⛏️ Two Teams Enter the Mountain

To speed up construction, excavation of the tunnel began simultaneously from the north and the south. One team entered the mountain from the side of the Agiades spring, while the other began from the side facing the city. Each team had around half a kilometre of solid rock ahead of them – and neither could see what the people working on the opposite side were doing.

This is precisely what made the project so extraordinary. Excavating the tunnel itself required an enormous amount of effort, but the greatest challenge was setting the direction and level for both teams precisely enough for them to meet inside the mountain after working through hundreds of metres of rock.

Herodotus drew attention to this very feature of the structure, describing the tunnel as “amphistomon” – excavated from both ends. He did not, however, explain how Eupalinos achieved it. And this is where the most fascinating part of the story begins.

📐 How Did Eupalinos Know Which Direction to Dig?

This was the most difficult problem of the entire project. Eupalinos could not simply draw a straight line through the mountain and check whether the two entrances lay exactly opposite each other. He first had to determine the route of the tunnel across the surface of Mount Kastro and then transfer that direction underground.

How exactly did he do it? We do not know. Costas Zambas, however, has proposed an interesting reconstruction. According to his theory, a series of wooden surveying structures may have been erected at regular intervals on the slopes of the mountain. Their horizontal beams would have been carefully levelled, with successive points aligned within the same plane. In this way, it would have been possible to establish a line across the surface of the mountain connecting the two sides of the massif, precisely determine the locations of the entrances and calculate the lengths of the planned tunnel sections.

These would therefore not have been “towers” with a rope stretched across the summit. Instead, the structures would have served as a series of surveying points, allowing Eupalinos to maintain the correct direction despite the uneven terrain.

This, however, remains only a hypothesis. No archaeological remains of such structures have yet been discovered, and Zambas himself emphasises that their existence would need to be verified through archaeological investigation of the mountain’s surface.

In this way, Eupalinos may have known where to begin and in which direction to proceed. But an even more difficult problem remained: how could he maintain that line once the teams were deep inside the mountain?

🤝 Meeting Inside the Mountain

At some point, only a thin layer of rock remained between the two teams. The final blows of their pickaxes opened a passage, and the two independently excavated corridors became a single tunnel.

The point where they met has survived to this day and clearly shows that the two sections did not come together like two perfectly aligned lines. The passages join at different angles and at different heights – all the adjustments made during the final stage of excavation can still be seen here.

What matters most, however, is that after excavating roughly a kilometre through the interior of the mountain, without being able to check the position of the other team, Eupalinos succeeded in bringing the two tunnels together.

💧 Where Did the Water Flow?

The large tunnel itself was not the channel through which the water flowed. It served primarily as a working and access passage. Along its eastern wall, a much narrower trench was excavated, and ceramic pipes carrying water from the Agiades spring to the city were laid at the bottom of this trench.

This was necessary because the main tunnel runs almost horizontally. For the water to flow by gravity, the bottom of the trench had to descend gradually – from around 4 metres below the tunnel floor in the north to as much as 8–9 metres in the southern section. This meant that people could move through the main passage while the water flowed several metres beneath their feet.

It was this deep trench containing the ceramic water pipes that formed the true heart of the entire system.

🏺 The Aqueduct Operated for More Than a Thousand Years

The system built in the 6th century BC served the inhabitants of Samos for more than a thousand years. During that time, it was repaired and modified, and the ceramic pipes were replaced when they no longer functioned properly. Over time, increasing amounts of limescale accumulated inside them, reducing the flow of water.

Eventually, the aqueduct ceased to perform its original function, but the tunnel itself continued to be used. In later centuries, it served, among other things, as a refuge for local inhabitants. Fragments of vessels, oil lamps and coins found inside provide evidence of this later use.

Only many centuries later did the forgotten structure once again attract the attention of researchers – and the story of its rediscovery began.

🛡️ When the Tunnel Stopped Being an Aqueduct

Over time, the water supply system ceased to function, but the tunnel itself was not abandoned. In the 7th century AD, when Samos was vulnerable to raids, its interior began to be used as a place of refuge. Additional defensive features were added to part of the tunnel, adapting it to its new purpose.

Archaeological finds – oil lamps, vessels and coins – provide evidence of people sheltering inside. The tunnel, built more than a thousand years earlier to supply the city with water, had therefore acquired an entirely new purpose: to protect its inhabitants.

🔎 How Was the Forgotten Tunnel Rediscovered?

For many centuries, the Tunnel of Eupalinos was almost forgotten. Serious attempts to locate it began in the 19th century. In 1853, the French traveller and archaeologist Victor Guérin discovered the spring at Agiades and a roughly 400-metre section of the underground water conduit leading towards the mountain. He did not, however, reach the main tunnel.

The breakthrough came in 1882, when the monks Kyrillos and Theofanis discovered the tunnel entrances. The first surveys and investigations soon followed – Ernst Fabricius produced, among other things, drawings of the tunnel and of the point where the two teams of ancient builders had met.

A full understanding of the structure, however, had to wait for almost another hundred years. In the 1970s, the tunnel was cleared and systematically investigated by archaeologists from the German Archaeological Institute. It was this research that made it possible to reconstruct the course of Eupalinos’s remarkable aqueduct.

👀 What Will You See in the Tunnel of Eupalinos?

When visiting the tunnel, you will primarily see the narrow, hand-cut passage through which two teams of workers advanced more than 2,500 years ago. In many places, the uneven rock surface, changes in direction and sections of ancient reinforcement supporting the walls and ceiling are clearly visible.

A deep water channel runs along one of the walls – the same trench at the bottom of which the ceramic pipes were laid. In some places, you can look down into it and see just how far it descends below the level of the main passage.

The most fascinating features, however, are the places where the structure reveals how the ancient builders worked: the bends and corrections in the tunnel’s course, markings left on the walls, and the point where the two passages excavated from opposite sides of the mountain finally met. It is here that the solutions described earlier can be seen directly in the rock.

🕯️ How Did They Maintain the Correct Direction Underground?

Establishing the axis of the tunnel on the surface was only the beginning. Once the workers entered deep into the mountain, they could no longer see the surveying points positioned on its slopes. They needed a way to transfer the previously established direction into the tunnel and keep checking it as excavation progressed.

Costas Zambas believes that Eupalinos may have used light for this purpose. On the northern side, a light source could have been placed at night at a carefully determined point on the opposite slope. Looking out towards the entrance from inside the tunnel under construction, the workers would have been able to see it in the distance. As long as the light remained exactly on the established axis, they knew they were excavating in the correct direction. Importantly, the northern light point was also positioned at the level of the tunnel floor, allowing them to monitor both direction and elevation at the same time.

On the southern side, the situation was more difficult because of the considerable difference in ground elevation. Zambas suggests that the corresponding light point may have been placed on a tall wooden structure, approximately 45 metres high, erected in the direction of the sea. Viewed from inside the dark tunnel, the light would have provided a distant reference point.

A problem arose on the northern side. As excavation progressed, water began to enter the tunnel – it had already been present much earlier, but over time the quantity became great enough to interfere with the work. Zambas suggests that this may have been why the floor of the passage was gradually raised, allowing the water to drain away from the area where the miners were working. At one point, the floor level was already around 2.4 metres higher than before.

This created another problem. The rising floor made it increasingly difficult to see the northern light point, while the water increased the humidity inside the tunnel. With poor ventilation, the air became hazy and the distant light could no longer be seen. According to Zambas, the loss of this reference point may have forced Eupalinos to change the original plan for the northern passage.

All of these explanations remain hypotheses; none of these methods has been confirmed archaeologically.

↪️ What If They Missed Each Other?

The closer the two teams came to one another, the more significant even a small surveying error became.

For approximately 393 metres, the southern passage ran almost perfectly straight. Its direction was then changed by about 32° to the east, another 30 metres were excavated, and work from this side came to an end.

The final section of the northern tunnel no longer runs straight. About 30 metres before the end of the southern passage, it begins to change direction several times in the horizontal plane. This final, curving section is 43.9 metres long.

How did the workers know they were getting close? Zambas suggests that during the final stage they may have been able to hear the blows of the other team’s tools and adjust their direction accordingly.

Changes can also be seen vertically. Over the final 27 metres before the meeting point, the northern passage rises noticeably. As a result, its ceiling ended up around 2.5 metres higher than the ceiling of the southern tunnel. Zambas links this difference to the arrangement of the rock layers and concerns about a possible collapse when the two passages broke through into one another. However, he does not explain in detail how the builders protected themselves against the possibility of the two tunnels missing each other vertically.

Hermann Kienast, author of the most important archaeological study of the Tunnel of Eupalinos, interprets this section differently. According to his reconstruction, the height of both passages was deliberately increased before the planned meeting point. On the northern side, the floor was kept at the same level while the ceiling was gradually raised – by a total of around 2.5 metres. On the southern side, the opposite was done: the ceiling level was maintained while the floor was lowered by around 0.6 metres. This created a larger vertical space in which the two passages could intersect even if there was a slight difference in elevation.

Which interpretation is correct? Today, we cannot say for certain. We know the shape of the tunnel, but we can only reconstruct why Eupalinos chose these particular solutions.

What matters most is that the plan succeeded – the two passages were connected inside the mountain. Their course near the meeting point shows, however, that success did not depend solely on perfect surveying accuracy. According to Zambas, some of the deviations visible today were deliberate, while others may have resulted from the need to respond to problems encountered during excavation. Eupalinos was able not only to plan an extraordinarily difficult project, but also to modify that plan when conditions inside the mountain required it.

👷 Who Was Eupalinos?

We know surprisingly little about the man who designed one of the most remarkable tunnels of antiquity. The most important information comes from Herodotus, who names him as Eupalinos, son of Naustrophos, from Megara. We do not know where he acquired his knowledge or what other structures he designed. Nor have any of his plans or descriptions of the methods he used during the construction of the tunnel survived.

Eupalinos worked in the 6th century BC, at a time when Samos was one of the wealthiest and most rapidly developing Greek city-states. For many years, the construction of the tunnel was associated primarily with the tyrant Polycrates, who ruled the island in the second half of the 6th century BC. The matter, however, is not quite so straightforward. Hermann Kienast, who studied the aqueduct for many years, dated the beginning of its construction to around 550 BC based on his research into the monument. If he was correct, work began before Polycrates came to power.

We also do not know whether Eupalinos devised all the solutions used in the project himself or how many specialists worked alongside him. What we do know is that the undertaking required far more than simply supervising workers. The terrain had to be surveyed, the route of the aqueduct determined, the level and direction of the tunnel established, measurements taken as excavation progressed, and solutions found to problems encountered inside the mountain. The traces preserved in the tunnel show that its course was carefully monitored and corrected whenever necessary.

⛰️ Why Did They Have to Tunnel Through the Mountain?

The Agiades spring lay to the north of Mount Kastro, while the ancient city was located on its southern side, by the sea. The aqueduct could have been routed around the mountain, but the route would have been longer and a conduit running along the slope would have been easier to find and destroy during an attack. The Samians therefore chose a much more difficult solution – to carry the water through the interior of the mountain. Costas Zambas believes that security may have been one of the most important reasons for this decision.

The Tunnel of Eupalinos was only one part of the entire system. From the spring, the water travelled approximately 890 metres to the northern entrance, then 1,036 metres through the mountain, before continuing underground on the southern side towards the ancient city.

Tunnel of Eupalinos on Samos – How Did Two Teams Meet Inside the Mountain?

More than 2,500 years ago, two teams of stonecutters stood on opposite sides of Mount Kastro on Samos. Their task was to excavate a tunnel more than a kilometre long – one team working from the north, the other from the south – and meet inside the mountain.

There was no GPS, no lasers and no modern surveying equipment. They had simple instruments, geometry and the calculations of Eupalinos of Megara. After several years of work, the two tunnels did indeed meet.

But how did Eupalinos determine their direction? And why did the Samians undertake such an extraordinary project in the first place?

💧 The City Needed Water

Ancient Samos, present-day Pythagorio, developed at the southern foot of Mount Kastro. As the city grew, it needed increasing amounts of water – for drinking, public fountains and everyday life. The problem was that a good water source lay on the other side of the mountain, in the Agiades area. This was most likely the “great spring” mentioned by Herodotus.

The Tunnel of Eupalinos was therefore not the entire aqueduct, but only its most difficult, central section. From the spring, the water first travelled for about 890 metres through an underground conduit to the northern side of the mountain. It then had to pass through the Kastro massif via a 1,036-metre-long tunnel, before continuing for more than another kilometre on the southern side to reach the city. The entire system was therefore more than 2.5 kilometres long.

It would have been possible to try to bypass the mountain and run the aqueduct along its slopes. However, the route would have been longer, the conduit more difficult to maintain and, above all, easier to disrupt during an attack. The Samians therefore chose a much more difficult solution: to tunnel straight through the mountain and conceal the water supply underground. Zambas believes that security considerations may have been one of the most important reasons for choosing this particular route.

📜 Costas Zambas – an engineer and specialist in historic monuments who participated in the restoration work on the Tunnel of Eupalinos between 2013 and 2016. Based on new measurements and observations, he proposed his own reconstruction of how the tunnel was surveyed and excavated.