I'm working on this concept. A floating island a mile long and 200 meters wide that would travel eternally in a wide circle between Western Europe and the Americas.

I am going to make some diferente questions about this concept. My preliminary description of this is as follows:

Preliminary description of the floating island

1- Epoch: during the Roman Empire, approximately 100 BC to 300 AD.

2- Manufacture: it would be made of Roman concrete, which can resist the effect of sea water for centuries. Its exterior appearance would be "rocky" and in its interior there would be large hollow compartments to give it buoyancy. Concrete boats exist today and are very strong but not very popular. Modern concrete doesn't last long at sea, but the Roman does.

3- Route: a priori I believe that the circuit of currents of the Atlantic Ocean would allow this island to travel in circles from one side of the sea to another without stranding with only a minimum of human control. I'm not sure if a boat of this weight would be controllable with sails or paddles, but we could have a handwave engine.

4- Economics and politics. It would be inhabited by a small tribe of merchant-navigators. Maybe Phoenicians, Carthaginians or Hebrews. When the "island" approached Hispania they could trade with the Romans using smaller ships and in the Caribbean they would trade with the Amerindians.

4-1- I think that the economy of this small nation would be viable as a bridge of merchandise between the two continents, without cultivating or fishing more than a small percentage of its food.

4-2- I supose that this small nation would be able to defend its island from possible invaders, Romans, Celts or Caribbean, but I'm not completely sure.

5- Ecology: with the help of gardeners, an island like this one could maintain thick vegetation with only large ponds to store rainwater.

Obviously the technological and economic elements of making such a big "island" out of concrete in Caesar's time would be totally handwaved, but I will appreciate your opinions. My first question about this concept is this:

(1) Could this island be made of roman concrete or some other "rocky" material known by the most advanced peoples from the Antiquity?

  • 4
    $\begingroup$ This is a really cool concept but you're asking a lot of different questions at once. This makes your question too broad. Please limit yourself to one question per post. $\endgroup$ – sphennings Oct 17 '17 at 15:58
  • 2
    $\begingroup$ I would consider making the island wider. I would be concerned that at only 200 meters wide, extreme ocean waves might make it subject to being rocked substantially. Not sure if it could actually founder, but the residents would have a tough time of it. Really cool concept. $\endgroup$ – Paul TIKI Oct 17 '17 at 16:17
  • 1
    $\begingroup$ Start with the design & manufacture. That alone would get very long detailed answers. Once you have that, it can inform your questions about life on the island, as far as ecology & supplies. 3rd question should be about the route and trading. A 4th question, could be, given the setup you have (route, ecology, structure) how the society might work, based on merchant ship culture or even on cruise ships! Each of these questions are pretty dense! $\endgroup$ – Erin Thursby Oct 17 '17 at 16:24
  • 2
    $\begingroup$ You have edited to put in a main question but forgotten to remove the other questions. $\endgroup$ – Bellerophon Oct 17 '17 at 16:28
  • 1
    $\begingroup$ Seems like the ecology and residents would need to survive the enormously powerful storms (including hurricanes!) of the North Atlantic on a frequent basis. Quite a design problem. $\endgroup$ – user535733 Oct 17 '17 at 16:32

While concrete ships, boats and even canoes have been made, I'm not clear if people in the ancient world would have the understanding or desire to build ships out of concrete (Roman or otherwise).

The principle of displacement wasn't fully understood until Archimedes discovered it in the period between 287 and 212 BC, so it is unclear if ancient people would realize or recognize that a ship could be made out of materials other than wood, reeds or skin. Certainly the idea that something made of concrete would float would be counterintuitive to people living then (and indeed it is rather counter intuitive for many people today).

The other issue, assuming someone has determined the principle of displacement and developed a means of casting a thin structure out of concrete, is how to assemble the structure. It is impractical to cast a single concrete structure of that size even today, so I would expect the ship to be made out of a large number of modules attached together something like an egg crate or ice tray:

enter image description here

Ice cube tray model of the ship

Each module can be cast as an open topped cube, but the real difficulty comes in attaching each cube to the others in the ship's structure. Attachment points are going to be areas where stresses are concentrated (especially during rough weather) and could also be prone to leakage, meaning a large part of the crew's duties will be bailing out the ship, and constantly moving around and checking attachments, tightening or otherwise shorting them up and fixing packing, gaskets or other seals which are used to keep water out. For that level of technology, I would expect large bolts or lashings to be used to attach the pieces together.

Another issue is that concrete is strong compressively but not in tension. If the pieces are rubbing against each other or get slammed against each other they can break. This is also an issue when considering how the ship is run. Wooden ships have carpenters aboard who can fix the ship when under weigh, but fixing a hole in a module of a concrete ship at sea will be challenging to say the least.

So while this is a "possible" project, ultimately it is both improbable given the technological restraints of the day and impractical as a seagoing vessel.

  • 1
    $\begingroup$ No no no, fixing hole in concrete is not hard. Roman concrete can set underwater, so it's only a matter of canvas to hold it in place, or a matter of holding a "plug" piece in place when glue like layer sets. If anything, it is easier than carpentry. $\endgroup$ – Mołot Oct 17 '17 at 20:29
  • 1
    $\begingroup$ @Molot Fixing a hole with concrete against the pressure of water filling in? No way. You are thinking about filling a hole with concrete in a dry environment, but a hole in the concrete island would be like trying to stop a high-pressure broken pipe by filling it with mud. $\endgroup$ – Rekesoft Oct 18 '17 at 8:07
  • $\begingroup$ And that's why concrete is used in emergency boat repair when dry dock is not possible, right? Because it does not work? <irony mode off> $\endgroup$ – Mołot Oct 18 '17 at 8:14
  • $\begingroup$ Archimedes didn't discover the concept that non-buoyant materials could be shaped to float - ancient peoples were not that clueless (and could see basic reality in front of them). Water clocks comprised of small ceramic or copper bowls with tiny holes gauged to specific inflow rates were used for the consistent measurement of time for thousands of years before that. As far as large concrete boats go, the torsional forces from floating at sea would destroy such a vessel very quickly from the bending and twisting over the waves. $\endgroup$ – pluckedkiwi Oct 18 '17 at 14:31
  • 1
    $\begingroup$ @Mołot that link is a question about concrete patches working not a statement, and it is referring to epoxy cement which is something else entirely. $\endgroup$ – John Oct 18 '17 at 15:12

The possibility of floating islands was well known to the Romans, and the Greeks before them. Delos is a little island that according to myth, floated around the Aegean until it agreed to risk the wrath of Hera and become the birthplace of the twin gods Apollo and Artemis. Poseidon anchored it in gratitude. It was a rich and fancy place for hundreds of years, despite a dearth of any natural resources.

painting of Delos
painting of Delos by Carl Anton Joseph Rottmann; from Wikipedia article

Leto spoke winged words, asking her: "Delos, if you are willing to be the home of my son, Phoebus Apollo, and to found a fat temple,
no other will ever touch you nor forget you. I do not think you will become fat in cattle and sheep, neither will you bear ripe fruits or produce abundant plants. But if you have the temple of Apollo Far Darter, all men will bring you hecatombs and gather here and the insatiable savor of fat will always rise upwards. You will feed those from another's hand those who inhabit you, since your soil is not fat." Thus she spoke.


I propose your Romans do not build the island from scratch, but discover naturally occurring floating islands and then put them together and caulk over them with concrete. Giant rafts of pumice do float the ocean after volcanic events (even today) and maybe this is where the idea of floating islands came from. I am hoping for larger pieces. One needs pumice for roman concrete in any event!

giant floating pumice raft giant raft of pumice from http://volcano.oregonstate.edu/floating-pumice-%E2%80%93-oceanic-hazard

Patching up and reinforcing a natural floating pumice island would require much less concrete than building from scratch.

I think that having the island be a center for ceremony and worship (as done on Delos) might work better than having it be a cumbersome, barely steerable ship. You could definitely have rainwater capture; looking thru images of Delos there is still an ancient cistern there.

I am less optimistic about the possibility of defense. The Irish monks knew about the Vikings and had time to build and the Vikings still came and took all their stuff. That also happened to Delos, eventually and was the end of it. But if the keepers of your island knew the secret of Greek Fire...


Don't plan it. Find it.

Since pumice isn't very strong as parts weather the raft people make repairs using the best materials available. Over decades pumice patched to pumices wears out, but the cement they used to connect parts is still buoyant by trapping air where the pumice was. A much smaller (though still very large) all cement craft is the final product.

  • $\begingroup$ Very imaginative, I'm dubious over the long term stability especialy in a big storm, but one of the best ideas yet $\endgroup$ – Slarty Oct 18 '17 at 20:57

A pontoon city.

It is impossible as a single structure, roll and twist would rip a steel ship even a fraction of that size apart, a concrete one would not stand a chance. If you attached engines too it it would break apart so fast the paint wouldn't even have a chance to dry. It gets even worse if it is hollow. the larger a ship is the more flex it is subjected to. On the other hand, as a mass of linked but independent structure it might be possible. Thousands of independent floating blocks. because it is not rigid it can flex easily.

NSBD's idea about pumice is not bad, small natural pumice rafts have existed. The best way to do it would be to cut many many giant cubes of pumice, drill a few holes through them to run rope and poles and you may have something viable. The bigger they are the deeper they float the less waves will be able to bother them. The same thing could be accomplished with lighter than water concrete, the concrete will float a lot lower in the water as well but it will also cost a fortune to make and concrete like that may be beyond roman technology.

the island's biggest problem is a complete lack of natural resources, they have to import everything, which costs money building this thing would bankrupt an empire and keeping it supplied would require it being one of the largest trading hubs in history because they have nothing to work with. Remember they have import everything but food and maybe water, all the timber all the fuel all the metal all pottery, they have zero natural resources except fish meat, and possibly rain water and some high out put food crop if they are lucky. They cannot even manufacture replacements or repairs for their concrete without outside help.

Lastly your island is not going to make a circuit it is going to end up in the north atlantic gyre where it is a huge pain to get ships to and from. No engine is going to be able to overcome the drag this structure will incur when moving against the currents. they would be better off anchoring it somewhere.

  • $\begingroup$ +1 for bringing up the en.wikipedia.org/wiki/North_Atlantic_Gyre $\endgroup$ – user535733 Oct 18 '17 at 0:23
  • $\begingroup$ Importing materials isn't that much of a concern - their whole purpose is to be an enormous cargo ship, so trade goods they will have plenty of access to. Rather than getting stuck in a gyre (which still leaves them intact if in a difficult to row out of situation), I would be more concerned with running aground, or getting wiped out by a storm. $\endgroup$ – pluckedkiwi Oct 18 '17 at 14:49
  • $\begingroup$ But just like any cargo ship supplies area bigger concern than cargo, and this thing can't pick up or off load directly, and unloading and loading ships is by far the most time consuming portion of early shipping. Made even worse since it does not have a safe harbor for waiting ships. It also costs vastly more to build and maintain than it could ever generate through trade. $\endgroup$ – John Oct 18 '17 at 15:07
  • $\begingroup$ Seawise Giant is an oil tanker that is 1500 feet long. This island would only be a bit more three times that. $\endgroup$ – zeta-band Oct 18 '17 at 19:58
  • $\begingroup$ the seawise giant flexes by multiple feet across its length, doing so to a concrete structure would reduce it to rubble. $\endgroup$ – John Oct 18 '17 at 21:40

I'm going to Crib a little off of Thucydides as far as some of the modular aspects of construction and add a little something else that is possible to address some problems with cement sections banging into each other.

Pykrete might be something that your ancient Roman types could use to help make this floating island possible. Here is how I see this working:

They make a few concrete boats and go sailing to the far North. Of a period of a year, they learn about mixing wood shavings with water and freezing it to form a buoyant and very strong substance. They link the 2 ships with pykrete into a sort of oversized catamaran. This is successful, so they return to Rome and have family begin construction on more concrete boats. When they get back north, they decide think, why have family on a different boat, lets join the new boats to our catamaran, since that worked out. See where I am going here...

The Pykrete will fill the gaps between the concrete boats and act as both a connective material as well as a collision buffer. Once it's frozen, the concrete boats are going to be protected from smacking each other. You could go on, building out year after year. Eventually get to the poiint that the oldest boats are there for structure and storage and build up. As your surface increases you might even be able to move some dirt on top of some of the structure for some basic agriculture. Maybe think about the hanging gardens of Babylon.

What you are going to need, eventually, is some sort of propulsion so you can return to the north to re-freeze everything every so often. I don't know precisely how, maybe with massive sails on masts made from huge trees. Maybe with alchemy or some other handwavium. Sounds like an interesting premise

Additional information on a planned WWII supercarrier from pycrekte is available here Big boats are possible when you realize that Ice is Nice!

  • $\begingroup$ Pykrete does still melt even though not as fast as pure ice does, but there is no way it survives a trip to the Mediterranean, especially not with relying upon sails and currents. $\endgroup$ – pluckedkiwi Oct 18 '17 at 14:54

This is a fascinating idea but sadly is not at all practical. Roman concrete is an interesting and surprisingly strong material, but it would not be suitable for constructing such a large ocean going vessel.

No large structure could realistically be made out of Roman concrete, or any other form of concrete, unless it was steel reinforced. And even assuming the far superior tensile strength of reinforced concrete over unreinforced concrete, such a large structure would be very vulnerable to destruction in a storm due to heave and torsional strains.

As severe storms are not uncommon in the Atlantic I doubt it would last long.

Having a segmented structure would help in one way but would present other serious issues such as how to design joints that can move successfully with 6 degrees of freedom whilst retaining structural cohesion.


So I'm going to attack this problem at points 2,3, and 5 because the choice of building material isn’t the most problematic part. The biggest question is, how big will this thing have to be to support just one person? From there, we can scale. Let's figure out how big an island would have to be to support a single person.

The island/ecology

Now that we've got an estimate for how much we need to keep afloat, we can calculate the size of our 1-person island. Although other commenters have pointed out the need for compartmentalization and structural reinforcement, we're going for a minimum size, so I designed a spherical shell-like boat made of pozzolan. We're adding weight with this shell, but it's also displacing water- which is what we want. The general equation we're solving is $weight\ of\ shell + weight\ of\ soil = weight\ of\ water\ displaced$

Weight of shell

Weight of the shell is complicated, but I assumed a 1m thick wall which simplifies the volume to $4\pi r^2t*\frac{1500kg}{m^3} \approx 9000*r^2kg$

This is the maximum buoyancy we can have- others have proposed modular systems and the like, but this was just a thought experiment to see what's theoretically possible. Anything modular or reinforced will quite possibly never float at all.

Weight of soil

The heaviest thing actually on the boat will definitely be the soil needed to farm. We can avoid the weight of the freshwater ponds by storing it in containers suspended over the side of the ship- that way they'll actually help a little with buoyancy. A person needs ~2000 calories per day. On a boat in the middle of the ocean, we have the advantage of being able to harvest seaweed- in this case, it'd be Sargassum. Not very tasty, but edible, and certainly enough of it. I'll say this knocks our calorie requirement down to ~1000 calories per day, because we need to get other nutrients that seaweed doesn't have. The two plants that come to mind are potatoes and nuts, which grow well together. Unfortunately, these plants weigh a lot because of the soil they require- potatoes require about 10 liters of soil and peanuts about 16. Given the calories in a potato, how many potatoes a plant produces a year, and the soil required by one plant, I end up with 7,000 liters of soil per person per year. For peanuts, by a similar calculation, I end up with 5,000 liters of soil per person per year. With a soil density of 1.5kg/L, that's 18,000kg of soil we need to support and keep afloat. At minimum. For one person.


Now we can return to our original question- substituting the equation earlier with our new numbers becomes $9000r^2kg+18000kg=\frac{4}{3}\pi r^3*1000\frac{kg}{m^3}$ This solves nicely to output 1500m as the radius for our spherical boat topped with soil. Here, I almost gave up on the question- there's no way a pre-modern society would be able to manufacture this, and it'd just be for one person. It does scale back with more people because the soil requirement grows linearly and the volume grows as the cube of the radius, but it's not going to get smaller. However, given that you're okay with a mile-long boat I'll push on.

Ocean currents

You say you've plotted a course that would allow you to float indefinitely in circles, and I'll admit I'm skeptical. I assume you're using the North Atlantic Gyre to circulate, but objects that are suspended in the gyre don't circulate nicely indefinitely- they're deflected into the middle of the gyre by the Coriolis force, which is why we have things like the Great Pacific Garbage Patch and the Sargasso Sea. Check out this video from the UCLA spin lab- it does a great job of explaining it: https://www.youtube.com/watch?v=yP6eG9iXmKc.

However, we've got a boat- a rudder wouldn't work because we don't have a keel, but maybe some sails on the surface could solve that. They'd be only raised when a favorable wind is blowing away from the center of the gyre, and that might be enough to keep you out of it.

The biggest problem with relying on ocean currents is that they are slow. It takes about 5 years for water to circulate around a gyre, and that'd be highly variable. This means that for time spans of multiple years at a time, you'd be in the middle of the ocean and essentially impossible to reach, let alone find. The Canary Current is especially slow- moving only 0.03 m/s means it could take you a couple years just to float from north to south. Even worse, this slow speed means you'd spend a year floating in freezing cold waters near the Arctic, then two years floating near the equator in the dreaded doldrums. That's why the biology would be vital- but your crops would die every time you moved north or south. Perhaps crop rotation could help? If you had two different strains of potatoes or nuts, you could swap them out as you got warmer or colder.


I hope that gives you some perspective on the challenges you're facing. In essence, your floating island would have to be massive, you'd have to cultivate multiple types of crops and manage them perfectly, and you'd have to periodically avoid getting swept into the gyre. Trade would be nearly impossible- nobody will be able to find you, and as soon as a sailing ship leaves to trade with the mainland it'd be impossible to find the way back.

Of course, that assumes everything goes perfectly- but a lot of the other answers in this chain deal with problems like patching the ship or storms. However, those kind of problems are often the ideal focus points for stories such as yours. Good luck!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.