Imagine two parallel, perfectly reflective polished mirrors. Now imagine a few photons bouncing between the mirrors, exactly perpendicular to the mirrors. Question: will the photons bounce forever? Will they be absorbed eventually by the surface of the mirrors, or gradually lose energy by some mechanism?
At a practical level, what would happen if I stood between two really good, expensive mirrors? Would my reflection persist forever, until something came in between to absorb the photons?
No, I know where this is going, the answer is no.
No
Yes
Even if such perfect reflectors existed, and the photons did bounce back and forth forever, you wouldn’t be able to see the reflected image - because seeing it involves intercepting the photons.
Any question which starts with “What if there were a perfect ____” is always setting up for failure, no matter what goes in the blank, because there’s never a perfect whatever-it-is. One can make extremely good mirrors, that reflect 99.99% of the light that hits them, but even that small percentage that doesn’t reflect adds up quickly, when you’re talking about something that moves as fast as light.
Moderator Note
Thread title edited to more clearly indicate the topic. Please use more descriptive thread titles in the future.
Two mirrors with a photon bouncing between them is actually a fairly important concept to understanding relativity. There’s a brief explanation of it here:
http://www.physicsforidiots.com/relativity.html
In the real world though (which is what the OP is asking about) you have a few problems. As was already pointed out, in the real world we can’t make perfectly reflective mirrors. We can’t make perfectly flat mirrors or make them perfectly parallel either.
And if we somehow could make two perfectly flat parallel mirrors, you have a simple geometric problem in that there is no way to insert a photon from outside of the mirrors in such a way that it will end up bouncing between the two mirrors on a path that is perpendicular to the mirrors.
That makes all of the applications of two parallel mirrors purely theoretical.
Why can’t you just place them far enough part that you have time to move one of the mirrors into place after sending the photon toward the other one?
a - they have to be in a vacuum
b - those buggers move real fast…
Ok, but if we can make perfect mirrors I’m sure we could place them a few 100,000 miles apart in space.
Also, I have no idea how it works or relates to the initial question, but what if there was some Bose-Einstein condensate in between the mirrors slowing down the photons? Do they follow a straight path through that stuff?
“…Lovers—were not the other present, always spoiling the view…”
from the Eighth Duino Elegy
…
We’ve never, no, not for a single day,
pure space before us, such as that which flowers endlessly open into:
always world, and never nowhere without no:
that pure, unsuperintended element one breaths,
endlessly knows, and never craves.
A child sometimes gets quietly lost there, to be always jogged back again. Or someone dies and is it.
For, nearing death, one perceives death no longer,
and stares ahead—perhaps with large brute gaze. Lovers—were not the other present, always
spoiling the view…
draw near to it and wonder. . . .
Behind the other, as though through oversight, the thing’s revealed . . .
…
Translation by J.B. Leishman and Stephen Spender
Wir haben nie, nicht einen einzigen Tag,
den reinen Raum vor uns, in den die Blumen
unendlich aufgehn. Immer ist es Welt
und niemals Nirgends ohne Nicht: das Reine,
Unüberwachte, das man atmet und
unendlich weiß und nicht begehrt. Als Kind
verliert sich eins im Stilln an dies und wird
gerüttelt. Oder jener stirbt und ists.
Denn nah am Tod sieht man den Tod nicht mehr
und starrt hinaus, vielleicht mit großem Tierblick.
Liebende, wäre nicht der andre, der
die Sicht verstellt, sind nah daran und staunen . . .
Wie aus Versehn ist ihnen aufgetan
hinter dem andern . . .
Well, since we’re making stuff up, let’s make one of them a perfect one-way mirror! 
You can make a mirror that comes really close to being perfectly reflective. But you can’t make a mirror that is at all, in the slightest degree, “one-way”. What people really mean by a “one-way mirror” is a partially-reflective window between two rooms, where one of the rooms is dark. Flip the light switches in both rooms, and without doing anything to the mirror itself, it will suddenly appear to be “one-way” the other way.
If you had two perfectly flat, parallel perfect mirrors, and you had a photon travelling between them perfectly aligned, and they were far enough apart that you could slide one of the mirrors in place between bounces, the photon still wouldn’t bounce forever, because it would spread transverse to its propagation direction (i.e. radially). A photon (or electromagnetic wave) of finite extent will still spread transverse to its propagation direction. Eventually, most of the photon would not be hitting the mirror, and it would scatter in some other direction.
The construct of parallel mirrors I intended to use merely to present the problem of what happens to photons as they bounce around a highly/perfectly reflective surface. You can equally imagine a long closed cylinder with the same “perfectly reflective” mirror coating inside to explore this problem. The requirement of parallel surfaces is no longer relevant.
Same question: what happens to the photons - do they bounce around forever or do they lose energy in some way? Would a significant fraction of photons be lost to tunneling? Would they persist for years until some sensor “consumed” them?
No. Taking you strictly at your word, that you have perfectly reflective and perfectly parallel mirrors, and ignoring any fiddle-faddle weaseling about practicality, your photons can still *occasionally * scatter off of each other, a process that will ruin your collimation and allow photons to escape your cavity.
So the mirrors have to be movable, liking play Pong to keep the photon in the box.
Sure, but in an ‘assume spherical cows’ sense, it’s not invalid - if, for example, we want our thought experiment to ignore the imperfection of the mirrors, in order to concentrate on some other thing in the scenario.
Looked this up; this appears to be exceptionally rare, with the LHC detecting about 20 events per year. So sure let’s say about that many photons are duly lost. what about the rest?
I’m not sure what you’re trying to get here. People have given a whole bunch of diferent reasons why photons won’t bounce back and forth forever, and you keep saying “OK, but ignoring that reason, what would happen?”
As far as we know now, photons do not intrinsicly decay, so yes, if you assume you can confine them perfectly inside a sealed, perfectly reflecting cavity, then they’ll stay perfectly confined ‘forever’. Does that answer your question?
Cavity ring-down spectroscopy is based on light trapped between highly reflective mirrors. However, I always think of that more as a standing wave than bouncing particles. Do remember that photons are waves too; they don’t necessarily behave like bouncing balls.
The photons will impart momentum on the mirrors. Not a lot, mind you, but you’ll want them attached to each other.
Law: When the OP poses a thought experiment, one of the first posters in the thread must insist that the premise is impossible, thus completely missing the point of the thread.