Disclaimer: the overall idea of the world as presented is basically lifted from Strugatsky Brothers "Inhabited Island". The details of the question are for a derivative work in that universe.

Imagine a planet whose atmosphere has an unusually large refraction, resulting in the observer getting the impression that the surface of the planet is concave. The upper atmosphere is very dense and opaque, constantly, phosphorescent, resulting in no observable sun or stars from the planet's surface.

If a sentient species with civilization developed on such a planet (without - to date - access to high level atmospheric flight, but otherwise, a 1900-1940 level of overall scientific/technological development - what would be the likely cosmology be for such a species?

Their direct visual experience is that they live on an inner circle of a large bowl or a sphere.

Would they likely be able to detect that they indeed live on an outer surface of a spheroid, by gravitational observation/experiments? Geographical travels?

Would they be able to somehow deduce the existence of extra-planetary space, their solar system/sun, or other astronomical bodies/structures?

Assume that all the physical laws of the universe are 100% identical to Earth, aside from whatever needs to be minimally bent to produce such an unusual atmosphere.

Desired answers will be based on physics/astronomy/cosmology, as well as history of science.

  • 1
    $\begingroup$ This sounds suspiciously like the lead up to the Krikkit wars in Hitch Hiker's Guide to the Galaxy. hitchhikers.wikia.com/wiki/Krikkit $\endgroup$
    – Culyx
    Commented Oct 6, 2014 at 16:44
  • 2
    $\begingroup$ I've posted a question inspired by this one on physics.se, we'll see what that leads to. You could possibly link to that question in your question body. $\endgroup$
    – overactor
    Commented Oct 7, 2014 at 12:23
  • $\begingroup$ I think this is what neutron stars already do. Their gravity is so intense (next only to black holes), that they bend light around them, causing the observer to see a bit beyond the horizon, which creates the impression that they are inside a depression. $\endgroup$
    – Alastor
    Commented Mar 24 at 9:15

2 Answers 2


The Ship in the Distance

One of the early arguments for a spherical earth went something like this:

You are on a large sea, and you see a ship approaching you. Why do you see the masts first, and not the ship all at once? If the earth (and ocean) was flat, you'd always be able to see it, all at once.

Now, if they would see the same ship in the sky, this argument still works. Why can we see the sky-ship all the time, but the real ship shows up as described above? Someone would figure this out eventually, and it would make a good argument for a convex planet surface. I am sure there must be other ways to figure this out, but this is simple and came to mind.


Now, not being able to see stars, their cosmology would be very geocentric. Assuming they could see nothing but the sky above them, they would have no choice but to conclude the planet is the universe. The discovery of the non-visible spectrum is going to be a big deal for them, especially if portions of this non-visible spectrum can pierce the atmosphere. (Odds are, some wavelengths will!)

If they attempted to get into the sky, they may have a big surprise when they get through it! This is especially true if the sky acts like a rather large mirror; earth-people would want to get to the sky-people, only to find out that they don't exist. Some people may observe that such sky-people mirror our every actions, and that could have huge effects on philosophy and religion. It could also be exploited for communication as well, but that's off-topic here.

Tides will, of course, be a huge mystery to them. It was a huge mystery for us, after all, until heliocentrism took off, despite the fact that Seleucus of Seneca proposed the idea around 150 BC.

Since our atmospheres can require certain depths (and temperatures) to refract light correctly, there is a possibility that high mountains may allow them to see (or hint at) the existence of their sun or moon. This could be leveraged to learn about the sun and moon much in the same way that humans have.

Really, the more technology they have, the more likely they'll figure out that they live on the outside surface of a sphere, not the inside of one. Advances in flight and optics will be key to figuring this out.


I believe they would figure out they're really on a ball.

1) At that tech level they understand refraction and would be able to figure out that the atmosphere refracts. Thus they could calculate where the light rays really went.

2) Ballistics. Artillery rounds would go long. Someone would figure out what was up. (Note: There might not be artillery on such a world. Consider Venus--it looks like a bowl. The atmosphere is dense enough artillery isn't going very far.)

3) Gravity. Climb a hill, gravity goes down.

4) The travel time of electromagnetic radiation in the atmosphere. Like with the ballistics issue everything would take a little longer than it should because it's actually taking a longer path than it looks like.

  • $\begingroup$ @TimB: They don't cancel each other out at all. The mass of the mountain is minuscule compared to the mass of the entire planet. It's not added mass, either. When you're standing directly on level ground, you have underneath you the equivalent of that mountain with all the space next to it filled in with additional mass. The difference is very small, though. $\endgroup$
    – Jay Vogler
    Commented Oct 6, 2014 at 20:25
  • 1
    $\begingroup$ I'm not talking about the mass of the mountain. I'm talking about being farther away from the mass of the planet. Climb a mountain, gravity goes down. (That's why high precision scales have to be calibrated to the location they are in.) $\endgroup$ Commented Oct 7, 2014 at 0:08
  • $\begingroup$ Gravity going down doesn't necessarily indicate a spherical world. Moving away from a flat plane results in less gravitational attraction. $\endgroup$
    – HDE 226868
    Commented Oct 7, 2014 at 2:25
  • 1
    $\begingroup$ @HDE226868, but the gradient tells you a lot about the object creating it. $\endgroup$ Commented Oct 7, 2014 at 2:52
  • 1
    $\begingroup$ @LorenPechtel: Correct. My comment was directed at a now-deleted comment by TimB which proposed the opposite. $\endgroup$
    – Jay Vogler
    Commented Oct 7, 2014 at 4:14

You must log in to answer this question.

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