chroot wrote
There is no such thing as a conservation of mass. There is a conservation of mass-energy, the sum of the two combined.
That is incorrect. In fact many derivations on relativity were based on the postulate of conservation of mass. Even if one wishes to define "mass" as "invariant mass" as Taylor and Wheeler do in
Spacetime Physics Second Edition there is still mass conservation. I.e.
http://www.geocities.com/physics_world/stp/pg_249.htm
"Thus part of the mass of
constituents has been converted into energy; but the mass of the
system has not changed/"
By the way, "mass-energy" is a synonym for "relativistic mass"
HallsofIvy
In the general theory of relativity, gravity works by changing space-time itself and has nothing to do with "mass".
That is incorrect. Mass acts as the source of gravity in the same way that charge acts as the source of the EM field. The complete description of mass is the energy-momentum tensor (or mass tensor if you prefer) just as the complete description of charge is the 4-current 4-vector.
Since the time-time component of the stress-energy-momentum (SRM) tensor is proportional to relativistic mass, i.e. T
00 = c
2(mass density), and since the SEM tensor is divergenceless it follows that mass is conserved. But then again that tensor is almost designed that way so one expects that to me the case. So long as mass is defined so that p = mv (p and v are 3-vectors) is conserved then mass is conserved.
For details see
Cosmological Principles, Peacock, Cambridge Univ. Press, (1999). See pages 17-18
http://assets.cambridge.org/0521422701/sample/0521422701WS.pdf
rocketcity
Why is a box of light heavier than a box of dark?
Because the light transmits forces to the walls of the box and it is that transfer of momentum that causes the box to weigh more. This should be measureable in principle. There was a paper published on this subject in the American Journal of Physics. It's called
The mass of a gas of massless photons, H. Kolbenstvedt, Am. J. Phys., 63(1), Jan 1995
chroot wrote
Pressure counts too. It gets wrapped up in the stress-energy tensor, which determines the amount of gravitational curvature.
Pressure is just one component of the SEM tensor. And you need to SEM tensor to describe the inertial mass of a box. Not just to describe the gravitational mass of a body. If the SEM tensor is divergenceless then the mass of the entire system will always transform as M = gamma*M
0.
*Would* this increase the measured weight of the box, to someone looking at a scale outside the box (even assuming an ideal scale)? I'm trying to compare this to *gas molecules* inside a box increasing its weight just by using a kinetic theory argument, i.e., the molecules collide with the bottom of the box more frequently, or at higher speeds, than those that strike the top of the box.
Yes. That is correct.