Chronos said:
The used to sell light pinwheels in novelty shops [still do, i suppose]. The 'clockwise' face of each vane is painted white, the opposing face is painted black. It is suspended in an evacuated glass container to eliminate air resistance. When exposed to bright light, the wheel spins clockwise.
These pinwheels actually work on heat differential, not solar pressure. It's virtually impossible (at least with today's technology) to test the effects of solar pressure on Earth. To come close, the evacuated glass container would have to be a perfect vacuum, vs. the extremely low pressure it actually is.
I'm not alone in trying to figure out how to correlate solar pressure with the idea that light will leave a 'perfect' mirror with the same energy it entered.
Thomas Gold (Center for Radiophysics and Space Research, Cornell University) has some doubts, as well.
http://www.arxiv.org/html/physics/0306050
I have some problems with his paper, though. The biggest is that satellites reportedly do experience a force from solar pressure. Solar pressure can cause torques which disturb satellite attitude and changes the eccentricity of geosynchronous satellites, given enough time.
Louis Friedman, working on the Cosmos I, a satellite which will use solar sail propulsion defends the idea that solar pressure does exert a force on satellites:
http://www.planetary.org/solarsail/ss_and_physics.html
I have some doubts about how well the experiment will work so close to Earth and even they realize the observed effects are likely to be very small. (Other effects, such as gravity gradients, tend to dwarf the effect of solar pressure for low orbiting satellites.
If solar pressure does exert a net force on a satellite, then there has to be energy conservation. The light has to lose some energy in transition. This letter comes closest to explaining how energy balances out:
http://solarsails.info/news/newscientistletter.html
Since the energy of light depends on its intensity and frequency, the light reflected off the mirror has to see a reduction in one or the other.