I have lush fertile mountains separated by miles of flat, barren desert. Each mountain is home to a community. The distances between mountains vary from 2 - 20 miles.

Every occupied mountain has at least one river that drains into the desert and eventually reappears at the sea after travelling underground for many miles. There is a rain cycle.

The inhabitants use water power (not steam) to operate mills and other machinery.

The transport system

Over the centuries, these mountain 'islands' have been increasingly connected by inclined roads. A traveller gets on or loads a cart that rolls from one mountain to another. Braking at the other end is done by rolling partway up the destination mountain. Of course the original height is not reached due to friction along the way.

EDIT - There are two roads between any two cities - their slopes are in opposite directions. That way you can roll from A to B or from B to A.

To complete the journey, the cart is pulled up to the necessary height using water power. Before rolling back, the cart is moved to the new start height and the opposite slope.


The technology is pre-industrial revolution but ball bearings have been invented that reduce the friction of the cart's wheels.


Are there any snags to this system or will it just work as is? Suggested minor improvements/corrections will be welcomed.

  • 4
    $\begingroup$ So what happens if you get halfway down the track from mountain A and mountain B, and a goat wanders onto the track, or a rock was left there by a storm, or you hit a pothole? Is there any way to restart a journey once it's interrupted by inconvenient events? $\endgroup$
    – Cadence
    Mar 9, 2023 at 23:32
  • 7
    $\begingroup$ if you one road in each direction then you have a huge heigh difference, meaning sand will pile up on the low portion of the road and the high one acts like a wind break. Sand moves around in a desert, and it moves a lot. $\endgroup$
    – John
    Mar 10, 2023 at 1:04
  • 1
    $\begingroup$ internal-consistency is the closest to the old reality-check tag that now exists. science-based would probably suit you though $\endgroup$
    – Separatrix
    Mar 10, 2023 at 13:59
  • 4
    $\begingroup$ A gravity-powered system to passively send things downhill screams canal, not road. $\endgroup$ Mar 10, 2023 at 16:40
  • 3
    $\begingroup$ "You need [handwavingly] low friction between the cart wheels and the surface." - This isn't even achievable with modern technology. WORLD'S LONGEST DISTANCE COASTING IN NEUTRAL 38m video but it's all downhill. Can't find furthest coasting on level ground because that's like two blocks no matter what and wouldn't make for good footage. - Quarries use stone rails, but that's for down hill only, and the oxen are only to brake it. $\endgroup$
    – Mazura
    Mar 10, 2023 at 23:41

9 Answers 9


Not feasible

Just for fun, take a bicycle and start pedalling until you are going at 10 km/hr on the flat. Then stop pedalling and see how far the bike will coast before it stops. Repeat the experiment by taking the bike up to 20 km/hr and then see how far you can coast - assuming level ground and equivalent wind conditions, it will be much less than twice the distance of the first test, and it will be nowhere near a kilometre, let alone 2+ miles.

The big problem with the concept is that there is not only rolling friction, but air resistance - and air resistance is proportional to the square of your speed. (I'm assuming that a planet with lush forests and a water cycle has an atmosphere too!) Which means that lots of the potential energy the cart has at the top of a high mountain will be lost in air resistance rather than completely converted into the cart's kinetic energy as it attempts to roll across the gap between mountains. (This is even allowing the infeasible idea that the carts and tracks are so perfectly engineered that the carter does not need to brake at all on the way down.)

In brief, even if the cart's wheels turn with no friction at all, air resistance will make a cart slow to a halt soon after gravity stops providing acceleration.

  • 9
    $\begingroup$ Quasi-relevant XKCD. $\endgroup$
    – Cadence
    Mar 9, 2023 at 22:52
  • 4
    $\begingroup$ @chasly-supportsMonica the question stated that the mountains are between 2 and 20 miles apart - assume this meant at the base, not the peak, but it does not matter. The cart is not going to get more than a few hundred metres/yards past the halfway point. A pre-industrial one mile-long rope that is somehow wound in by water power is barely credible, a ten mile-long rope is not even remotely plausible. (Double rope lengths if they are a loop, like a typical cable car or ski lift arrangement.) $\endgroup$ Mar 10, 2023 at 1:03
  • 2
    $\begingroup$ @Vesper the question was updated with the text you are quoting after this answer was posted. It changes matters somewhat, though the resources and energy required for a pre-industrial society to build a smooth road with constant slope 20 miles long are fairly unbelievable. Will look further, the xkcd what-if that Cadence has referenced makes me suspect it is still infeasible. $\endgroup$ Mar 10, 2023 at 11:22
  • 4
    $\begingroup$ @KerrAvon2055: Does the Roman Empire count as industrial? They had aqueducts that carried water over 30 miles with a near constant, gentle slope. That's stonework and concrete, and very primitive surveying tools. $\endgroup$
    – JRE
    Mar 10, 2023 at 11:48
  • 1
    $\begingroup$ @JRE - that is an excellent point. The rivers could be diverted and instead of carts, i could use boats that float along with the current. $\endgroup$ Mar 10, 2023 at 14:59

Many issues, but workable

When you say "pre-industrial," I have to presume that you mean "pre-railways" since iron rails are an important hallmark of industrialization. If you don't have iron railways, then your friction and wear characteristics of your cart are going to make them unmaintainable. In that case, no, you have no chance of making this work.

If you presume iron railways, then you will need somewhere between a .5% (1:200 grade) and 2% (1:50) drop for the entire distance, depending on how much friction the dust causes. Given that, your plan will work as long as the sand doesn't pack together when it's crushed.

I'm curious how you plan to anchor this. If there's nothing but sand underneath, your trestle will lean towards the windward side in short order. Maybe you're driving the support piles deep into the sand, cross-bracing it, then burying the cross-bracing?

  • 1
    $\begingroup$ 2 miles (3.2km) at 1:200 gives an initial elevation of 16 meters, not too bad, but still quite a bit of engineering. On the other side of the spectrum of your estimatoin, 20 miles (32km) at 1:50 would give 640 meters. Not really feasible, to have such a structure spanning across the desert $\endgroup$
    – njzk2
    Mar 12, 2023 at 15:40

Maintenance issues will make this insurmountable

So you have there mountains, connected by sloping roads. This means that the start of each sloping road should be halfway up the mountain. This in turn means that the ground has to be formed into several hundred meters high slope, which is WAAAY too high for pre-industrial society to be able to operate. Water power is no steam, and to move that sheer amount of mass around requires too long to accumulate, and then there are rains that would deteriorate those roads at quite a decent speed.

The concept, however, is feasible, but with more than carts - you need a good road too, to minimize loss to friction. Railroad would be perfect, yet railroads are right into the industrial revolution, even if discounting engines required to haul cargo, as here the pull would be done by gravity.

And finally, there is an XKCD about that with some data about a much longer road, so maybe tweaking your base concept according to data there could help.

  • $\begingroup$ FYI, pre-industrial railway did exist. I think they were why some conventions exist in later railroads. $\endgroup$
    – Argyll
    Mar 10, 2023 at 22:39

Just about possible, but why? I think it would lose out to other more commercial solutions. It requires a huge amount of community investment in a resource. The nearest equivalent I can think of is the Stockton-Darlington railway, which was 25 miles long. This was profitable because the roads could not take the bulk of the coal to the ships, and the cotton to the looms. Other countries that did not have this bottleneck developed railways later.

If you have pack animals you could cross 20 miles of desert in a day. The Silk Road went for thousands of miles, some of it through desserts with this sort of spacing. Your case needs no public investment: if you had a camel you could take it on.

You could have a water-powered funicular to cart your goods up and down the mountain, but water-powered funiculars seem to have had a brief vogue in the 1880's. There are no ancient ones as far as we know. But your goods are already on a pack animal, so you could save your fare if the funicular existed.


If you want to use water to move goods up and down mountains then barges and locks are your best bet. This is 10th century technology, it just requires digging, water, and simple boats. There are already connecting rivers across your desert too.

If the "deserts" are salt flats, then you could use land yachts enter image description here

but for anything remotely desert like that wont work.

I think it would be awesome to think they got some of these sand walkers to work for transporting goods on rougher terrain, but I suspect they are way too fragile and are barely able to move themselves let alone take a load.


Yes, this would work - but there's a couple of changes to make it work even better.

There's some real life examples of something similar to this where 2 'lifts' are powered by counterweights.

There's an example where Water is used to increase the weight of the descending counterweight, so as to lift the ascending lift. Once it reaches the bottom, the water is returned to a storage tank.

Water Powered Funicular Railway is what I'm thinking of - and probably what would work well in your setting

The only issue you may have is this type of system works really well when you've got a steep incline, but on flatter inclines it isn't really going to work

  • $\begingroup$ Yes, that way of raising the cart at the destination end would be ideal. The free-rolling journey from one mountain to the next would be much more exciting than on a funicular! $\endgroup$ Mar 9, 2023 at 22:39
  • $\begingroup$ Alternatively, you can use a water wheel to provide the mechanical energy, or gravity, or perhaps a combination of the three. $\endgroup$ Mar 10, 2023 at 11:24

Pure gravity rollers will make the roads as tall as the mountains

You need to decide whether there is energy storage/conversion somehow for the carts. Or whether they roll by their own gravity.

If you need them to roll by their own gravity, then it is simple to see why the idea has issues. A 1% grade for 10km wide would be 100m high at the starting point. (100km divided by 100.) Having such a road is basically extending the mountains themselves.

It's impractical to construct and maintain such a tall and bulky structure if it was to be raised above ground purely for the road network, even with modern engineering. So you would need the terrain to naturally form gentle slopes for the transportation method to work. Say your mountains are 500m high. And your gentle slopes between mountains start at 100m up one mountain and gradually descend until reaching the foot of another mountain. The problem ofc is to explain why there would be bi-directional slopes.

You probably need steel railway

Even then, 1% grade is small. You need low friction between the cart wheels and the surface. It probably needs to be railed but, for the carts to continue rolling at 1% grade, it is still probably impossible without steel.

Tall structure or not, you would need a lot of earthwork for railway or any road on desert to work. Sliding directly on sand would be too much friction. You would need constant maintenance for such earthwork. How do the communities arrange for such maintenance? So there needs to be explanations for such maintenance.

Maybe some novel energy storage/release mechanism that drives carts with cables

Now, if you have some cable system where you drop a weight (slowly) in order to pull a cart to some distance, you may be able to do away with a lot of the bulk of the slope that would be required for purely free rolling carts driven by their own gravity. Maybe they mostly free roll part of the way and mainly rely on some cable system the rest of the way.

Also, if your carts drive along natural slopes, they should not be uniformly 1% grade anyway. So having some alternative driving mechanism to overcome flat parts or even rising parts makes more sense regardless of how the slopes come about.

How big are these ramps?

Having ramps of height much more then a few tens of meters is beyond practical for most preindustrial peoples. The great pyramid of Cheops is 150m in height. These ramps would be at least double that in height, and many times the length especially if trying to avoid air friction.

Egypt comparison

Egypt had and has a major transport route which historically allowed for a single shared religion and strong state building. A strong state combined with a highly concentrated population that for periods of the year could spend time on optional projects, with major transport route that could supply a large workforce with food and supplies.

The mountain islands will have at most a strong kingdom per mountain. More likely multiple regional powers per sub-valley due to transport difficulties. So not strong state, no unifying transport.

Yet question calls for project that will need cooperation of many kingdoms over decades, maybe centuries.

If the ramps actually extend for kilometers then the costs of completions will go up to centuries.

Track maintenance

There will be debris or even malicious opportunistic placed obstacles. How much effort will be required to be spent to maintain the track? Likely required to have a set of workers permanently keeping the track clean. Who provides workers at the ends to assist travelers? With forest mountains as frequent as the question proposes, drifting sand is unlikely, but with the very low frictions required any and all debris must be cleaned off.


If this were magically built as planned, how much would it speed things up? It would be faster for couriers and a limited number of merchants, limited due to not being able to send many vehicles per hour to avoid collisions. A caravan over good roads can do something like 12 to 20 km a day. So maybe an hour with ramps vs a day without.

Techs needed to lower death rate.

Low friction bearings

Some civilizations didn't have wheels, it takes a bit to be able to build tolerably low friction wheels and bearings. It takes even more to have high quality low friction bearings. High speeds without good bearings will mean spectacular failures every so often. Enough to be a known thing that some will refuse to use the system.


Without some means of keeping the vehicles on the ramp they will fall off resulting in death and destruction. so this requires flanged wheels and rails or some other guide arrangement.

Rack and pinion steering

Alternative to rail would be steering that has mechanical centering via feedback. Which is not easy to invent or build. Poor implementations will result in deadly failures. Without the self centering positive feedback loops and no rails vehicles encountering a track non perfection frequently go off the track.

Project completely infeasible.

The project is far too expensive for multiple reasons.

  • Requires multiple near or post industrial technologies to avoid high death and or loss of vehicles.
  • Requires ramps that will cost enormous extremely long term investment
  • Considering a centuries long construction time and very limited benefits, The return on investment is below zero almost certainly. There is some benefit of look how awesome we are if it is completed.
  • Requires long term multi-generational international cooperation of sovereigns
  • Requires strong unified states with good internal transport to be able to construct and maintain.
  • Requires different geography then what is described in the question.The mountain countries will not be unified enough to need better transport.

The extreme costs and the cooperation required are going to be deal breakers. Distant secondary is insufficient tech to make it safe.

  • $\begingroup$ Hmm, you make many good points. I need to rethink. $\endgroup$ Mar 13, 2023 at 23:15

Use aqueducts, not roads.

As a previous answer states, carts on a track can work at a grade of somewhere between 1:200 and 1:50... but this requires industrialization to make that much steel. But even if you could get that kind of efficiency, that still means that to go 20km you need a ramp that is 100-400m tall which again would be impossible to build with pre industrial technology.

... but since you're already using water power so heavily, you should lean into that. An aqueduct can maintain a good flow at much lower grades closer to 1:1500. This means that to travel 20km, your starting point only needs to be about 13.4 meters higher than the ending point. We know this is doable using pre-industrial technology because the Romans did it over 2000 years ago. So, you could collect melting snow from each mountain into reservoirs that are above the starting point, and open them up to create a sort of artificial river on which to place small boats that can then be lifted up on the far end powered by the outflowing water.

To conserve water, you can cork the flow

Perhaps there is a compromise between wheels and boats that would make since though. Running aqueducts big enough for your boats 24:7 will consume a lot of water, maybe more than you have. So you may want a solution that only uses as much water as you need. The problem with carts is that the heavier you make the cart, the more inertia you need to overcome to get it moving, but if your cart blocks the flow of water, then you can add "weight" without adding inertia allowing it to roll under the pushing force of the water behind it, even at a very low grade.

So your cart could block the flow, you release water to flood the aqueduct behind the cart, and the weight of that water will push it despite the very low grade but without wasting all the water you would need to float a boat on.

enter image description here

Then if you need to send a large payload at once, you can even float boats in the built up water behind the cart and they will move like a train.

enter image description here

  • $\begingroup$ "if your cart blocks the flow of water, then you can add weight without adding inertia" - what? How are you drawing this conclusion? $\endgroup$
    – Cadence
    Mar 14, 2023 at 16:34
  • $\begingroup$ @Cadence I may have explained this oddly, but if you have a cart that is 1 ton on a shallow slope, and place 9 tons of water in it, then the weight that wants to move the cart forward is 10x as high, but so is the inertia of the cart; so, its weight does not impact its ability to move at a given incline. But if you put 9 tons of water behind the cart, you also have ~10x as much forward force, but the water does not add to the cart's inertia, it follows fluid resistance rules now. Thier resistances to move will average out based on the ratio of mass. $\endgroup$
    – Nosajimiki
    Mar 15, 2023 at 14:32
  • $\begingroup$ So a 1 ton cart with 9 tons of water behind it will move much more like flowing water than like rolling cart. $\endgroup$
    – Nosajimiki
    Mar 15, 2023 at 14:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .