers were certainly aware of water flow along channels, as attests the example of Euripos Straits in Central Greece (Fig. 1). In fact Strabo mentions that Eratosthenes, a famous 3rd century BC geodetist, had attributed the flow to differences in sea level on either side of the straits (Strabo, 1, 3, 12, 36, 55). This provided an alternative explanation to that of Aristotle, who in his Meteorologika had assigned flow to hydrodynamic effects. Archimedes suggested that sea waters follow the laws of ideal liquids, and hence excluded the possibility of uneven sea levels supported by “geodesists” such as Eratosthenes (Strabo, Geography, I, 3, 14). Interestingly, similar debates arose in the nineteenth century, with famous mathematicians and physicians such as Laplace and Fourier rejecting the geodetic measurements along the Suez canal on the basis of physical laws (see Marlowe, 1964).
34It appears therefore that the heart of the problem is not the geodetic measurements but rather the corresponding hydrographic conditions, and notably differences in tidal range.
35There is another parameter which has a role to play, at least in the case of the Suez canal. Partly due to sea-level variations (Gvitrzman, 1994), several lagoons exist between the Bitter Lakes and the Red Sea, in which a fauna and flora different to that of the Mediterranean exists (Por, 1971). As Pliny noted, a flow of water from the Red Sea towards the Bitter Lakes, Wadi Tumilat, etc. (Figs. 2, 3) threatened to contaminate the freshwater reserves.
36Thus, two groups were opposed: (1) those trying to open marine routes for financial and military reasons, and kings wishing to associate their names with important works; and (2) farmers, worried about the contamination of freshwater reserves, fishermen, some military officers and those living from traditional trade.
37It is therefore likely that the abandonment of the early project for a complete Suez canal in antiquity was not due to a “geodetic error”, as is widely accepted, but rather to detractors’ exaggeration of hydrographic and ecosystem differences either side of the isthmus leading to a fear for flooding of coastal areas. In the case of Corinth, the contrast in the physiographic and hydrographic conditions at the two inlets of the planned canal was minor. The possibility of an error in the sea-level estimations was moderate. The abandoned Suez project and observations at the Euripos Strait, in combination with the huge amount of material to be excavated for the canal (walls up to 80 m high) led to the collapse of the Corinth project.
38It is usually assumed that gross (up to 10m) errors in leveling revealed a considerable difference in sea-level at the two inlets of the Suez and the Corinth canals, and this led to fears that (1) a continuous flow of water would render both canals non-navigable, and (2) it would lead to drowning of coastal areas. Such fears are assumed to have prevented the excavation of the two canals during antiquity. Examination of the evidence reveals that such large of errors are unlikely in the ancient geodetic surveys. More important is the contrast in hydrographic and physiographic conditions at the two inlets of the canals. These differences fuelled debates on water currents through the planned canals. Exaggeration of possible negative impacts, in particular of the risk for flooding of coastal areas, was used by detractors to force the abandonment of the projects during antiquity.