Imagine that there's a light ray, with source at point A, and it's directed towards point B (which is very far from point A) and it continues for a huge distance. How will an observer at point B perceive the light ray if it gets moved be a certain angle so that it's now directed towards point C.

If the distance AB is large enough, then according to theory of special relativity "something should happen" either with light ray (will it behave like a stream of water, i.e. like matter = matter at the 'end' of stream will fall behind the matter at the 'start') or with perception of the observer so that the speed of light is not exceeded.

  • $\begingroup$ Please clarify what you mean by "according to special relativity something should happen". $\endgroup$ Apr 25, 2013 at 4:21

1 Answer 1


It's not totally clear what the situation is you're asking about, but it seems to be something like this:

illustration of two rays from A, bent, going toward B and C

If something is causing light to bend - a glass wedge, a gravitational field, magic unicorn powder, fuzzy green fog - then a ray (yellow) initially heading toward B will miss B. It will hit C if there happens to be such a thing in the right place. Anyone at B will not be aware of this yellow ray of light.

However, another ray (orange) leaving A at the right angle, after being bent by whatever, may end up heading just the right direction to arrive at B. An observer at B will see A, not in their view straight across (white dashed line) but appearing above the expected direction.

Also note that they won't see the exactly the right half of A, as expected if light weren't being bent, but rather have a view of the upper right part.

Special and general relativity have nothing to add to this basic geometry, although general relativity might explain the cause of the bend e.g. Schwarzchild geometry - the gravity around any massive object.

Further reading: search for Eddington, Einstein, Africa and eclipse. Try this site: http://undsci.berkeley.edu/article/natural_experiments

  • $\begingroup$ Okay, at the moment when something bends the light ray, should the observer at C notice it immediately? I mean, if i takes time t1 for the light ray to change its direction at the point of deflection and the distance BC is larger then t1 * c ? $\endgroup$
    – Artur Udod
    Apr 23, 2013 at 9:46
  • 1
    $\begingroup$ @Artur: The principle is the same as if you are looking at a torch in my hand and I turn the torch off. You don't find out about it until $x/c$ seconds later (ignoring relativistic effects and assuming we're in a vacuum etc etc). Photons currently travelling in your direction are unaffected by subsequent torch waggling. You may need to edit your question and add a diagram and a more careful explanation of what is puzzling you. $\endgroup$ Apr 23, 2013 at 11:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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