CHAT: Propagation of light

> Date:         Fri, 9 Jun 2000 21:45:25 -0400
> From: Robert Hailman <robert@...>
> Wasn't that 1/300 times as fast? I remember that someone somewhere was
> able to slow down light to more reasonable, but still incredibly fast by
> human standards, speeds. 35 sticks in my mind, but I can't remember what
> the units were... m/s? I remember it involved putting it through a
> certain dense gas at a low temperature. Of course, I read that in
> Popular Science, not the New York Times.
Those are two different experiments. The 35km/h or whatever it was was
in a supercold gas/liquid (same thing at that temperature, which I
think was in the microKelvin range).

The new one used a gas of excited atoms, like the first lasers did.
(This would be quite hot, but I don't know the actual temperature).

My understanding of the news is that the special lasing chamber
locally reverses the phenomenon that makes light move slower in any
other medium than vacuum.

Any pulse of light must necessarily contain more than one wavelength.
(Fourier analysis and all that). However, in very simplified terms,
you can also 'decompose' it into an early part and a late part, both
propagating at the vacuum speed of light --- and what the transmission
medium does is that it 'robs' energy from the early part and gives it
to the late one. (And once the late part grows, its energy starts
being transferred to an even later part). The net effect is that the
peak of energy in the pulse moves ahead slower than each of the parts.

In the lasing chamber, the early part of the pulse is 'allowed' to
borrow energy from the excited atoms, and that debt is then paid by
the late part --- so in effect, the energy peak of the pulse jumps
forward.

Depending on the distance between the parts of the pulse and the size
of the chamber, you should be able to get any apparent speed you want
for the energy peak --- including coming out before it goes in. But
that is no more a real speed than the one you get by sweeping a laser
beam past the moon.

Lars Mathiesen (U of Copenhagen CS Dep) <thorinn@...> (Humour NOT marked)