Measuring the speed of unseen light

[altergnostic]

The light-clock thought problem is used to demonstrate time-dilation. From the diagrams, in addition to the accepted assumptions, we can derive the time transform.

Here's a typical diagram:

http://galileo.phys.virginia.edu/classes/252/srelwhat_files/image017.gif

 

It has always bothered me that it assumes we, as observers in the moving frame, actually see the beam bouncing between mirrors, but we don't. Neither the relative position of the observer nor the process by which we observe the beam are ever clear.

 

Also, the diagrams seem to conflate an observer at the origin of a moving frame with an absolute frame in relative motion - the aether.

 

Let's assume the light-clock moves to the right and the beam+mirrors can be described by a right triangle ABC, where A is the point of emission from the bottom mirror at T0, B is the point of reflection on the top mirror at T1, and C is the position of the bottom mirror at T1. The origin is at A and the right angle at C.

 

How do we, as observers in relative motion, get the times and positions of the beam AB? If we were trying to diagram the beam in the aether frame, we could say that, as the beam moves through the aether, it describes AB at c by direct contact, but MM showed we can't check this assumption. We, as observers, would not see such a path. We wouldn't see the beam at all: it goes from one mirror to the other, but where are we in the diagram? We have no defined position yet. If we are not in line with the path of the beam, we can't observe it - we only see light that reaches us directly.

 

Let's place the observer fixed at the origin A. As the beam moved towards B, he has no way to keep track of the beam's position. That's why we always measure the speed if light with a round trip.

 

As it is, the light clock diagram works only by assuming that the path AB would be described at c for the observer, but the observer doesn't even see that beam, let alone know specific coordinates.

 

We can update the setup so that the observer at A receives the beam's coordinates at B with a signal from a stationary emitter placed at B, but if we want to be really rigorous, we must assume c only applies to this signal, as seen from A. We must, because experiment shows that. But experiment doesn't show that an undetected beam, emitted from a moving source and travelling in an arbitrary direction, would also have an observed speed of c. Furthermore, everywhere we have c in SR's transforms, it applies to detected light - light from the observed event that reaches the observer directly.

 

Do you know of any experiment that measures the speed of such a transverse beam? Is there any experimental evidence that shows that light travelling in the rest frame at c would also describe its path at c relative to an observer in relative motion?

[michel123456]

Good question. The light-clock diagram bothers me to.

 

Scenario 1.

 

I am at point A.

I throw a stone to the mirror in front of me and start immediately to run to point C.

 

If the mirror don't break and reflects the object, the stone will come back at point A,

At point c I will miss the stone.

 

Scenario 2.

 

I am on a train moving.

When the train passes through point A, I throw a stone perpendicularly to the rails.

The Stone hits at point B on a gigantic mirror that extends parallel to the rails.

At point C the stone comes back and bumps my head. (i must be very unlucky to manage doing that)

 

Scenario 2 supposes that the stone will have lateral initial velocity.

I wonder what happens for light. Do photons carry lateral motion? (what you call 'transverse beam")

[altergnostic]

Do photons carry lateral motion? (what you call 'transverse beam")

If you want to keep c constant, then there's no lateral component. But then you must redefine the magnitudes of the units of length and time, so that x/t=c even though x is presumably larger. That's what the lorentz transformations do, but back when they were developed, it was assumed that light moves at c wrt the aether.

If you add a lateral component you have violated the constancy of the speed of light postulate, and then you have to explain why it disagrees with experimental evidence here.

 

My point is that experimental evidence says nothing about a beam that doesn't reach the observer directly, as far as I know, and I would like to know if there's any attempt to measure the speed of light in a moving frame, where the light beam being meadured is not directly detected by the observer.

Edited January 6, 2013 by altergnostic