Chance can play a large part in many of our activities and it was a book published a few years ago entitled ‘The Man Who Invented History: Travels with Herodotus’ by Justin Marozzie that sparked this author’s interest in the Eurpalinos’ tunnel.
Driven simultaneously from both ends of the alignment by the Greeks in the 6th century BC (Sandstrom 1963) this 1,040mlong tunnel was successfully constructed to provide a reliable water supply to the ancient city of Samos, now called Pythagorio (in memory of Pythagoras, the philosopher and mathematician of hypotenuse fame who was born on that island circa 570 BC).
Background
Samos, a comparatively green and fertile Greek Island, is located in the North Aegean Sea just off the coast of Turkey from which it is separated by the narrow, 1.6km wide Mycale Strait. About 45km long and 13km wide at its greatest extent, it has an area of almost 480km2 and a population of around 34,000. A large portion of the island is covered with vineyards and apart from grapes the main agricultural products are honey, olives and olive oil, citrus fruit, dried figs, almonds and flowers.
Although the early history of Samos is obscure, it was relatively quick to develop, and by the 6th century BC it had become one of the main commercial centres of Greece largely due to the island’s proximity to trade routes. A succession of tyrants culminating in the rule of Polycrates saw the island reaching the height of its prosperity around 535 BC.
Engineering works
These were interesting and exciting times for Samos. There was a growing threat from the Persian Empire at its doorstep on the nearby mainland, while Polycrates with the most formidable fleet in the Aegean had expansionist ambitions. Thus defensive measures were needed on the island to provide a secure basis for conquest and piracy abroad. The Eupalinos’ tunnel formed part of these defensive works, which also included city walls at just over 6.4km long, 6m high and 2m wide, with 31 towers, a moat to the west of the city of Samos and a breakwater extending some 400m out into the sea, which was up to 18m deep to provide a harbour sheltered from southerly winds. Contemporaneously the Heraion, the largest temple in the world at that time, was under construction to the west, further along the southern coastal plain. Indeed, this part of Samos must have looked like a building site for much of the 6th century BC.
Writing a century later, Herodotus considered the tunnel, breakwater and temple the three greatest works of the Greeks. He judged the tunnel to be pre-eminent while providing only brief details of the other two works.
The Eupalinos’ tunnel
This water supply tunnel which remained in use until late Byzantine times was rediscovered by Abbot Kirillos in 1881 and found to be essentially as Herodotus had described it. As already indicated it is 1,040m-long and was driven through Mt. Kastro, a 239m high hill, to convey water from the Agiades spring to the north into the city of Samos. The tunnels consists of two parts:
An upper section which is generally squarish, 1.8m wide and high with a horizontal floor
A lower part where open topped gutters have been laid at the bottom of a trench located along the east wall of the tunnel. This trench varies in depth from 3.4m below the tunnel floor at the northern entrance to 8.5m at the southern exit: hence the gradient of the water channel is about one in 200 or 0.5 per cent. See figures 1 and 2.
Herodotus tells us that the architect for the tunnel was Eupalinos, son of Naustropos, who hailed from Megara on the Greek mainland. Unfortunately he does not tell us how Eupalinos determined the alignment and level for the tunnel or how it was constructed. Mt. Kastro is composed of Miocene rocks – soft marls and limestone to the north and hard limestone to the south (Higgins and Higgins 1996). The workmanship was of a high standard. The ceilings and walls are generally bare rock that appears to have been peeled off in layers and chisel marks can still be seen.
The tools for excavating were hammer, chisel, candle or lamp and bucket. The interesting question is the rate of progress achieved. Estimates vary from five to 15 years to complete the tunnel; the meeting point between the two drives is 620m from the northern entrance giving an average daily rate of progress of within the range 0.11 to 0.34m per day on this leg of the tunnel. It has been suggested that the southern leg of the tunnel could have been started some years after the northern leg but there is no mention when positing this that the limestone is harder to the south.
Working conditions were almost certainly poor and work would probably have gone on round the clock seven days a week: given the harshness of the conditions it is very likely that the tunnellers were slaves and it is hard to imagine may of them surviving from start to finish of the project. However one looks at it, this was a colossal achievement for manual labour.
Setting out the tunnel
Equally impressive was the determination of the line and level on which to construct the north and south legs of the tunnel so that they would rendezvous somewhere beneath Mt. Kastro. And there has been much speculation on how Eupalinos solved these problems. The two ussues are ensuring that the two legs were correctly aligned and that the floor of the upper section of the tunnel at the north and south portals was at the same elevation. Both problems can be solved by going either over the top or around the side of the mountain, or indeed by using both methods in conjunction as any first rate engineer – and Eupalinos certainly was that – would do if both were available at the time.
Aligning the tunnel
With the over-the-top method it is relevant to remember that the tunnel was just over a kilometre long so that 100 people along the line would be only 10m or so apart. It would not be too difficult to organise these into a straight line between the portals and a quite clever method for doing this has been described (Belloc 1925). Even a rope, string line or chain could have been used. Once determined, the line would have been permanently marked.
With the around-the-side method, a stepped series of lines at right angles to each other would have been laid out around the mountain between the tunnel portals as shown in figure 3. Irrespective of the number of steps involved, a right angle can be projected through the mountain with its hypotenuse marking the line of the tunnel. The method depends on the Samians being able to set out right angles and although opinion is divided whether this was so (Rihill and Tucker 1997, Apostol 2004) the likelihood is that they could, given that they had been around since Egyptian times.
Take your pick then on what was used but on the basis of simplicity and straightforwardness the over-the-top method would appear to be the more likely.
Once the line of the tunnel had been determined externally, projecting it once the tunnels were inside the mountain is the problem. Again there is no information on how it was done. Indeed it would appear that they made life difficult for themselves since the entrances of both ends have been dug off line and do not follow the direct alignment until some 7m and 15m inside the hill of the north and south ends of the tunnel respectively. How they did it all was certainly clever stuff for those times.
Finding the level
This appears to be the more difficult problem. With the over-the-top method, measuring poles would be used to measure the height up one side of the mountain and down the other. Logically they would have started from the north portal, not necessarily but preferably in a straight line up the mountain, sticking in measuring poles as needed to ensure that they could sight horizontally from the upper part of the lower pole to the bottom part of the one above. On reaching the top they would add up all heights measured on the poles going up the north side of the mountain. The process was then reversed coming down the south side of the mountain until the sum of the heights equaled that for the north side which was the level at which the south portal should be excavated to provide a horizontal floor to the upper section of the tunnel. In principle the method is straightforward and simple but the rise and fall over the mountain with a portal level at the 55m contour and a summit of 239m would have required some 200-400 difference of elevation measurements, each subject to error.
It is likely the outcome would not be sufficiently accurate.
On the other hand going around the mountain is more promising for this task. Rudimentary levelling devices using water – a method still used in setting up surveying instruments – had been invented in Egypt earlier. Using such an instrument, the level at the south portal would be found by going around the mountain from the north portal preferably on the level. This would have been quite difficult following the 55m contour but would have been quite feasible at a lower elevation. Again it is tempting to speculate which method was adopted but in this case it would appear that carrying levels around the mountain would have been easier to accomplish than the alternative.
Driving the tunnel
The timescale of construction of the tunnel has already been considered, and had the tunnellers trusted the setting out and continued on the straight-line direction on which they began, a near perfect junction would have eventuated. However there were obviously doubts in their minds and there would have been ample time for them to fester and cause concern. About half way to the meeting point, the northern leg begins to zigzag and changes direction a number of times before finally making a sharpish left turn to intersect the south leg. Floor level in the north leg was 580mm higher than in the south leg, just another of all the remarkable achievements on this project.
Construction of the inclined channel for the water supply would have been another time-consuming operation given that the width of the excavation was just sufficient for a single person to carry out the excavation. Located along the eastern side of the tunnel and up to 8.5m deep, the amount of excavation involved was approximately 3,000m3 and only a little less than from the upper section of the tunnel. Its completion would have been the final act in the construction of the tunnel. Nobody knows how long it took to complete the project and, as already mentioned, estimates range from five to 15 years on the basis that things usually take longer than expected, this author’s preference would be for the longer period.
Resourcing
Quite significant amounts of labour and materials were required for the building works in the city of Samos and its environs for much of the 6th century BC. Highly skilled labour was needed for the planning, design and setting-out and supervision of the water supply system, temple, fortification and breakwater. In the case of the water system, the Samians obtained the services of an outsider, Eupalinos of Megara. On the other hand two locals, Thoikos and Theodoros, were responsible for the Heraion temple. It is unlikely too that all the skilled craftsmen and semi-skilled workers could have been obtained within the island of Samos and some would almost certainly have had to come from elsewhere. Indications are that Polycrates paid high wages to immigrant craftsmen.
Working conditions within the tunnel have already been mentioned with most of the workforce likely to have been slaves given Polycrates’ penchant for piracy. One can visualise two, perhaps three men working at the face by the light of smoky candles or oil lamps with their noses and mouths protected by cloths from the dust their chiseling is creating. Behind, their helpers are cleaning out and carrying away the excavated rock, removing blunted chisels and supplying re-sharpened or new ones as well as providing food and drink. There was only natural ventilation. Excavation probably continued 24 hours a day and we do not know how many shifts there were. But we can almost certainly be sure that working conditions were awful and nobody had heard of health and safety!
By comparison with other building projects on Samos dealing with the 3,500m3 or so of excavated rock from the upper section of the tunnel at a rate of about 1m3 or two tons per day would not have been difficult. The supply of the 2.5km of clay pipes and guttering at an average rate of about a metre a day would be quite easy.
Conclusion
Engineering world firsts – which the Eupalinos’ was in both length and being driven from both ends – are of their very nature remarkable achievements and indicators of the pinnacle of engineering expertise and capability at that particular time. It is an enduring monument to its skilful designers and the indomitable spirit and herculean efforts of the tunnellers who constructed it. It is also the outstanding project of the tyrant Polycrates. Tyrants, dictators and absolute monarchs can be obnoxious individuals, prone to self-aggrandisement and the wish that posterity will think well of them. Some of them see infrastructure as a route to immortality.
It is fascinating for those interested in the history of engineering and other social matters to reflect on the differences and ambiguities that appear in the various descriptions of the same events. But this is only to be expected and the reader is left to draw his own conclusions. Even now it is an everyday occurrence to find each daily newspaper giving its own view of ‘events’ and at times differing substantially with each other on the ‘facts’ themselves. And on many matters new information may emerge years after an event.
In the above piece we may appear to criticise Herodotus for not telling us more. But writing a hundred years later with only hearsay and little else to go on, was he as much in the dark concerning the details of the works as we are?
Figure 1, plan of Eupalinos Tunnel Figure 2, cross section of Eupalinos Tunnel Figure 3, around mountain method of finding alignment
