You're losing the forest for the trees. Maybe check the diagrams I drew again, they seem pretty direct.StevenO wrote:Now you are talking nonsense to me. Your first option is equal to what I described above and should give no fringe since light just moves through the device as a wave with velocity c+v. Your second option also does not give a fringe. I find the bullet model less attractive since it supposedly should fire bullets at an angle if the bullets only move at speed c.altonhare wrote:I could assign the light pulses a velocity relative to the aether, which is always the same, and calculate the velocity relative to the device, which will vary depending on the device's motion through the presumed aether. The result I get, a fringe, is consistent with a "field wave" model of light.StevenO wrote:I see what you mean, but that would mean you would assign the velocity of the device to all lightwaves? That simply means all lightwaves move through the device at c+v and you will find no fringe. Miles mentions this as a valid representation.altonhare wrote: Any individual light wave has the same velocity wrt the device, i.e. the outgoing wave and the incoming wave are 2 different light pulses. We could debate over whether it's one pulse being reflected or a pulse being absorbed and another pulse being emitted, but the more important point is the latter is how I intended the statement.
Alternatively I could assign the light pulses a velocity relative to the pulses' emitters (the device), which is always the same. The result, no fringe, is consistent with a "bullet" i.e. corpuscular model of light.
It was not known a priori which was correct.
First, assume no aether. Sitting in the device we just see a pulse go there and back, whether we are in the parallel arm or the perpendicular arm. No fringe.
Sitting outside the device with v(device) = 0. We see the same thing. Sitting outside the device wth v(device) != 0, we calculate a staggerred arrival of the pulses to their respective mirrors, but it is compensated for on the return trip.
The above is what Miles did.
Now, assume an aether. Sitting in the device we see pulses go there and back. If the device is moving wrt the aether at v, in the parallel arm we calculate c-v/c on the way out and c+v/c on the way back. In the perpendicular arm we calculate sqrt((v/c)2+(x/c)2) on the way there and the way back, where x is the length of the arm.
The best way to see this is in terms of distances-traveled. In the parallel arm, on the way out the pulse must traverse a longer distance through the aether than on the way in to get from source to target, if the device is moving wrt the aether. The distance through the aether is longer by v/c on the way out ad shorter by v/c on the way back in the parallel arm. The distance through the aether in the perp arm is sqrt((v/c)2+(x/c)2) both ways.