The usual derivation of E=mc2 uses the following sloppy mathematics,
about which the student must not murmur. (Actually, no student
dares question this most sacred amongst sacred cows.)
E = c p = c mc(1-v2/c2)-1/2 = c mc(1+v2/2c2)= m c2 + (1/2)mv2
and what is not kinetic energy [(1/2)mv2] is something else, and that
something else is the famous E=mc2. But the third quality [right-hand
side, first line above Ed.], is not an equality. Relativistic enthusiasts
are always letting things move at velocities approaching the speed of
light (c), but note the rotten approximation obtained from the first two
terms of the appropriate binomial expansion (Taylor series, if you like):
(1-v2/c2)-1/2 (1+v2/2c2)
v=0.5c 1.1547 1.1250
v=0.9c 2.2941 1.4050
v=0.95c 3.2026 1.4512
When E=mc2 is used, it is benignly used as an exact expression of physical
fact, not rudely spoken of as an approximation; a bad one at that.
Let me add my school-boy
derivation of E=mc2 which at
least bears somewhat in some
way on the physics of the matter
and avoids the perplexing
notion of converting matter to
energy (whatever that means).
Consider a blob of mass
radiating (giving off) things
called photons (Figure 1), and
let them be given off by shells
of thickness Dx and moving
out from the blob at velocity v=c.
Assuming that Newton got F=ma right, we write the force on the shell as:
F = ma = mDx/Dt2 = m (Dx/Dt)/Dt.
But the energy associated with this motion is:
E = F Dx = m(Dx/Dt2)Dx = m(Dx/Dt)2 = mv2 = mc2.
Once relativists slip (1+v2/2c2) by you, they get clocks to slow down,
measuring rods
to shrink, and other Alice-in-Wonderland things by somehow
replacing v2 in this expression by whatever man can grab, e.g., see
Humphrey's wonders when he sticks in F (p. 101, Starlight and Time,
Master Books).
NOTES AND REFERENCES
1
Soldner, J. 1804. Berliner Astr. Jahrb., p. 161. Reprinted in Ann. der
Phys., 65:593, 1921.
2
E.g., in A Treatise on Electricity and Magnetism, vol. 2, pp. 437-441.