Relativistic Energy

Ask a physicist for the most famous equation in science and you will get $E = mc^2$ . Ask what the $E$ means and the trouble starts, because the equation as usually quoted is not the general relation between energy and mass. It is a special case — the one where the body is standing still. The general statement is $E = \gamma mc^2$ . Put $v = 0$ , so that $\gamma = 1$ , and it collapses to $E = mc^2$ : the rest energy, the energy a body has when it has no motion left to contribute. Every other value of $v$ gives something larger, and as $v \to c$ it grows without bound. Conflating the two — writing $E = mc^2$ and then wondering why a fast particle carries more energy than that — is the single most common error in all of physics pedagogy, and it is committed by people who can do the algebra perfectly well. The deeper question is why energy should have anything to do with $\gamma$ at all, and the answer reframes what energy is. Momentum turned out to be the spatial part of a four-vector. That four-vector has a fourth slot — the time slot — and something must be sitting in it. What is sitting in it is $E/c$ . Energy is not a separate quantity that happens to be conserved alongside momentum; energy is the time component of the momentum four-vector, and its conservation and momentum's are one statement, seen from two angles. Newton needed two conservation laws and could explain neither. Relativity has one, and the $\gamma$ was never optional.

Key takeaways

Signature $(+,-,-,-)$ : $\eta_{\mu\nu} = \mathrm{diag}(+1,-1,-1,-1)$ , so $U\cdot U = +c^2$ and $p\cdot p = +m^2c^2$ .
$E = \gamma mc^2$ is the general relation. $E = mc^2$ is its $v = 0$ case — the REST energy $E_0$ . They are not the same equation and must never be swapped.
Owned result: energy is identified, not defined. $E \equiv c\,p^0$ — the time component of $p^\mu = mU^\mu$ , times $c$ . Because $p^\mu$ is conserved as a four-vector, its time slot is automatically a conserved scalar in every frame. Energy conservation is not an extra law; it is momentum conservation looked at from the time direction.
Owned result: $E$ is frame-dependent; $E^2 - (pc)^2$ is not. Under a boost $V$ , $E' = \gamma_V(E - Vp_x)$ — the energy changes, and can even be made smaller. But $E^2 - (pc)^2 = (mc^2)^2$ is the same number in every frame. Worked below with real numbers: an electron's energy drops $13.33\%$ under a boost while the invariant does not move at all.
Owned result: $E = pc^2/v$ — so a massless particle is forced to travel at exactly $c$ , and anything travelling at $c$ is forced to be massless. It is an if-and-only-if, not a convention.
Key idea: the rest energy $mc^2$ is the term Newton could not see, because it is CONSTANT. Newtonian energy is defined only up to an additive constant, so a term that is the same before and after every mechanical process is mechanically invisible. It becomes physics the moment mass can change — and then it is the largest number in the room.

The core insight

The low-speed expansion $E = \gamma mc^2 \approx mc^2 + \tfrac12 mv^2 + \tfrac38\frac{mv^4}{c^2} + \cdots$ contains the entire history of the subject in one line. Newton found the second term. The third term is a correction he could never have measured. And the first term — vastly the largest — he could not even have detected in principle, because classical mechanics only ever cares about differences in energy, and $mc^2$ never changes in a mechanical process. Relativity's claim is that this invisible constant is real energy, available in full, if you can find a process that changes $m$ . The twentieth century found several.

Where this lives today

Every electron-volt is a use of this equation. The unit itself is defined by the energy an electron gains crossing a one-volt potential difference, and the entire scale of particle physics is quoted in it precisely because rest energies are the natural yardstick: the electron is $0.511$ MeV, the proton $938.272$ MeV, the Higgs boson $125$ GeV. To say "the Higgs has a mass of 125 GeV" is to quote $mc^2$ , and everyone in the field reads it that way without comment. The invariant $E^2 - (pc)^2$ is the workhorse of collider analysis. A detector measures energies and momenta in the laboratory frame — a frame with no special status, in which the interesting collision is usually moving — and the first thing any analysis does is form frame-independent combinations, because those are the only quantities that describe the physics rather than the accident of the laboratory's motion. The mass of a short-lived particle is reconstructed as exactly this norm, taken over its decay products. And the frame-dependence of energy is not an abstraction: it is the design constraint of every fixed-target experiment ever built. A beam of enormous laboratory energy can be feeble in the frame that matters, because most of that energy is the motion of the centre of momentum and is unavailable to make anything new. Quantifying that loss — and the argument for colliders that follows from it — needs the system invariant, which is Invariant Mass's subject.

The physical picture

Momentum is the spatial part of $p^\mu = mU^\mu = \gamma m(c, \vec v\,)$ . Nothing in that construction singles out the spatial part; it is a four-vector, and it has four slots. So write out the one we have been ignoring:

p^0 = \gamma m c \qquad\Longrightarrow\qquad c\,p^0 = \gamma m c^2.

Now, what is this quantity? Three facts identify it, and they leave no room for argument. It is conserved — because the whole four-vector is conserved in a collision, and a four-vector equation holds slot by slot. It has the dimensions of energy. And at low speed it reduces to $mc^2 + \tfrac12 mv^2$ — a constant plus exactly the Newtonian kinetic energy. A conserved quantity, with the units of energy, that reduces to the Newtonian energy in the Newtonian limit, is the energy. There is nothing else it could be. This is worth savouring, because it inverts the usual logic. In Newtonian mechanics energy is a slightly mysterious thing: a quantity you construct, verify to be conserved, and then use. In relativity you do not construct it at all. It was already there, in the slot you were not looking at.

Figure

PUNCHLINE — the mass hyperbola. Every state of one electron lies on the SAME curve E² − (pc)² = (mc²)², and a boost merely SLIDES it along. In frame S the electron has β = 0.6: E = 0.638749 MeV, pc = 0.383249 MeV. Boost by V = 0.8c and it lands at E′ = 0.553582 MeV, p′c = −0.212916 MeV — the energy has FALLEN by 13.33% and the momentum has REVERSED. Yet E² − (pc)² = 0.261120 MeV² at both points, and √0.261120 = 0.510999 MeV = mₑc² exactly. The energy is a shadow that depends on where you stand. The hyperbola is the electron

Look at where $E = mc^2$ actually lives on that figure. It is a single point: the intercept, where the hyperbola crosses the energy axis, where $p = 0$ . It is not the curve. The curve — every point of it — is $E = \gamma mc^2$ . Writing " $E = mc^2$ " and meaning "the energy of a moving particle" is a claim that the whole hyperbola has collapsed to its vertex, and it is wrong everywhere except at one point.

The core equations

Energy is read off the four-momentum, not postulated:

p^{\mu} = m U^{\mu} = \left(\frac{E}{c},\ \vec{p}\,\right), \qquad \boxed{E \equiv c\,p^{0} = \gamma m c^{2}}, \qquad \vec{p} = \gamma m \vec{v} \qquad (1)

where: $m$ — the invariant rest mass, a Lorentz scalar; $\gamma$ — the Lorentz factor of the particle's speed. Physical reading: this is a definition by identification. We do not decide what energy is and then check it is conserved; we observe that the time slot of a conserved four-vector is a conserved quantity with the units and the low-speed limit of energy, and conclude that it is the energy. Setting $v = 0$ costs nothing and gives the most-quoted equation in science:

\boxed{E_0 = mc^2} \qquad (\gamma = 1) \qquad\text{and in general}\qquad \boxed{E = \gamma mc^2 = E_0 + K} \qquad (2)

where: $E_0$ — the rest energy; $K = (\gamma - 1)mc^2$ — the kinetic energy, developed in Relativistic Kinetic Energy. Physical reading: these are two different equations and the difference is the whole point. $E_0 = mc^2$ is a statement about a body at rest. $E = \gamma mc^2$ is a statement about a body in your frame. What $E_0 = mc^2$ actually means — that mass is not conserved, that internal energy weighs — is the subject of Mass–Energy Equivalence. The energy transforms like a component, because that is what it is:

E' = \gamma_V\big(E - V p_x\big), \qquad p'_x = \gamma_V\!\left(p_x - \frac{V}{c^2}E\right), \qquad E^2 - (pc)^2 = (mc^2)^2 \qquad (3)

where: $V$ — the speed of the new frame along $x$ ; $\gamma_V$ — the Lorentz factor of the boost, not of the particle. Physical reading: $(E/c, \vec p\,)$ boosts exactly as $(ct, \vec x\,)$ does — the same matrix $\Lambda$ . So energy is as frame-dependent as time is. What survives is the norm, and the norm is the mass. Finally, dividing the space slot by the time slot:

\vec{v} = \frac{\vec{p}\,c^2}{E} \qquad\Longleftrightarrow\qquad E = \frac{pc^2}{v} \qquad (4)

where: $\vec v$ — the particle's ordinary velocity. Physical reading: $\gamma$ and $m$ both cancel, so this holds for anything, massive or not. Set $m = 0$ : the invariant gives $E = pc$ , and then (4) gives $v = pc^2/(pc) = c$ — exactly $c$ , forced, with no freedom. Run it backwards and $v = c$ forces $m = 0$ . Masslessness and lightspeed are the same fact.

Where the energy comes from

Derivation AEnergy is the time component of the four-momentum, and its conservation is momentum conservation seen sideways

The claim to be established is not a formula but a reclassification, and it is the reason relativity is a simplification rather than a complication. Newtonian mechanics has two great conservation laws, energy and momentum, which are logically independent: you prove them separately, from different assumptions, and nothing in the theory explains why a universe should have both. Relativity has one law. We show here that conservation of the four-vector $p^\mu$ contains both, that the quantity in its time slot is forced to be the energy, and that the two Newtonian laws are the time and space components of a single covariant statement.

Step 1. Write the four-momentum out. $p^\mu = mU^\mu = m\gamma(c,\vec v\,) = (\gamma m c,\ \gamma m \vec v\,)$ . Its spatial part is the relativistic momentum, established in Relativistic Momentum. Name the time part $E/c$ , so that $E = \gamma mc^2$ . Why: at this stage $E$ is just a label for $c\,p^0$ . The content is in the next three steps, which prove that the label is deserved.

Step 2. Establish that it is conserved. In a collision, four-momentum conservation reads $\sum_{\text{in}} p^\mu = \sum_{\text{out}} p^\mu$ . This is a four-vector equation, so it holds component by component — and its $\mu = 0$ component is $\sum_{\text{in}} E_i = \sum_{\text{out}} E_i$ . Why: this is the entire derivation of energy conservation, and it took one line. The Newtonian proof requires work-energy theorems and assumptions about the forces; here it is a corollary of the statement that a four-vector is conserved.

Step 3. Establish that it is conserved in every frame. Under a boost, $p'^\mu = \Lambda^\mu{}_\nu p^\nu$ , and $\Lambda$ is linear. So if $\sum_{\text{in}} p^\mu - \sum_{\text{out}} p^\mu = 0$ in one frame, applying $\Lambda$ gives $\Lambda^\mu{}_\nu\big(\sum_{\text{in}} p^\nu - \sum_{\text{out}} p^\nu\big) = 0$ in every other. Why: this is the deep point, and it explains a Newtonian coincidence. A boost mixes $E$ into $\vec p$ . So if energy alone were conserved and momentum were not, a boost would produce a frame in which energy was not conserved. The two laws must both be true, or neither is — they are not independent, and relativity is what reveals that.

Step 4. Establish that it is the energy, by taking the low-speed limit. Expand $\gamma = (1-\beta^2)^{-1/2} = 1 + \tfrac12\beta^2 + \tfrac38\beta^4 + \cdots$ , so E = \gamma mc^2 = mc^2 + \tfrac12 mv^2 + \tfrac38\frac{mv^4}{c^2} + \cdots The second term is exactly the Newtonian kinetic energy — not approximately, not up to a constant. Why: a conserved quantity with the dimensions of energy whose low-speed limit is the Newtonian energy plus a constant is the energy. The identification is now forced. And the constant is $mc^2$ .

Step 5. Explain why Newton missed the constant. Classical mechanics determines energy only up to an additive constant — only differences are ever measured, and $mc^2$ is the same before and after any process in which mass does not change. It is not that $mc^2$ is small; it is that it cancels. Why: this is the honest reason the largest term in the expansion went unnoticed for two centuries. It is invisible to mechanics not because of its size but because of its constancy — and it stops being invisible the instant a process changes $m$ . *Verified symbolically in verify_aq-sr-relativistic-energy.py: $c\,p^0 = \gamma mc^2$ from $p^\mu = mU^\mu$ with $\eta = \mathrm{diag}(+1,-1,-1,-1)$ ; the series $\gamma mc^2 = mc^2 + \tfrac12 mv^2 + \tfrac38 mv^4/c^2 + \tfrac{5}{16}mv^6/c^4 + \cdots$ with coefficients confirmed by sympy; and the linearity argument checked by boosting an explicit two-in/two-out collision and confirming that energy balance in one frame implies it in the other.*

Go deeper · Derivation B · 2 minThe energy is frame-dependent, the invariant is not, and a massless particle is forced to move at exactly c

Derivation BThe energy is frame-dependent, the invariant is not, and a massless particle is forced to move at exactly c

Two things are true at once and beginners routinely collapse them into one. The energy of a particle is a component of a four-vector, so different observers assign it different values — a fast electron in your frame may be at rest in mine, and we will disagree about its energy by any factor we like. And yet there is a combination of $E$ and $\vec p$ on which every observer agrees exactly. Holding both facts at once is what it means to understand relativistic energy, so we prove them together, with explicit numbers, and then extract the sharpest consequence: that masslessness and travelling at $c$ are the same statement.

Step 1. Boost the four-momentum. $(E/c, \vec p\,)$ transforms with the same $\Lambda$ as $(ct, \vec x\,)$ : $E' = \gamma_V(E - Vp_x)$ and $p'_x = \gamma_V(p_x - VE/c^2)$ , with $p_y, p_z$ unchanged. Why: nothing here is special to energy. It is a four-vector component, and it does what four-vector components do.

Step 2. Form the invariant. E'^2 - (p'_xc)^2 = \gamma_V^2\Big[(E - Vp_x)^2 - \big(p_xc - \tfrac{V}{c}E\big)^2\Big] = \gamma_V^2\Big(1 - \tfrac{V^2}{c^2}\Big)\big(E^2 - p_x^2c^2\big) = E^2 - p_x^2c^2, since $\gamma_V^2(1 - V^2/c^2) = 1$ . The transverse momenta cancel identically. Why: the cross terms $\pm 2Vp_xE$ cancel between the two squares, and the leftover prefactor is exactly $1/\gamma_V^2$ . This is the same one-line miracle that makes the spacetime interval invariant — because it is the same algebra.

Step 3. Evaluate the invariant once and for all. Compute it in the rest frame, where $\vec p = 0$ and $E = mc^2$ : it equals $(mc^2)^2$ . So $\boxed{E^2 - (pc)^2 = (mc^2)^2}$ in every frame. Why: the licence to evaluate an invariant in the most convenient frame and trust the answer everywhere is the single most useful technique in relativistic dynamics. Here the convenient frame is the rest frame, and the answer is the mass.

Step 4. Read off the velocity. From equation (1), $\vec p/E = \gamma m\vec v/(\gamma m c^2) = \vec v/c^2$ , so $\vec v = \vec p c^2/E$ . Both $\gamma$ and $m$ have cancelled. Why: an expression for the velocity that contains no mass will survive the limit $m \to 0$ , where $\gamma m\vec v$ does not. That is exactly what we need next.

Step 5. Set $m = 0$ and watch $v = c$ fall out. The invariant becomes $E^2 = (pc)^2$ , so $E = pc$ . Substituting into Step 4: $v = pc^2/E = pc^2/(pc) = \boxed{c}$ , exactly. Why: this is not the assumption that light travels at $c$ — that was a postulate about light. This is a derivation that anything with zero rest mass must travel at $c$ , whatever it is, forced by the energy–momentum relation alone. Note also what we did not do: we never wrote $\gamma = \infty$ . $\gamma$ is undefined for a massless particle, and no step above used it.

Step 6. Run the argument backwards. Suppose $v = c$ . Then $p = Ev/c^2 = E/c$ , so $E^2 - (pc)^2 = E^2 - E^2 = 0 = (mc^2)^2$ , forcing $m = 0$ . Why: with Step 5 this makes it an if and only if. Zero mass $\Leftrightarrow$ speed exactly $c$ . Nothing with mass can reach $c$ ; nothing without mass can do anything else. The light cone is not a speed limit imposed on massless particles — it is the only place they are allowed to be. *Verified numerically and symbolically in verify_aq-sr-relativistic-energy.py: the boost identity $E'^2 - (p'c)^2 - E^2 + (pc)^2 = 0$ for symbolic $E, p, V$ ; and the explicit case of an electron at $\beta = 0.6$ ( $E = 0.638749$ MeV, $pc = 0.383249$ MeV) boosted by $V = 0.8c$ to $E' = 0.553582$ MeV, $p'c = -0.212916$ MeV, with $E^2 - (pc)^2 = 0.261120$ MeV $^2$ in both frames and $\sqrt{0.261120} = 0.510999 = m_ec^2$ ; and $\beta' = p'c/E' = -0.384615$ cross-checked against the velocity-addition law $(0.6 - 0.8)/(1 - 0.48)$ .*

Figure

PUNCHLINE — E = mc² is the INTERCEPT, not the curve. The true energy E = γmc² starts at 1.0 mc² when the particle is at REST and diverges at β = 1. Newton's ½mv² starts at ZERO. The gap between them at β = 0 is exactly mc² — and it is CONSTANT, which is precisely why two centuries of mechanics never noticed it: classical energy is only ever defined up to an additive constant. The rest energy is not small. It is invisible — until something changes the mass

Limiting cases

Case	Result	Meaning
$v = 0$	$E = mc^2$	the rest energy — the famous equation, and only here
$v \ll c$	$E \approx mc^2 + \tfrac12 mv^2$	Newton, plus the constant he could not see
$c \to \infty$	$E - mc^2 \to \tfrac12 mv^2$	the Galilean world: the rest energy runs away and only $K$ is left
$v \to c$ , $m > 0$	$E \to \infty$	unattainable: no finite energy suffices
$m = 0$	$E = pc$ , $v = c$ forced	$\gamma$ undefined; masslessness $\Leftrightarrow$ lightspeed
$E \gg mc^2$	$E \approx pc\big(1 + \tfrac{m^2c^2}{2p^2}\big)$	ultrarelativistic: the mass switches itself off
any boost	$E$ changes, $E^2 - (pc)^2$ does not	the component varies; the particle does not

Common trapE = mc squared is the REST energy. The general relation is E = gamma m c squared

This is the most-conflated pair in physics, and the confusion is not harmless. $E = mc^2$ holds only when $v = 0$ . A moving body has $E = \gamma mc^2$ , which is strictly larger, and the excess $(\gamma - 1)mc^2$ is its kinetic energy. Two consequences follow immediately. First, you cannot compute the energy of a moving particle from $E = mc^2$ — try it on an LHC proton and you will be wrong by a factor of $7247$ . Second, and more insidiously, you cannot rescue the formula by letting $m$ grow with speed: that is the discredited "relativistic mass", and it fails for reasons set out in Relativistic Momentum. The mass is invariant; the energy is not. If a source writes $E = mc^2$ for a moving body, it is either using $m$ to mean $\gamma m$ — in which case it has abandoned the invariant mass — or it is simply wrong.

Common trapEnergy is a component, so its value is an observer's opinion

There is no such thing as "the energy" of a particle, full stop; there is only its energy in a stated frame. The worked example below boosts an electron and its energy falls by $13.33\%$ — no energy was taken from it, we merely changed our minds about where we were standing. Conservation and invariance are different properties and both matter: energy is conserved (the same before and after, within one frame) but not invariant (different observers get different values). Momentum is the same. Mass is the reverse: it is invariant across frames — but for a system, it is not additive, and that is Invariant Mass's subject. When a problem seems to violate energy conservation, the first question is always: am I adding up energies measured in the same frame?

Worked examples

1 · Boost an electron and watch its energy change while its mass does not

An electron ( $m_ec^2 = 0.510999$ MeV) moves at $\beta = 0.6$ along $x$ in frame $S$ . Frame $S'$ moves at $V = 0.8c$ along $x$ . Find $E$ and $pc$ in both frames, and evaluate the invariant in each.

Step 1 — identify. Energy and momentum are the time and space slots of one four-vector, so they boost with the standard $\Lambda$ . The invariant $E^2 - (pc)^2$ must come out identical, and it must equal $(m_ec^2)^2$ .

Step 2 — frame $S$ . $\gamma = 1/\sqrt{1 - 0.36} = 1/0.8 = 1.25$ . So $E = \gamma m_ec^2 = 1.25 \times 0.510999 = \mathbf{0.638749}$ MeV and $pc = \gamma\beta\, m_ec^2 = 1.25 \times 0.6 \times 0.510999 = \mathbf{0.383249}$ MeV.

Step 3 — boost to $S'$ . $\gamma_V = 1/\sqrt{1 - 0.64} = 1/0.6 = 1.666667$ . Then $E' = \gamma_V(E - \beta_V\,pc) = 1.666667\,(0.638749 - 0.8 \times 0.383249) = \mathbf{0.553582}$ MeV, and $p'c = \gamma_V(pc - \beta_V E) = 1.666667\,(0.383249 - 0.8 \times 0.638749) = \mathbf{-0.212916}$ MeV. The energy has fallen and the momentum has reversed direction — in $S'$ the electron is moving backwards, because $S'$ overtook it.

Step 4 — the invariant, twice. In $S$ : $E^2 - (pc)^2 = 0.638749^2 - 0.383249^2 = 0.408001 - 0.146880 = \mathbf{0.261120}$ MeV $^2$ . In $S'$ : $0.553582^2 - (-0.212916)^2 = 0.306453 - 0.045333 = \mathbf{0.261120}$ MeV $^2$ . Identical.

Step 5 — the boundary number. $\sqrt{0.261120} = \mathbf{0.510999}$ MeV $= m_ec^2$ , exactly. The energy dropped by $13.33\%$ and the momentum changed sign, and the invariant did not move in the eighth decimal place. That is what it means to say the mass is the particle and the energy is only a shadow of it. As a cross-check, the electron's speed in $S'$ is $\beta' = p'c/E' = -0.212916/0.553582 = -0.384615$ — precisely what the velocity-addition law gives: $(0.6 - 0.8)/(1 - 0.48) = -0.2/0.52 = -0.384615$ .

2 · A 1 GeV electron: how much of it is mass?

An electron is accelerated to a total energy of $1$ GeV. Find $\gamma$ , and the fraction of its energy that is rest energy.

Step 1 — identify. "Total energy" means $E = \gamma mc^2$ , so $\gamma$ is a ratio of energies — no speed required.

Step 2 — apply. $\gamma = \dfrac{E}{m_ec^2} = \dfrac{1\times10^{9}\ \text{eV}}{0.510999\times10^{6}\ \text{eV}} = \mathbf{1957}$ .

Step 3 — the split. Rest energy $0.510999$ MeV; kinetic energy $K = E - mc^2 = 1000 - 0.511 = \mathbf{999.489}$ MeV.

Step 4 — the boundary number. The rest energy is $1/1957 = \mathbf{0.051\%}$ of the total. Over $99.9\%$ of what this electron carries is kinetic energy — the mass has become a rounding error. This is the ultrarelativistic regime, and it is why high-energy physicists so often work with $E \approx pc$ and put the mass back only when they must. It is also, read the other way, a warning: for this electron, $E = mc^2$ would understate the energy by a factor of nearly two thousand.

Physics of the portrait

The portrait here is not a person but a term in an expansion — the one that sat at the front of the energy of every object in the universe for two hundred years without anyone noticing, because it never changed.

Origins · 2 minRest energy and how this was discovered

Rest energy

1905–1932 · Einstein's September postscript · the constant Newton could not see · the term that had to wait for the nucleus

The rest energy is the strangest quantity in classical physics: overwhelmingly the largest energy any ordinary object possesses, and completely undetectable by any classical experiment. The reason is bookkeeping, not magnitude. Newtonian mechanics fixes energy only up to an additive constant — you may add a billion joules to every energy in a problem and nothing observable changes — and $mc^2$ is exactly such a constant, identical before and after any process in which the masses do not change. Chemistry does change them, but by parts in $10^{10}$ : burning a kilogram of coal converts about $3\times10^{-10}$ of its mass, far below what any nineteenth-century balance could weigh. So the term was there, in Newton's kinetic energy, hiding at the front of a series nobody had reason to write down. Einstein wrote it down in September 1905, in a three-page follow-up to the relativity paper, and drew the conclusion that a body radiating energy $L$ must lose mass $L/c^2$ . It then took a generation and a new kind of physics — the nucleus, where binding energies reach parts in $10^{3}$ — before anyone could arrange a process that changed a mass by enough to see. When Cockcroft and Walton split lithium in 1932 and compared the energy released against the mass that vanished, the invisible constant finally became a measurement.

History

1905

— Einstein's "Zur Elektrodynamik bewegter Körper" (Ann. Phys. 17, received 30 June) derives the work needed to accelerate an electron and finds it diverging as v approaches c.

1905

— Einstein's follow-up "Ist die Trägheit eines Körpers von seinem Energieinhalt abhängig?" (Ann. Phys. 18, 639; received 27 September, published 21 November) draws out the mass–energy consequence.

1906

— Planck's relativistic dynamics (Verh. Deutsch. Phys. Ges. 8, 136) puts energy and momentum on a common footing through F = dp/dt.

1908

— Minkowski's "Raum und Zeit" makes energy and momentum the time and space parts of one four-vector; E = c p⁰ becomes a definition rather than a discovery.

1909

— Lewis and Tolman (Phil. Mag. 18, 510) reach the same dynamics from collisions alone, without electrodynamics.

1928

— Dirac's relativistic electron equation (Proc. Roy. Soc. A 117, 610) takes both roots of E² = (pc)² + (mc²)² seriously; the negative-energy solutions lead to antimatter.

1932

— Cockcroft and Walton (Nature 129, 649) split lithium with accelerated protons and compare the energy released against the mass difference.

Going further

Texts. Rindler, Introduction to Special Relativity, 2nd ed., §§26–27 · Taylor & Wheeler, Spacetime Physics, 2nd ed., ch. 7 · Griffiths, Introduction to Electrodynamics, 4th ed., §12.2 · Morin, Introduction to Classical Mechanics, ch. 12 · Hartle, Gravity, ch. 5 · Okun, "The concept of mass", Physics Today 42(6), 31 (1989).

Provenance. Signature $(+,-,-,-)$ , so $U\cdot U = +c^2$ and $p\cdot p = +m^2c^2$ . Everything above is recomputed in verify_aq-sr-relativistic-energy.py: $E = c\,p^0 = \gamma mc^2$ from $p^\mu = mU^\mu$ ; the expansion $E = mc^2 + \tfrac12 mv^2 + \tfrac38 mv^4/c^2 + \tfrac{5}{16}mv^6/c^4$ with sympy-confirmed coefficients; the boost invariance $E'^2 - (p'c)^2 = E^2 - (pc)^2$ for symbolic $E, p, V$ ; the massless implication $m = 0 \Rightarrow E = pc \Rightarrow v = c$ and its converse; and the numbers — electron at $\beta = 0.6$ : $E = 0.638749$ MeV, $pc = 0.383249$ MeV; boosted by $V = 0.8c$ : $E' = 0.553582$ MeV, $p'c = -0.212916$ MeV, a $13.33\%$ energy drop; invariant $0.261120$ MeV $^2$ in both frames, $\sqrt{\ } = 0.510999$ MeV $= m_ec^2$ ; $\beta' = -0.384615$ against velocity addition; and a 1 GeV electron at $\gamma = 1957$ with rest energy $0.051\%$ of the total.

Scope. The norm $p\cdot p = +m^2c^2$ and the relation $E^2 = (pc)^2 + (mc^2)^2$ are derived at Four-Momentum and used here. $\vec p = \gamma m\vec v$ is derived at Relativistic Momentum. The kinetic energy $K = (\gamma-1)mc^2$ is stated here and derived from the work–energy theorem at Relativistic Kinetic Energy. What $E_0 = mc^2$ means — that mass is not conserved and internal energy weighs — belongs to Mass–Energy Equivalence. The extended ultrarelativistic treatment is Energy–Momentum Relation's, and the photon's is Massless Particles and the Photon's.

Builds on. Four-Momentum · Four-Velocity · Relativistic Momentum · Four-Vectors. Leads to. Energy–Momentum Relation · Relativistic Kinetic Energy · Mass–Energy Equivalence · Massless Particles and the Photon · Conservation of Four-Momentum.

So is E = mc² wrong?

No — it is right, and specific. It is the correct energy of a body at rest, and it is the equation that tells you mass is a form of energy. What is wrong is using it as the energy of a moving body, which is $E = \gamma mc^2$ . The two differ by the kinetic energy. Einstein's own 1905 statement was about a body at rest, and the modern convention keeps it that way by writing $E_0 = mc^2$ with the subscript. Nothing is being taken away from the famous equation; it is being given its correct domain.

If energy depends on the frame, is it really conserved?

Yes, and the two ideas are independent. Conservation compares before with after, within one frame: in that frame $\sum E$ is the same after the collision as before. Invariance compares one frame with another, at the same moment. Energy is conserved but not invariant; the mass is invariant but (for a system) not additive. The subtle and beautiful point is that conservation would fail if only energy were conserved and not momentum — a boost mixes them, so a frame would exist where the balance broke. The two Newtonian laws stand or fall together, and relativity is what reveals why.

What is the energy of a photon, given that γ is undefined?

$E = pc$ — obtained from the invariant $E^2 - (pc)^2 = 0$ , never from $\gamma mc^2$ , which for $m = 0$ is the meaningless product $\infty \times 0$ . In quantum terms $E = h\nu$ , so a photon's energy is set by its frequency and by nothing else. Note the pleasing consistency: a photon does have energy and momentum despite having no mass, and it is precisely the vanishing of the invariant $E^2 - (pc)^2$ that ties them together as $E = pc$ . Zero mass does not mean zero four-momentum; it means a null four-momentum.

Why does the energy of a moving body diverge as v approaches c?

Because $\gamma$ does, and the physical content is the work–energy theorem: pushing a body faster requires work, and the work needed for each additional increment of speed grows without limit as $v \to c$ . The divergence is not a mathematical artefact to be regularised away; it is the statement that $c$ is unattainable for anything with mass. Note that the speed stays perfectly finite and bounded while the energy runs to infinity — the energy is what diverges, never the velocity. The detailed accounting is Relativistic Kinetic Energy's.

Is the rest energy really available, or is it just bookkeeping?

It is really available, but only in processes that change the mass. In a purely mechanical process the masses are unchanged, $mc^2$ cancels from both sides, and the rest energy is indeed pure bookkeeping — which is why Newton never needed it. Chemistry changes masses by parts in $10^{10}$ , too small for any classical balance. Nuclear processes change them by parts in $10^{3}$ , which is why the rest energy first became visible in a nuclear experiment, and it is the whole story of Mass–Energy Equivalence. In annihilation the conversion is complete: an electron and a positron at rest have $2 \times 0.511$ MeV of rest energy, and they leave behind two photons carrying exactly $0.511$ MeV each.

Why is E = pc²/v such a useful relation?

Because it contains neither $\gamma$ nor $m$ , so it never breaks. It gives the velocity of anything from its energy and momentum — $\vec v = \vec p c^2/E$ — which is exactly what a detector needs, since detectors measure $E$ and $\vec p$ , not $v$ . And it is the clean route to the massless case: with $E = pc$ it returns $v = c$ without ever dividing by zero or invoking an infinite $\gamma$ . Whenever a relativistic formula threatens to produce $\infty \times 0$ , look for the version of it that has cancelled the $\gamma m$ away. Usually there is one, and it is usually the more fundamental statement.