You approach Diu from the water, and this is the first thing the fort teaches you. From the land side there is a gate, a road and a ticket counter. From the sea there is eight metres of stone with the swell breaking against the base of it, and no way in at all. The building was designed to be seen from the deck of a ship, by somebody who had to decide whether to come closer.
That is what a coastal fort is. An inland fort answers the question "how do men get up?" — and the whole vocabulary of Indian hill forts follows from it: scarps, curtains, gateways placed where the approach narrows, water storage because a siege is a contest in patience. A coastal fort answers a different question: "how do ships get past?" The enemy is not a column on a road. The enemy is hulls, and the dimension that matters is not the height of the wall but the width of the water it can reach.
What was actually being defended
Diu is a small island off the southern coast of Gujarat, about fifteen kilometres long and five wide, cut off from the mainland by a creek. Its value was never agricultural. It sits at the mouth of the Gulf of Cambay, on the sailing line between the Gujarat ports and the Red Sea, and whoever held its channel could charge, or refuse, the trade that passed. The fort is on the eastern tip, with the sea on three sides and a channel on the fourth.
It was begun in 1535, and the circumstances are worth stating because they explain the shape. It was not taken by conquest. Sultan Bahadur Shah of Gujarat was facing an invasion by Humayun, and made a defence pact with the Portuguese: in exchange for help, they were allowed to build and hold a fortress at Diu. Two years later the sultan was dead in the harbour during a confrontation with the same Portuguese, and the arrangement collapsed into two sieges — 1538 and 1546, Ottoman and Gujarati forces against the garrison — both of which the Portuguese survived. The fort's reputation was made in those two defences. The alliance that produced it lasted about twenty-four months.
Thickness, not height
The section rebuilt under Dom João de Castro is described in the Portuguese inventories as eight metres thick and some 280 metres long. Eight metres of masonry is not a wall built to keep men out; a wall of two metres does that. It is a wall built to be hit by cannon and not come down — to absorb, spread and dissipate the energy of a ball, and to hold its shape while the outer face is knocked about. Where a hill fort gains protection from elevation, a coastal fort buys it with mass, because on a low island there is no hill to stand on.
Behind the thickness the plan is layered, and the layers are the interesting part. An outer wall follows the shoreline. An inner wall carries the bastions and the guns. Between them runs a moat, and there are two of these: the outer one is tidal, connected to the sea and rising and falling with it; the inner one is cut out of the island's sandstone and stays dry.
The moat is a machine with no moving parts
A dry ditch stops a ladder and a mine. A tidal ditch does something a dry one cannot, and it does it twice a day without anybody being paid. As the tide rises the moat fills; as it falls, the water runs back out through the connection to the sea, and the flow scours the bed on its way. Silt that would otherwise accumulate and give an attacker a ramp is carried out by the same water that brought it.
Read as engineering rather than as scenery, this is a piece of maintenance automation: the fort outsourced its dredging to the moon. And it is a reminder that in this pillar the adversary is often not an army. Here it is the tide, and the design does not resist it — it recruits it. The moat is not a barrier that happens to be wet. It is a barrier that uses being wet to stay a barrier.
Pushing the perimeter into the water
The last piece of the design is the one you can see from the ramparts at dusk: a small fortress standing in the channel itself, about a nautical mile offshore, on a rock with water on every side. The Portuguese called it Fortim do Mar; in Gujarati it is Pani Kotha, the water house. It is a rare thing in India — a fort purpose-built in the sea rather than on the shore.
Its function follows from the geometry. A gun on the tip of the island can reach across the channel; a gun in the middle of the channel can reach along it, in both directions at once, at the point where the fairway is narrowest and a ship has the least room to turn. Put a battery there and the approach stops being a matter of passing a fort and becomes a matter of passing a corridor. It is the same logic as a tower in the middle of a bridge rather than at each end.
The walls did not save it
It is worth keeping the record straight. In 1670 raiders from Muscat took the town and plundered the fort, which is a useful corrective to any account that treats a good section as a guarantee. Fortifications buy time and impose cost; they do not decide. And the end, when it came, did not come through the walls at all: in December 1961 the Indian armed forces took Diu along with Goa and Daman, and four and a half centuries of Portuguese presence on this coast ended in about two days.
The odd afterlife of a fort that stopped working
Today the fort holds a lighthouse whose beam is said to carry thirty-two kilometres, a jail, a garden and a great many cannon lying where they were left. The sea is eating the outer wall in places, and the fabric is a patchwork of seventeenth- and eighteenth-century repairs over a sixteenth-century idea.
This is where Diu closes the pillar's argument from the other side. The coral stone house and the stilt house were replaced, and the working dimension went with them. Diu was not replaced. It was decommissioned, and because it stopped doing its job it could be kept almost exactly as it was. Nothing needed to be modernised about a wall that no longer closed anything. The channel it was built to command is now managed by buoys, by pilots, by a signal on a screen — which is to say, by range and bearing rather than by masonry.
So the fort survives as a shape. Its gradient, its tide and its guns are gone; what remains is a very large, very well-made object that people visit. That is not a failure — it is the most that any piece of infrastructure can hope for, and it is worth being clear that the price of preservation here was obsolescence. The building we can still walk through is the one that stopped being useful in 1961, not the one that was still needed.
Sources and checks
- Portuguese heritage inventory (FCSH, NOVA University Lisbon) Diu — the fortress, its walls, and the Portuguese presence on the Gujarat coast
- Daman and Diu tourism Diu Fort and the Fortim do Mar (Pani Kotha) — site descriptions
- Archaeological Survey of India Diu Fort — conservation status and the fabric of the Portuguese fortifications
- Accounts of the sieges of Diu The Ottoman and Gujarati sieges of 1538 and 1546, and the Muscat raid of 1670
- Histories of the Estado da Índia Diu as a chokepoint on the Gujarat–Red Sea route, and the treaty of 1535
- Coastal fortification studies The design of sea-facing fortifications: fields of fire, tidal moats and the siting of outworks
Last checked 3 November 2026 Yellowsea Huang Corrections