The Tunnel of Eupalinos

(by Dan Hughes, with appendix by Hans J. Keller)

I've just had a small but splendid adventure, the recounting of which requires some background that will take up the beginning of this story. It all took place in the week 8 to 15 April 2000. (It's written approximately as it happened, and things are reported the way we thought they were. However there are some addenda at the end, with additional or more accurate information that I, or we, came across later.)

The Greek island of Samos is close enough to the Turkish coast that a really strong swimmer could make the crossing easily I think. Samos is unusual for Greek islands in having a lot of forests, and is probably more prosperous than the average Greek island too. This prosperity existed already before tourism reared its head after WW2, and has only been increased by the Germans and Dutch and English and French visitors during the past half century. In the 7th and 6th centuries B.C. Samos was an important part of Greece, prosperous then too, with a much admired capital city and with many cultural achievements: Pythagoras came from here. The walls enclosed a city, on the south coast and then called Samos also, that stretched more than 1500 metres - about a mile - west from its harbour and had a North-South width of 750 metres, half a mile, between its southern edge, the sea, and the high ground to the north. The village on that site today clusters around the harbour and occupies no more than one-tenth of the area of the ancient city, and is no longer called Samos: that name is reserved for the more important newer town over on the north side of the island, where the Piraeus ferry-boats land and where the island's business activities are largely concentrated. Besides tourists, the present village exploits fishing.

Pythagoras left Samos in about 530 B.C., after a possible unpleasantness with the island's ruler. Ten or twelve years later, the `tyrant' of Samos, Polycrates, engaged the engineer Eupalinos to carry out various public works. One of these was to build a mole and modernise the harbour and hence make the port of Samos city more practical, more useable. This Eupalinos did so successfully that the modern mole - and the harbour line too, they say - can be found on top of his foundations. The shape of this mole and the harbour it enclosed were so suggestive when seen from the hills above that sometime during the 25 centuries from then to now the declining town acquired the name `Tigani': Greek for `frying pan'. The military dictators of the country (`the Colonels') in the late 1960s and early 1970s thought this a vulgar name for a place with such a history and changed its name to `Pythagoreion'. Yet still today, while the buses from the main city might say PUQAGOREION, if you get off the ferry boat from Piraeus and ask for the bus to Pythagoreion, you may well be told that `the bus to Tigani leaves from down there'.

Polycrates also engaged Eupalinos on another project: to build a tunnel. The northern walls of the town ran along a mountain ridge about 250-300 metres high, and at the western extremity of this small mountain there was a spring on the north side, called Agiade today. Polycrates wanted the water piped into the town. Eupalinos organised the work so that the tunnel was begun from both sides of the hill and the two teams met in the middle. The estimates for the time required range from 5 to 15 years: the mountain is solid limestone and one has to suppose that many of the slaves doing the work died. The tunnel's existence was recorded by Herodotus (as was the mole and harbour, and the third wonder of the island, the great temple to Hera, thought by many to be the largest in the Greek world). The precise location of the tunnel was only re-established in the 19th century by German archaeologists. The tunnel proper is 1030 metres - 3432 feet - long and visitors can enter it.

I knew some of these things when I first went to Samos in the early 1970s with a friend, and we found the southern entrance to the tunnel of Eupalinos. One could walk as far as one's courage held out, but of course some sort of light was necessary. It was dirty, often very low and also often narrow. It was quite fascinating to me, and I went back to Samos several times over the next 20 years, always taking some sort of torch or flashlight in order to enter the tunnel. Once I went very far and found what seemed to be the meeting point of the two teams of slave-diggers, but my companion and I were rather scared by it all and seeing what looked like a landfall in front of us a little further on we returned to the entrance. (In the village a young man told us that we could have passed that `landslide' however.) Then a few years ago when I went in I found lights running one or two hundred metres into the tunnel; this rather disappointed me. But I was more disappointed when the tunnel became an official site and one had to pay money to get in, and observe the timetable that the Greek Ministry of Culture had established; worse yet, there was a gate after 100 metres and visitors could go no further.

Then last year a friend and colleague from Brussels, Jean Doyen, gave a talk about the tunnel at a mathematical conference (held on Samos, in sentimental honour of the connection with Pythagoras). Jean gave a very plausible explanation, using techniques and instruments known to the ancient Greeks, of how the two teams of diggers were set off in the right directions in the first place: there was no place on the mountain from which both ends of the tunnel could be seen. There is also the problem of levels: a watercourse must obey certain rules. Jean had gathered quite a lot of information about the tunnel, and we decided that it would be a good project to try to enter it and see what we could see. Susan Walker of the British Museum in London, an old friend, put us in touch with Hermann Kienast, of the Deutsches Archäologisches Institut in Athens who had done a lot of modern work on not only the tunnel but a number of other interesting things on Samos; then Alex Bacopolous of the Polytechnic in Athens, and especially his friend Sissi Papathanassiou, helped us with the Ministry of Culture who sent us a letter of permission to pass the 100 metre gate.

Kienast's papers and books were available in Brussels and Jean made a copy of a good bit of the most relevant work. Here we learnt a great deal that we had not realised. For instance, there are two tunnels really: the upper one, that can be walked through, and the lower one, as much as 10 or 12 metres further down, containing the terra cotta pipes for the water, with frequent shafts from the upper one down to the lower one. (In fact Herodotus contains this information too, but I at least did not absorb that: I had somehow drawn the conclusion from my visits that the water course was a few feet down in the ominous ditch that ran on one side of the upper tunnel.) And the lower one is in fact much longer, for it runs several hundred metres through the pine forests on the northern slope of the mountain from the spring Agiade before reaching the tunnel proper, and then runs a fair distance on the south side before finally feeding tanks and cisterns there, inside the ancient city. These extensions of the water-course were probably dug as ditches and then covered over, unlike the tunnel through the hill. Kienast gave quite detailed maps of the lines of these various water-courses, as well as very accurate information about the tunnel - or better, the two tunnels.

So in April of the year 2000 we went to Samos to look at all this. We wanted to walk through the tunnel, at least as far as possible, to find and study the meeting point of the two teams of diggers, to explore the northern side of the mountain and find the spring, and to see whatever else might present itself. Jean also hoped to investigate the ancient walls, a project that I was less enthusiastic about, but only because climbing great hills is no longer so attractive for me. We met at Athens Airport on the Sunday evening, and flew together to Samos, where we had booked a hotel and a car.

The hotel, the `Hera II', was just as I remembered it from last year, up the hill over Pythagoreion with great views from our balconies. Despina, the woman who ran it, was as nice as the year before too. We walked down into the town for the first of the many fish meals we were to have in the next few days, and Jean brought out all the maps and documents that he had prepared. The car that we had arranged didn't seem to exist: calling the office of the car rental company got no answer. (Not such a surprise however, and this being Greece I was prepared for that: in the morning Despina's husband found us another car that cost us 30,000 drachmas for four days: about £14, or $22, a day. That he would so naturally and smoothly solve our problem is also because this was Greece.)

On Monday, when the tunnel is shut anyway, we went scouting around the ancient walls near us, and found a lot of things that I had never seen nor even known about. The walls were sometimes in the back gardens of village houses or country farmers, and the great towers that once stood at important points were maze-like jumbles of enormous shaped stones. We ate fish again, with retsina, in the evening.

Then on Tuesday we went up to the south entrance to the tunnel itself. (I remembered how hard it was to find 30 years ago, in contrast to today when there are signs and even a road.) The guard gave us keys and off we went (with a slight hitch, quickly repaired, since he had given us the wrong keys). The next part is hard to describe, simply because walking in a tunnel is not the sort of experience many of us have often. It varies in size and is normally about square in cross-section, maybe 1.5 m by 1.5 m - that is, 5 ft by 5 ft - or even more. But often it assumes other shapes: sometimes very narrow, just wide enough for one person, and perhaps almost as tall as a man; other times it becomes higher and we could stand up straight in it; however once in a while it is less high, not much over a metre. In my memory of 20 or more years ago, the tunnel was a dirtier and messier place, much narrower, and of course totally dark: today much of it is illuminated by lights along the ground, every 10 or 15 metres, that made a very attractive pattern disappearing in the distance in the places where it ran almost straight for a while. It was always dark: the lights simply broke the darkness. Clearly it has been much cleaned out, and the low sections are places where the floor was covered with earth that probably could be removed some day. On one side there is a discontinuous ditch, and when we peered down this ditch, using our lights, we could see the other tunnel 10 or 12 metres, 35 or 40 feet, below. This ditch had been one of my great fears when I entered the tunnel in the past, but now all the open stretches were covered by a sort of grill: an object (or a small child) could easily fall through, but an adult would have to work at it: this was a great comfort (yet one had to feel that some authenticity had been lost too).

I could hardly stop thinking about the slaves who dug the tunnel: with the tools of the 6th century B.C., it was impossible for me to imagine that it could be done. The problem of how the lower tunnel was dug engaged us, and anyway, why dig an upper tunnel if you're going to dig a lower one that was designed to do the important work (carry water into the city)? One possible answer: the upper one was dug first, then the shafts dug down to give access to the lower level, where a much more accurate line was necessary as well. But the slaves, the horror of knowing what lay ahead of them, either for years or until death: inside the tunnel, the reality of their position hit both of us strongly.

The chapel.

At one point we passed what Kienast's book described as a mediaeval Byzantine chapel: I vaguely remembered it from my earlier visits, but it made less sense then than it did today. Without being told what it was, I doubt if either of us would have guessed. This part was rather wet, with stalactites hanging from the ceiling: this water dripping down is possibly the reason for the chapel, for the water might well have been considered holy or magic. We found the point where the two halves met, and this was interesting.

junction seen from the S junction seen from the W

The tunnel coming from the North side was maybe a metre - almost four feet - higher than the tunnel level coming from the South. The two tunnels meet almost at a right angle, which seems odd, but Kienast had an ingenious explanation: the teams were directed at a certain point to change line, each turning somewhat to its right and then the northern one turning sharply to their left in order to intersect the line of the southern team. This trick was designed to make sure that they did meet. This was about as far as we had gone in the 1970s, for at this point, as well as I can remember, the change in tunnel floor line alarmed us, in addition to the landslide in front just after the meeting point. Even with some lighting today, and the good lights we carried, it was a spooky place.

The tunnel ceiling was much higher for a while after that as we headed north, and here we met a bat. The bat was more upset by us then we were by it, and kept fleeing ahead of us, then turning and trying to get by - which Jean and I certainly wanted it to do. Finally, flattening ourselves against one wall, we freed enough space for the bat to return to what may have been its home. We wondered what the bat could eat in the middle of such a tunnel (when we asked the guard about this after coming out, he told us that there were other creatures in there, for instance snakes). Shortly after the bat the illumination, such as it was, stopped and our lights were our only accompaniment.

About 180 metres from the end of the tunnel (always according to the splendid diagrams and maps of Kienast) we gave up. The tunnel was becoming very narrow, and I in particular was feeling claustrophobic. It would continue to be that narrow, or even get worse, according to Kienast's plans, so Jean accepted my plea that we turn back. Jean is shorter than I am, and had hit his head once or twice, painfully: but I had done it even more often, and had sore bumps on my head for several days. (A Dutch friend in Italy, who used to be a mining engineer, remonstrated with me later: `you guys should have had helmets'. Quite right.) We watched for the bat on the return but didn't see it.

The tunnel had mysterious markings on its walls, mostly made in a sort of red paint. Often they were (Greek) letters, sometimes they were straight vertical lines; occasionally they were indecipherable strings, perhaps words. It's quite astonishing that such marks could survive 25 centuries, but Kienast was clear that they were ancient. We also saw a more modern graffito, almost at the end of the tunnel, where there were no lights: the word `FUCK' burnt onto the ceiling in black, presumably with the flame or smoke from an open light. That mark will probably not last so long. (The guard told us that a party with permission to pass the gate only comes along about once a year, so even the rare dedicated folks who want to enter this strange underground bit of history, and manage to get the necessary permission from the Ministry of Culture, might include some lumpens. That particular word has become internationally known, so it doesn't indicate necessarily that the writer was an English-speaker; also, for a non-Anglophone it doesn't carry the same power - offensive and vulgar, or potent and mysterious - that it has in its home language, it's more a kind of abstract obscenity. Still it's somehow slightly disappointing that people with such poor imaginations want to enter ancient tunnels. Only a few people probably make love here.)

We were very pleased with ourselves after emerging and we chatted for a few minutes with the guard. He said that, during our two hours or less inside, just one other party, a man and a woman, had visited, but they of course had just gone as far as the gate that we had locked behind us, carrying the keys, to let ourselves out. (This was only April, to be sure, and two days later we noticed a tourist bus parked near the guard's little hut, so there was more business; visitors had to pay 800 drachmas normally, but we got in free.) We enjoyed our mid-day meal - fish again, almost at the sea in the little village of Iraeon, where the great temple once stood - in a glow of achievement. I was also puzzling over all the conclusions I had drawn 20 and 30 years ago, conclusions that had turned out to be misleading or dead wrong.

There was foul weather on Wednesday: gusts of rain, a strong wind, and a chill in the air. So we went to the North Coast of the island, to Vourliotes. The daughter of one of our friends had written us about this little village, since her mother-in-law had a house there, and she had recommended a restaurant to us. I had been to Vourliotes before, but the restaurant was new to me: the `Blue Chairs', in the tiny main square of the village. We had a drink in a bar on the other side of the square, thinking that the restaurant was closed. But no, activity was visible inside so I went in to investigate: the owners invited us to eat, and we had a simple but splendid meal. We were their first customers of the year, indeed of the Millennium they thought (but we didn't explain that we thought that the Third Millennium will begin on 1 January 2001 - for after all there had never been a year `Zero', virtual though even year `One' was).

In the afternoon Jean went exploring around Pythagoreion, and I went in the car to see if I could find a road to the north side of the mountain, where the north entrance to the tunnel would be. I went west, passing through Chora and then north towards Mytilini: shortly there was a road turning right and I tried it. Not too bad a road, mostly asphalted, and suddenly on the right side of the road was a crude hand-lettered sign, with AGIADE written on it in rough red letters, pointing down a side road - pointing to Agiade. This road was a real dirt road, and it was necessary to go slowly: but after a bit there was a small group of houses, reminding me of some photos in Kienast's book. And by God, there was a spring: water bubbling out from under a building into a small tank. This was the first spring I had ever seen in Greece, perhaps the first island spring that I had even heard of. And it did indeed exist. I felt wonderful.

I returned to Pythagoreion excited to tell Jean about the spring. He was still out exploring, so I went out too, and managed to find some bits of the ancient city walls inside the modern village. There were walls built 10 or 20 metres in from what is today the harbour, leaving an area, between the port and the city, outside the walls (or maybe not: the line of the harbour front might have been different then). Jean had found things when he returned, and we were quite pleased with ourselves. Our plans for Thursday were obvious: go back to Agiade and then find the northern entrance to the tunnel. We enjoyed our fish that evening in a particularly good mood. One had to speculate about how interesting the life of an archaeologist could be, but then all the information in Kienast's book implied that a lot of digging and grubbing around was done, and so on. Maybe our amateur hunting-trip was sufficient for a couple of mathematicians.

The next day we went around to Agiade and explored a little then set off to find the northern entrance to the tunnel. Although the diagrams and maps we had were very good, it wasn't easy. The land north of the mountain was mostly flat farm country, but south of the spring the first slopes of the mountain began and there are big bushes, then lots of big bushes, and then pine and fir trees. Once in a while we could spot some sign of the ancient underground waterway: a buried ditch with a brick cover, sometimes even pipes buried in the side of the ravine: these pipes must have had something to do with the spring, but we couldn't tell if they were ancient or not. We wandered about quite a lot, each of us with a different opinion about where the north entrance ought to be. Finally we found it: unlike the entrance on the other side, here you could simply walk down and straight in, but we didn't do that, for there was a tremendous number of bees or wasps down in the entrance: another obstacle for the real archaeologist that we could avoid. The entrance had been tidied up and made rather attractive, but it was incongruous and lonely in the middle of a forest: a brick entranceway was built along to the steps that led down into the darkness, where all the bees or wasps were. There was in fact a metal gate, but it was simply lying on the steps (the guard at the south entrance told us later that he knew this: some visitors had torn it down and so far no one had remounted it). Maybe the people who wrote graffiti in the tunnel came this way.

Feeling quite satisfied with ourselves, we went back through the forest to the spring. Near the north entrance we had seen, across the little ravine that ran down from Agiade towards the sea, a very rough dirt road: we thought to return that way, instead of using the much longer route that we had come by. This road was truly awful, and we wondered if we had been foolish: the bloke who owned the car surely wanted it back in one piece. Then along the way we saw an aqueduct in front of us: it passed from the slope of the mountain on our left, the mountain that Eupalinos's tunnel went through, to the other side of the little ravine. Clearly Roman from its construction, and when we looked, we found it marked on our maps. We simply had not noticed it there before.

But why was this aqueduct here? Why did the Romans build it? While it seemed to be on the right level to receive water from the spring Agiade, it didn't look likely that it was to carry water into the city of Samos, now the village of Tigani or Pythagoreion. (The tunnel was used to carry water for perhaps a thousand years, so in the Roman times this aqueduct would have been taking water somewhere else presumably.) It led away from the town, towards perhaps the Temple of Hera, or the small town of Chora. And this recalled an old puzzle that I had thought about years ago: the name `Chora'. This word is widely used as the name of the chief town of an island, or perhaps a district on the mainland, and the actual Chora today is certainly not that, nor does it look like it ever was. But maybe in Roman times it was different: was the old city of Samos already in some sort of decline and a few kilometres to its west, a little higher and, more importantly, a little more distant from the sea and its dangers, this new town was the `capital' for a time? Or maybe the aqueduct was just carrying water out into the plain along the south coast, for irrigation.

A short distance on from the aqueduct the road was barred and we saw the signs of the Greek Army: a young soldier was rushing towards us as we drove up. This was the first person we had seen since noticing some farmers in the fields before arriving at Agiade, and after some linguistic fumbling we all began to speak in English. The soldier had North American inner-city immigrant English I thought, which seemed pretty strange. But such good English I asked him: how come? Because I'm from Toronto, he replied. His was indeed a family of Greek immigrants to Canada, and like many of the young men in that situation, he had to do his military service if he wanted to be able to visit Greece at all. But it's not too terrible, he smiled, and it will be over soon. He told us, apologetically, that we could go no further, it was forbidden. So we turned around and drove back up the awful road (which was worse, and scarier, than coming down it) and all the way around: several miles instead of the 200 yards that this tiny Greek Army camp occupied and that separated us from the Pythagoreion-Chora road.

Other things happened that week. Olympic Airways changed flights and didn't always tell both of us: but that's normal for Olympic Airways. Often they do much worse. I had gone to Athens on the Saturday, one day before Jean, and stayed in the Acropolis House and had my canonical souvlaki meal in Thanasis down by Monastiraki on Saturday evening. Jean stayed on in Athens for a few days after our Friday return from Samos and ate in Thanasis once too, observing later that while it was pretty good, it didn't compare to the splendid fish we had on the island. Quite right: those good Samian fish meals resulted from always going into the kitchen and asking to see what fish they had and then choosing exactly the ones that we wanted (a Greek restaurant-keeper likes you to do this, remembering to be friendly and polite while you do it). Also the island meals were much less dear than meals in Athens, or island meals in the summer: fish cost appreciably more in the summer, which is at least partly a supply-and-demand sort of thing and not only ripping-off-the-tourists. (Summer isn't the best time for fish in the Aegean, and lots of seafood is imported then, frozen, from half-way around the world.) The trip might have been worth it just for the good fish we ate, even without an ancient tunnel and all that it stood for.


There is an astonishing amount of wrong information circulating about this tunnel, which makes me feel less guilty about all the incorrect beliefs that I held for years. The length of the tunnel, where the watercourse was, whether it's passable or not, it purpose even: all these are told incorrectly in one place or another. This we knew before our tunnel trip. However I keep finding little new things, new corrections. For instance, Chora was indeed the capital of the island, but from around 1580 A.D. for maybe three hundred years. Nothing to do with the Romans. At least one source indicates an aqueduct running around the west side of the mountain and into the city, without making it clear if this is our aqueduct or something else. Kienast tells me that there's another spring over beyond the next hill to the west and it's bringing water from that spring to the side of the mountain where Eupalinos's tunnel is, from where it continues towards the city, to supplement the tunnel water, in Roman times. The length of this aqueduct, from its source to the city, was about 15 km and we saw just a small piece of it across the ravine.

Kienast also tells me that while the digging was surely unpleasant work, it's not at all the case that many of the workers died: these were specialists, even if they might have been slaves (and their skills must have been valuable). On each side the digging team would have consisted of about 15 men: 5 digging and the other ten resting. Only two men at a time were actually working on the rock face.

Another old source, not very reliable in many ways, says that `Agiade' is really `Haghiades', which makes some sense, since `Haghia', often abbreviated to `Aghia' or `Agia', is the (feminine) word for `holy' and the usual prefix for a female saint.

Then widespread information is that Aeakes, the father of Polycrates, may have started the tunnel even though Polycrates gets the credit for it on Samos itself, and in most of the guidebooks. It seems agreed too that Polycrates was lured to the nearby mainland by an enemy in the year 522 B.C., where he was skinned alive and then crucified. If that be true, then the unlucky fellow probably wasn't around to see the completion of the tunnel either. There's also the interesting footnote that Epicurus, of philosophical fame, came from Samos, although somewhat later than the time of Eupalinos, and that Pythagoras left the island as early as 535 B.C., exactly because of his opposition to `the tyrant': either Polycrates or Aeakes.

Jean found even more information some weeks after our return. Two British science-historians, Goodfield and Toulmin, visited the tunnel in the late '50s and Goodfield wrote a very interesting piece about the tunnel in the Scientific American (June 1964). (Goodfield found the tunnel narrower than today, and in general in a condition that fits my memory of my first visits, which was reassuring.) Then in 1995, the MAA and CalTech brought out a cassette and an accompanying article, written by Tom Apostol, about the tunnel. There is yet more information in these two sources, as well as considerable discussion of the two problems: how to lay the correct line for the tunnel, and how to get the levels right, digging from both ends. This has made me to conjecture: the lower tunnel was surely dug after (or at least following on) the upper one, and hence the direction problem was important for the upper tunnel, but the problem of levels only needed a rough solution. On the other hand, the lower tunnel, being connected to the upper one by vertical shafts which may have served for a number of purposes (access and disposal of rubble, for instance), had no directional problem, but here the problem of getting the level accurate was critical. However the Greeks had simple devices that could get levels right.

All the better sources of information about Eupalinos's tunnel spend a lot of time on the directional problem, and I at least am not totally convinced by any of the proposed solutions. Yet some technique was used, since the tunnel is there: will we ever know how Eupalinos did it?

Dan Hughes


Text: dan hughes
Photographs: jean doyen
Map+brochure: aeb
Greek letters lost by conversion

See also the article The Tunnel of Samos by Tom M. Apostel.

Appendix by Hans J. Keller

Dan's article has been the most helpful guide when my wife Christine and me visited the Eupalinos tunnel in June 2001. Thanks to his courtesy, I can add this addenda to his page, and I hope it will be helpful to those really interested in exploring the secrets of the tunnel.

The south entrance just above the city of Pythagoreion is now tourist attraction #1 and bus after bus brought hundreds of people, squeaking when passing the narrow tunnel entrance. This was rather disappointing. Nevertheless, it is certainly the first thing to do for those who visit the Eupalinos tunnel for the first time. Close to the entrance is the vertical shaft up to the day that was presumably used for triangulation. Thanks to the electrical lighting, the tunnel gives a nice view and perspective.
As getting a formal permit from the ministry of culture seemed rather difficult and out of reach during our 14 days stay, we decided to look for the north entrance, according to Dan's information and a map from Kienast we had (Fig.2).

We took the road from Mytilinii to the monastery of Triades (That is worth visiting too. There is only one old monk and one nun left there and they are very friendly). After 2.2 km, we turned right at the red AGIADE sign (Fig.1) that is still there. After 500 meters, the dirt road splits. Keeping right, we reached the few houses with the small church shown on Dan's picture. The source is just below the road. An old man was getting water for his mule and recommended the water for drinking.

A few meters above the source, and opening beside the road allows insight into the duct or underground water tunnel (Fig.3).
We followed its direction, crossed the road and headed south into the farmland. Close to a large olive tree, we found signs of a shaft: a stone square on the ground. Most of the duct from the source to the tunnel was built as a so called quanat. Shafts were lowered every 20 to 40 meters that helped excavating the waterway, allowed some daylight and made orientation easy. This method was largely used in ancient Iran. (Fig.4)

The duct crosses the first ravine, (A) on the map, in a stone channel of some 70 x 70 cm cross section, that can easily be found, and some openings allow insight. Then, it turns south-west to turn around a small hill (B), but finally crosses the hill in a tunnel. This tunnel is accessible for some 30 meters through an opening in the ground. In the beginning, it is built with a stone arch, then as a slot in the rock covered with large slate blocks, and finally as a real tunnel. (Fig.5)

After the hill, the duct follows east to the second ravine. This straight part is marked by the quanat shafts. (Fig.6) Many of them look like draw wells of about one meter in square. Once so often, a footpath promised easier walking than straight through the grass and thorny weed, but we always lost the tunnel and found ourselves on a terrace of the farmland. We realised that we better look to the terrain and maintain the level, and accepted that our shoes and socks were clustered with thorny seeds.
At (C), the duct approaches the road that leads to the city hill. The last shaft in the farmland seems to deliver some water and is used as a cistern to these days. Then, the duct crosses the second ravine again in a channel made from large stone plates. This point is directly accessible if one takes a left where the dirt road splits. The road crosses the ravine in a rock bed. We did not attempt to cross it with our rental car. The water duct is just a few meters to the right below the road. Here was a building, and many ruins around it. The road continued south, passed over a small third ravine, that is actually split in two, and inclined toward the hill, on top of which we noticed the ancient walls. If one does not want to follow the duct through the thorn bushes over the second ravine, one can walk up this road and head into the forest only 10 meters after the last ravine. There are some vague traces of footpaths there to the right. After the wood fire of July 2000, it is easier to access than before.

And here, a mere 15 meters from the road, we found the next shaft of the quanat tunnel. This first one was only 2 meters deep and covered with slate stone plates. There are a number of such shafts, becoming deeper and deeper as the hill inclines. Some are covered with wooden branches to prevent wanderers falling to death. Some are reconstructed on the top with stones or bricks, down to the level where they were cut into the solid rock. Where we met a steep rocky side of the small hill in front of us, the footpath (or what seemed to be one) lead rather downwards and towards the ravine on our right. There are also old stone walls from early cultivation that were leading us in the wrong direction.

From the map, we knew that the duct passes straight under the hill (D) so we climbed upwards. And here was the next, very deep shaft, looking like the ruin of a small building, covered with an iron grid. We continued over the hill, and headed directly to the tunnel entrance, that has been secured with large stone walls. Another shaft looked rather like a draw well. After 5 meters of brick reconstruction, a perfectly square shaft followed through the rock, with view to the water duct down there. The last shaft, closest to the entrance, looked like a small round building. For those who cannot find the tunnel, there is even a simpler way: Follow the dirt road south uphill for about 100 meters after the second ravine until there is another dirt road joining from the left (east). To the right, there are traces of a narrow road (or rather a path that was used by vehicles, presumably during restoration), that leads after some 500 meters directly to the tunnel entrance.

Into the Tunnel
We checked our lights and entered the tunnel. There were no wasps or other obstacles. The rusty gate was still lying on the side of the steps. Down the steps, there is a small room, where the water duct is joining the tunnel in the form of a huge, 10 meters high slot. Stone slabs on the walls that look like bookshelves may have served to put the lights on. The tunnel entrance is reinforced with stone masonry that seem to be of roman origin (round top). Although the raw tunnel cross section is the same as more south (some 1.8 x 1.8 m), the clearance of the passage is just large enough to walk through if someone is not too tall. Where this reinforcement starts or ends, it can be easily seen that that gap between the masonry and the tunnel has been carefully filled with perfectly fitting stones for maximum strength. The rock is very brittle in this area.

After some hundred meters, there is steel reinforcement, presumably installed by Kienast. The continuation is made in archaic pointed roof design, using large stone blocks. In some places, the pressure of the surrounding rock was so strong that even this construction was pressed together and the tunnel gets even narrower. The water duct seems to lead directly underneath the passage and is accessible by occasional shafts. The construction changes again to roman stile, and back to the archaic. We are passing some slippery zones with 5-10 cm of water.

In the dry areas, the masonry has been perfectly conserved over the 2500 years. It is easy to see how the tunnel has been carved with irons and hammer, and the stone blocks of the reinforcement look like if they had been cut just recently, with no trace of erosion. Unlike other tunnels and mines we used to visit back home, it was not cold (likely to be the annual average). Only where the tunnel is narrow, we felt the strong air draft that passes, as we filled most of the tunnels cross section. In the larger sections, we could not feel the wind. However, the wind may depend on daytime, weather and other external conditions. We felt comfortable in the tunnel. All dangerous shafts and holes are covered with grids throughout the tunnel. In a few places only, there are lose rocks on the ceiling, and we passed these areas without touching them. The tunnel has been flushed in some places with dirt, clay and gravel brought in by the water through cracks in the rock, what Dan noted as landslide, a not uncommon phenomenon in limestone caves. North of the middle or meeting point of the tunnel, much of these debris have been left in place on the east side of the tunnel, obviously to avoid huge transports. But one can pass comfortably.

I have to say that we are not totally inexperienced in underground explorations. Fanatic to explore old mines since childhood, and member of the Swiss Archaeological Mining Association (or whatever the translation be), we have been in many much narrower, dirtier, colder, and more exposed places. The Eupalinos tunnel can be entered without special equipment, as long as one walks slowly and carefully. We did not have helmets. We just got some spare flashlights and batteries just in case one run out or fell down to the water conduct. Potential visitors should be aware of the conditions they will meet. Bright light, such as a gas lantern may help against claustrophobia, at the risk of burnings. We spent 2 hours in the tunnel and did not think we were in for long.

Getting closer to the middle of the tunnel, the bat mentioned by Dan fluttered by and we found its home in a brownish crack in the roof. There are spiders in the tunnel, but I am sure there are no snakes. The water duct runs here 6-8 meters below the tunnel level, generally only accessible by shafts, but sometimes it is a continuous open ditch. There must have been at least the same quantity of rock to be excavated from this ditch than from the tunnel itself. 50 meters north of the meeting point, the installations of the electrical light start, that were switched off. We chose to visit the tunnel outside official opening hours to avoid hysteria of the tourists on the south side behind the gate. We passed the grid that was between 1986 and 1991 the end of the visitors area and came to the point where the two halves of the tunnel meet, and although we have heard and read much about this section we were very impressed. In fear of missing each other, they have enlarged the north tunnel vertically and curved horizontally toward each other, and it must have been a great satisfaction to the Greek diggers that they met without significant difference. The south tunnel is all in solid rock and with its full profile much spacier and comfortable.

What was even more fascinating to us than how they joined the two halves of the tunnel was how deep the water channel is. Even where this duct joins the tunnel at the north entrance, it is a large vertical slot. A small narrow tunnel would have done the job, specifically as digging in solid rock was certainly expensive and time consuming in those years. My feeling is that the constructors underestimated the slope required for the water to flow, probably because in small scale, water seems to flow horizontally. They probably lost already some 3 meters from the source to the tunnel entrance beyond the inclination they planned for. They had to dig deeper and deeper until they had the water there and the originally small tunnels extended to high slots. We noticed that throughout the tunnel, where visible, there is a ditch about 3 meters deep, that is filled back with stones over most of the length. This was probably attempt number one. As the tunnel is horizontal, the water did not flow either and they had to dig deeper again. The second tunnel 9 meters deeper at south is in the rock 6 meters below the floor of the original ditch and I am convinced that it was not deliberately made, but is a result of digging until they had the water flow.

The water duct exits to the day where the steps lead from the parking lot to the entrance. We did not explore its continuation towards the city. Maybe there is more to discover there.