Particle Decay Kinematics

A neutron at rest decays. If it went to a proton and an electron and nothing else, the kinematics would leave no freedom at all: two bodies, back to back, with momenta fixed by the three masses. The electron would come out at 781.6 keV, every single time. Not "on average", not "up to" — a line, as sharp as the neutron's mass. James Chadwick looked, in 1914, and found a smear. The electrons emerged with every energy from nearly zero up to about $780\ \mathrm{keV}$ , and the shape was a smooth continuum with no line anywhere in it. Something was wrong at a level that could not be patched. Ellis and Wooster settled it with a calorimeter in 1927: the energy really was missing, on average about two-thirds of it. For sixteen years the continuous $\beta$ spectrum was the most serious open wound in physics, and Bohr was prepared to abandon conservation of energy to close it. Pauli closed it instead, on 4 December 1930, with a letter to a conference at Tübingen proposing "a desperate remedy" — an unseen neutral particle carrying off the balance. His argument was this article's, run backwards. A two-body decay from rest produces a fixed, monoenergetic momentum. A three-body decay does not. The observed spectrum was a continuum, therefore there were three bodies, therefore there was a third particle. And the endpoint of that continuum — the energy the electron reaches when the invisible particle happens to get nothing — sits exactly at the line the two-body decay would have produced. The same fixedness, seen from the other side, is how particles get discovered. When Lattes, Occhialini and Powell photographed pions stopping in nuclear emulsion in 1947, every secondary muon ran the same distance, about 600 μm, before stopping. Identical range means identical energy means a two-body decay. The pion's daughter is monoenergetic, and so, at $4.12\ \mathrm{MeV}$ , is every muon in every one of those photographs.

Key takeaways

Signature $(+,-,-,-)$ : $p\cdot p = +m^2c^2$ on shell. $s$ is in energy² units, $\sqrt{s}$ an energy.
A decay is a collision run backwards: four-momentum is conserved, so $p_M^\mu = \sum_i p_i^\mu$ , and the parent's rest frame is the CM frame of the products ( $\sqrt{s} = Mc^2$ ).
Owned result — the daughters are monoenergetic. For $M \to m_1 + m_2$ from rest, the two products are back-to-back with a momentum fixed entirely by the three masses:

p^{*} = \frac{\lambda^{1/2}\!\big(M^2c^4,\ m_1^2c^4,\ m_2^2c^4\big)}{2Mc^3}, \qquad E_1 = \frac{\big(M^2 + m_1^2 - m_2^2\big)c^2}{2M}

with $\lambda(x,y,z) = x^2+y^2+z^2-2xy-2yz-2zx$ the Källén (triangle) function. There is no freedom left but the orientation.

Why it matters (1): a two-body decay gives a LINE; three bodies give a CONTINUUM. The continuous $\beta$ spectrum is therefore a proof that a third particle exists — which is how the neutrino was found.
Why it matters (2): the invariant-mass peak. Add the daughters' lab four-momenta and take the norm and you recover $M$ exactly, in every event, at any parent speed. Unrelated particles give no such number. That is why every particle discovery is a bump in a histogram.
$Q = (M - m_1 - m_2)c^2$ is the released energy, and it is shared in a fixed ratio — the lighter daughter takes almost all of it.
Key idea: a two-body decay from rest has ZERO kinematic freedom. That rigidity is not a curiosity — it is the signature by which particles are identified, and its ABSENCE is how the neutrino was inferred.

The core insight

Count the unknowns. Two daughters have six momentum components; four-momentum conservation imposes four equations; the two mass shells impose two more. Six constraints, six unknowns — the system is exactly determined, up to the three rotations that orient the whole thing. So a two-body decay from rest has no kinematic freedom whatsoever: the energies and the momentum magnitude are functions of the three masses alone, and nothing about the dynamics — the interaction, the matrix element, the lifetime — can change them. Add one more daughter and you add three unknowns while adding none of the constraints, and the rigidity collapses into a continuum. Everything in this article is a consequence of that count.

Where this lives today

Every particle discovery of the last fifty years is a picture of equation (4). You take a candidate set of decay products, add their measured four-momenta, take the norm, and put the answer in a histogram. Real parents pile into a peak at their mass; combinatorial accidents spread into a smooth background. The $J/\psi$ announced itself in 1974 as a spike at $3.097\ \mathrm{GeV}$ so narrow it stopped a conference; the Higgs boson was announced in 2012 as a bump at $125\ \mathrm{GeV}$ in the invariant mass of two photons. The physics that makes both plots readable is the rigidity derived above: kinematics contributes no width, so every feature of the peak is either the detector or the particle's own lifetime. The absence of that rigidity is just as diagnostic. Missing transverse momentum — the failure of the reconstructed four-momenta to balance — is how neutrinos are detected at hadron colliders today, and it is Pauli's 1930 argument with better instruments. When a $W$ boson decays to a lepton and a neutrino, the neutrino is never seen; what is seen is that the visible products do not add up, by an amount that reconstructs to the $W$ mass only when the missing piece is included.

The physical picture

Work in the parent's rest frame, which is where a decay is simplest and where it is also the CM frame of the products: $\vec P_{\mathrm{tot}} = \vec 0$ , so $\sqrt{s} = Mc^2$ (Centre-of-Momentum Frame owns the machinery; this article is its most rigid application). Momentum conservation then says the two daughters are back-to-back with equal momentum magnitudes — there is nothing else they can do, because their momenta must sum to zero. Energy conservation says $E_1 + E_2 = Mc^2$ . And each daughter sits on its own mass shell. That is enough. Two equations and two mass shells determine both energies and the common momentum, and the only thing left undetermined is which way the axis points. The consequence is worth stating starkly. A pion at rest cannot choose how energetic its muon will be. Every $\pi^+$ in the universe that decays to $\mu^+\nu_\mu$ produces a muon of kinetic energy $4.12\ \mathrm{MeV}$ and a neutrino of $29.79\ \mathrm{MeV}$ — not on average, but exactly, every time, because there is no other kinematically allowed outcome. In a nuclear emulsion that shows up as a muon track of constant length, and it was precisely that constancy — about $600\ \mu\mathrm{m}$ , in photograph after photograph — that told Powell's group in 1947 they were watching a two-body decay. Add a third daughter and the rigidity is gone at once. Now the two "other" particles can share their energy in any proportion, which means the first one's energy can be anything from zero up to a maximum. A line becomes a spectrum. Nature ran that experiment for us in $\beta$ decay, and the answer was a spectrum.

Figure

PUNCHLINE — the missing line IS the neutrino. If a neutron decayed to just a proton and an electron, two-body kinematics would force the electron to a single sharp energy: 781.6 keV, every event, no spread. Chadwick (1914) measured a smooth continuum instead, running from zero up to an endpoint that sits EXACTLY where the two-body line would have been. A continuum means three bodies. The third body is the antineutrino, and Pauli inferred it in 1930 from this graph alone

The core equations

A decay is four-momentum conservation with one particle on the left:

p_M^\mu = \sum_i p_i^\mu, \qquad\text{and in the parent's rest frame}\qquad p_M^\mu = (Mc,\ \vec 0\,), \quad \sqrt{s} = Mc^2 \qquad (1)

where: $M$ — the parent's invariant mass. Physical reading: the parent's rest frame is the products' CM frame. That is not an assumption but a consequence — the products inherit the parent's four-momentum, so their total momentum vanishes wherever the parent's does. Two daughters, and the system closes exactly. Back-to-back with common magnitude $p^{*}$ , energies summing to $Mc^2$ , each on its mass shell:

\boxed{\;E_1 = \frac{\big(M^2 + m_1^2 - m_2^2\big)c^2}{2M}, \qquad E_2 = \frac{\big(M^2 + m_2^2 - m_1^2\big)c^2}{2M}\;} \qquad (2)

where: $E_1 + E_2 = Mc^2$ identically. Physical reading: these are numbers, not distributions. Each daughter gets half of $Mc^2$ , plus or minus a term measuring which of them is heavier — and the heavier one gets more energy but, as we will see, less kinetic energy. The common momentum follows from either mass shell, and it is the marquee:

\boxed{\;p^{*} = \frac{\lambda^{1/2}\big(M^2c^4,\ m_1^2c^4,\ m_2^2c^4\big)}{2Mc^3} = \frac{c}{2M}\sqrt{\big[M^2 - (m_1+m_2)^2\big]\big[M^2 - (m_1-m_2)^2\big]}\;} \qquad (3)

where: $\lambda(x,y,z) = x^2+y^2+z^2-2xy-2yz-2zx$ is the Källén or triangle function. Physical reading: check the units — $\lambda^{1/2}$ has energy², and dividing by $Mc^3$ (mass × velocity³) gives energy²/(mass·velocity³) $=$ momentum. The second form is the same thing with the masses in mass units, and it is the one to memorise: $p^{*}$ vanishes exactly at $M = m_1 + m_2$ — the threshold below which the decay is forbidden — and grows from there. Centre-of-Momentum Frame derives this for a general system with $\sqrt{s}$ in place of $Mc^2$ ; here $\sqrt{s} = Mc^2$ and the result becomes a statement about a single particle. Two more quantities every experimentalist uses. The released energy, and the invariant-mass reconstruction that turns it into a discovery:

Q = \Big(M - \sum_i m_i\Big)c^2 = \sum_i T_i \qquad\text{and}\qquad \boxed{\;M^2c^4 = \Big(\sum_i E_i\Big)^2 - \Big|\sum_i \vec p_i\,\Big|^2c^2\;} \qquad (4)

where: the $E_i$ , $\vec p_i$ on the right are measured in the lab, at whatever speed the parent happened to have. Physical reading: this is the whole of experimental particle physics in one line. Because $s$ is an invariant, the right-hand side returns $M^2c^4$ in every event, however fast the parent was moving and in whatever direction. Collect a million events, histogram that number, and a real particle shows up as a spike; random combinations of unrelated tracks do not.

Deriving the decay

Derivation ATwo-body decay from rest: the daughters are back-to-back and monoenergetic

The claim is that a decaying particle has no say in how its energy is divided. Given the three masses, the two daughters' energies and their common momentum are fully determined, and the only freedom left is which direction the decay axis points. We derive this from four-momentum conservation and the two mass shells, and the algebra is short because the constraint count is exactly right.

Step 1. Go to the parent's rest frame and impose momentum conservation. The parent has $\vec p_M = \vec 0$ , so $\vec p_1 + \vec p_2 = \vec 0$ , hence $\vec p_2 = -\vec p_1$ and $|\vec p_1| = |\vec p_2| \equiv p^{*}$ . Why: back-to-back with equal magnitudes, before any dynamics is mentioned. A two-body decay is collinear, always.

Step 2. Impose energy conservation: $E_1 + E_2 = Mc^2$ . Why: the parent is at rest, so its total energy is its rest energy, and there is nothing else for the daughters to share.

Step 3. Subtract the mass shells. $E_1^2 - E_2^2 = \big(p^{*2}c^2 + m_1^2c^4\big) - \big(p^{*2}c^2 + m_2^2c^4\big) = (m_1^2 - m_2^2)c^4$ , because $p^{*}$ is common to both and cancels. Why: this is the standard move — subtracting the mass shells kills the momentum entirely and leaves a pure statement about energies. It is what makes the system linear.

Step 4. Solve. Factor $E_1^2 - E_2^2 = (E_1+E_2)(E_1-E_2) = Mc^2(E_1 - E_2)$ , so $E_1 - E_2 = (m_1^2-m_2^2)c^2/M$ . Adding this to $E_1 + E_2 = Mc^2$ and halving,

\boxed{E_1 = \frac{\big(M^2 + m_1^2 - m_2^2\big)c^2}{2M}}

and $E_2$ by swapping the labels. Why: fully determined — a function of three masses and nothing else. No matrix element, no coupling constant, no lifetime enters. Kinematics has already answered the question before the dynamics is even asked.

Step 5. Extract $p^{*}$ . Insert $E_1$ into $p^{*2}c^2 = E_1^2 - m_1^2c^4$ and factor the resulting difference of squares twice; the Källén function appears and factorises:

\boxed{p^{*} = \frac{c}{2M}\sqrt{\big[M^2 - (m_1+m_2)^2\big]\big[M^2 - (m_1-m_2)^2\big]}}

Why: the factorised form is the useful one. It shows at a glance that $p^{*} \to 0$ as $M \to m_1 + m_2$ — the daughters are created at rest relative to each other, with $Q = 0$ — and that $p^{*}$ is imaginary for $M < m_1+m_2$ , which is the algebra's way of saying the decay is forbidden.

Step 6. Read off the rigidity. $E_1$ , $E_2$ and $p^{*}$ are all functions of $(M, m_1, m_2)$ alone. The decay has zero kinematic freedom beyond the orientation of its axis. Why: this is the owned result, and it is a much stronger statement than it looks. It says that if you find the daughters of a two-body decay, their energies are a fingerprint — and if you find them with a spread of energies, it was not a two-body decay. *Verified in verify_aq-sr-particle-decay-kinematics.py: $E_1 + E_2 = Mc^2$ and $E_1^2 - (p^{*}c)^2 = m_1^2c^4$ are proved symbolically; the Källén factorisation is proved by expansion; $p^{*}$ is asserted dimensionally to be a momentum; $p^{*} \to 0$ at $M = m_1+m_2$ ; and the $\pi^+ \to \mu^+\nu_\mu$ numbers printed in this article are reproduced against PDG masses.*

Go deeper · Derivation B · 2 minWhy an invariant-mass histogram shows a peak, and why unrelated tracks do not

Derivation BWhy an invariant-mass histogram shows a peak, and why unrelated tracks do not

Experimentally, one almost never sees a decay at rest. Parents are produced with whatever momentum the collision gave them, they fly, and they decay in mid-flight — so the daughters' lab energies vary wildly from event to event and none of the numbers above can be read off directly. And yet the parent's mass can be, exactly, event by event. The reason is that the quantity to compute is not an energy but an invariant.

Step 1. Write the conservation law covariantly: $p_M^\mu = p_1^\mu + p_2^\mu$ , valid in every frame because it is an equality of four-vectors. Why: no rest frame is invoked. This is the version an experimenter can actually use, because a detector measures lab four-momenta.

Step 2. Square it. $p_M\cdot p_M = (p_1 + p_2)\cdot(p_1+p_2)$ , and the left side is $M^2c^2$ by the parent's mass shell. In energy² units:

\boxed{M^2c^4 = \big(E_1 + E_2\big)^2 - \big|\vec p_1 + \vec p_2\big|^2c^2}

Why: the right side is built entirely from lab-measured quantities, and it is a Lorentz scalar — so it returns the same number no matter how fast the parent was going. That single fact is what makes particle discovery possible.

Step 3. Expand the cross term to see what is being measured. Using the two mass shells, $M^2c^4 = m_1^2c^4 + m_2^2c^4 + 2\big(E_1E_2 - \vec p_1\cdot\vec p_2\,c^2\big)$ . Why: the parent's mass is encoded in the relative angle and energies of the daughters — which is exactly what a tracking detector is built to measure.

Step 4. Now consider two tracks that did not come from a common parent. Equation (4) still produces a number — you can always add four-momenta — but that number is not any particle's mass. It depends on the accidental angle and energies of two unrelated particles, so it varies smoothly from event to event and its histogram is a smooth combinatorial background. Why: this is the crux. Signal events all return the same value; background events return a spread. A peak on a smooth background is therefore a statement that some pairs share a parent.

Step 5. Ask why the peak has any width at all. Not from kinematics — that is exact. The width has two sources: detector resolution, and the parent's natural width $\Gamma = \hbar/\tau$ , which for a short-lived resonance means the parent does not have a single sharp mass to begin with. Why: so the sharpness of a bump measures a lifetime. A narrow peak (the $J/\psi$ ) means a long-lived state; a broad one means a fleeting one. The kinematic rigidity derived here is what makes that reading possible — every wobble in the peak is physics, because none of it is kinematics. *Verified in verify_aq-sr-particle-decay-kinematics.py: a $\pi^+$ is decayed at a rest-frame angle of $37^\circ$ and then boosted to $\beta = 0, 0.5, 0.9, 0.999$ ; the reconstruction $\sqrt{(\sum E)^2 - |\sum\vec p\,|^2c^2}$ returns $139.57039\ \mathrm{MeV}$ — the pion mass — to machine precision at every boost.*

Go deeper · Derivation C · 2 minTwo bodies give a line, three give a continuum, and that is how the neutrino was found

Derivation CTwo bodies give a line, three give a continuum, and that is how the neutrino was found

Pauli's argument was pure kinematics, and it is the most consequential use this chapter's machinery has ever been put to. The claim is that the shape of the electron spectrum in $\beta$ decay — not its scale, not its normalisation, just the fact that it is a continuum — is by itself a proof that more than two particles emerge. We prove the two halves of it, and then compute where the continuum has to end.

Step 1. Count the freedom for two bodies. Six momentum components; four conservation equations; two mass shells. Six constraints, six unknowns — determined, up to the three rotations that orient the axis. The daughter energies are numbers. A line. Why: this is Derivation 1, restated as a count. It is worth having in this form, because the count is what generalises.

Step 2. Count it for three. Nine momentum components; still four conservation equations; three mass shells. That is seven constraints against nine unknowns, leaving two genuine kinematic degrees of freedom beyond orientation. The energy of any one daughter is therefore no longer fixed — it can vary continuously. A spectrum. Why: adding a particle adds three unknowns and only one constraint. The rigidity of the two-body case was a coincidence of counting, and it does not survive.

Step 3. Find the endpoint. Daughter 1's energy is largest when the other two carry as little energy as possible — that is, when they are at rest relative to each other, so that their pair's invariant mass takes its minimum value $m_{23} = m_2 + m_3$ . But then the decay is effectively a two-body decay, $M \to m_1 + (m_2+m_3)$ , and Derivation 1 applies:

\boxed{E_1^{\max} = \frac{\big(M^2 + m_1^2 - (m_2+m_3)^2\big)c^2}{2M}}

Why: <strong>the endpoint of the three-body continuum is exactly the line the two-body decay would have produced.</strong> That is a striking and checkable statement, and it is the reason the missing line and the observed endpoint sit at the same energy on the punchline figure.

Step 4. Put the neutron in. For $n \to p + e^- + \bar\nu_e$ with $m_{\bar\nu} \approx 0$ , the endpoint condition is $m_{23} = m_p$ , so $E_e^{\max} = (m_n^2 + m_e^2 - m_p^2)c^2/(2m_n)$ . With PDG masses this is $1.2926\ \mathrm{MeV}$ , i.e. a kinetic energy of $\mathbf{781.6\ keV}$ . Why: and the $Q$ -value is $(m_n - m_p - m_e)c^2 = 782.33\ \mathrm{keV}$ . The two differ by $0.75\ \mathrm{keV}$ — the proton's recoil kinetic energy, which the naive $Q$ ignores. Both numbers are correct; they answer different questions, and confusing them is a real and common error.

Step 5. Draw the conclusion Pauli drew. The observed spectrum is continuous and ends at the two-body value. Continuity forces $\geq 3$ bodies. Only two are seen. Therefore an unseen third particle exists, it is neutral (charge is already balanced), and it is light (the endpoint sits where a massless third body would put it). Why: every one of those inferences is kinematics, and none of them needs a theory of the weak interaction — which is why Pauli could make them in 1930, four years before Fermi wrote one down. *Verified in verify_aq-sr-particle-decay-kinematics.py: the two-body $n \to p + e$ computation gives $T_e = 781.58$ keV against a $Q$ -value of $782.33$ keV, with the $0.75$ keV difference confirmed to equal the proton recoil; and the three-body endpoint formula is checked to coincide with the two-body line when $m_{23} \to m_2 + m_3$ .*

Limiting cases

Case	Result	Meaning
$M = m_1 + m_2$	$p^{*} = 0$ , $Q = 0$	threshold: the daughters are born at rest relative to each other
$M < m_1 + m_2$	$p^{*}$ imaginary	the decay is forbidden — the algebra says so by going complex
$m_2 = 0$ (one massless daughter)	$p^{} = \dfrac{(M^2 - m_1^2)c}{2M}$ , $\;E_2 = p^{}c$	the pion's neutrino: $E_\nu = p^{*}c = 29.79$ MeV
$m_1 = m_2$	$E_1 = E_2 = \tfrac12 Mc^2$	e.g. $\pi^0 \to \gamma\gamma$ : each photon takes $67.49$ MeV
$m_1 \gg m_2$	the heavy daughter takes almost all the energy, almost none of the kinetic energy	$\mu^+$ : $E = 109.78$ MeV but $T = 4.12$ MeV
three bodies	$E_1$ ranges over a continuum	endpoint at $E_1^{\max} = \big(M^2+m_1^2-(m_2+m_3)^2\big)c^2/2M$
parent in flight	daughters' lab energies vary	but $\sqrt{(\sum E)^2 - \|\sum\vec p\,\|^2c^2}$ still returns $M c^2$ exactly

Common trapQ is not the daughter's kinetic energy, and the difference is measurable

$Q = (M - \sum_i m_i)c^2$ is the total kinetic energy released, shared among all the daughters — never the kinetic energy of any one of them. For $\pi^+ \to \mu^+\nu_\mu$ , $Q = 33.91\ \mathrm{MeV}$ , and it splits as $T_\mu = 4.12\ \mathrm{MeV}$ plus $E_\nu = 29.79\ \mathrm{MeV}$ . The massless neutrino takes 88 % of it, even though the two carry equal and opposite momenta — because for a massless particle all the energy is kinetic, while the muon's $109.78\ \mathrm{MeV}$ is nearly all rest energy. The same trap sits in nuclear $\beta$ decay, where the electron's maximum kinetic energy ( $781.6\ \mathrm{keV}$ for the neutron) is not the $Q$ -value ( $782.33\ \mathrm{keV}$ ). The $0.75\ \mathrm{keV}$ shortfall is the proton's recoil, which $Q$ ignores because $Q$ is defined by masses alone. That is a real difference, and a modern spectrometer resolves it comfortably.

Worked examples

1 · The pion's monoenergetic muon

A $\pi^+$ at rest decays to $\mu^+ + \nu_\mu$ . Find the muon's momentum, energy and kinetic energy, and the neutrino's energy. Take $m_\pi c^2 = 139.57039\ \mathrm{MeV}$ , $m_\mu c^2 = 105.6583755\ \mathrm{MeV}$ , $m_\nu \approx 0$ .

Step 1 — identify. Two-body decay from rest, one daughter massless. Everything is fixed by the two masses; there is nothing to average over.

Step 2 — the momentum. With $m_2 = 0$ the Källén function collapses and $p^{*} = (M^2 - m_\mu^2)c/(2M)$ . Numerically, $(139.57039^2 - 105.6583755^2)/(2\times 139.57039) = (19479.9 - 11163.7)/279.141 = \mathbf{29.79\ MeV/c}$ .

Step 3 — the energies. The neutrino is massless, so $E_\nu = p^{*}c = \mathbf{29.79\ MeV}$ . The muon: $E_\mu = \sqrt{(p^{*}c)^2 + m_\mu^2c^4} = \sqrt{887.6 + 11163.7} = \mathbf{109.78\ MeV}$ . Check: $29.79 + 109.78 = 139.57 = m_\pi c^2$ ✓.

Step 4 — the kinetic energy. $T_\mu = E_\mu - m_\mu c^2 = 109.78 - 105.66 = \mathbf{4.12\ MeV}$ , which matches the PDG value.

Step 5 — the boundary number. $Q = (m_\pi - m_\mu)c^2 = \mathbf{33.91\ MeV}$ , and it splits as $4.12 + 29.79 = 33.91$ ✓. The massless neutrino takes 88 % of the released energy while carrying exactly the same momentum as the muon — the whole asymmetry is the muon's rest mass soaking up energy without soaking up momentum.

Step 6 — what it looks like. A $4.12\ \mathrm{MeV}$ muon has a range of about $600\ \mu\mathrm{m}$ in nuclear emulsion. Every time. That is the observation Lattes, Occhialini and Powell made in 1947, and it is why they could conclude the decay was two-body without ever seeing the neutrino.

2 · Reconstructing the parent from a boosted decay

The same pion now decays in flight at $\beta = 0.9$ . The muon comes off at $37^\circ$ to the flight direction in the pion's rest frame. Show that adding the daughters' lab four-momenta still returns the pion's mass.

Step 1 — identify. The individual lab energies will be nothing like $109.78$ and $29.79\ \mathrm{MeV}$ — the boost mixes energy and momentum. But equation (4) is an invariant, so it must return $m_\pi c^2$ .

Step 2 — boost. With $\gamma = 1/\sqrt{1-0.81} = 2.294$ , transform both daughters: $E_{\mathrm{lab}} = \gamma(E^{*} + \beta c\,p^{*}_{\parallel})$ and $p_{\parallel}^{\mathrm{lab}} = \gamma(p^{*}_{\parallel} + \beta E^{*}/c)$ , with $p_\perp$ unchanged.

Step 3 — reconstruct. Form $\big(E_\mu + E_\nu\big)^2 - \big|\vec p_\mu + \vec p_\nu\big|^2c^2$ from the lab values. The result is $\mathbf{139.57039\ MeV}$ — the pion's mass, to every digit.

Step 4 — the boundary number. Repeat at $\beta = 0$ , $0.5$ , $0.999$ : the answer is $139.57039\ \mathrm{MeV}$ every time. The reconstruction is exactly boost-independent, which is precisely why a histogram of it, filled with a million events at a million different parent momenta, still shows a spike one particle wide.

Figure

LEFT — a two-body decay at rest has zero freedom: back-to-back, p* = 29.79 MeV/c fixed by the masses alone. The muon takes 109.78 MeV of energy but only 4.12 MeV of kinetic energy; the massless neutrino takes 29.79 MeV, which is 88 percent of the released Q = 33.91 MeV. RIGHT — because that rigidity holds in the parent's rest frame while the reconstruction (ΣE)² − |Σp c|² is a Lorentz INVARIANT, adding the daughters' LAB four-momenta returns 139.57 MeV at every parent speed. Signal piles into a spike; unrelated tracks smear into a smooth background

Physics of the portrait

The pion is the particle that taught physics how to read a decay. It was predicted for one reason, found by mistake, confused with something else for a decade, and finally identified by the one thing this article says cannot vary: the range of its daughter.

Origins · 2 minPion and how this was discovered

Pion

1935–1947 · Yukawa predicts a carrier · a decade of mistaken identity · Powell's emulsions · the muon that always runs 600 μm

In 1935 Hideki Yukawa argued that the short range of the nuclear force implied a mediating particle of mass around $100\ \mathrm{MeV}/c^2$ — heavy enough to make the force die off within a nucleus, light enough to be findable. When a particle of roughly that mass turned up in cosmic rays in 1936–37 (Neddermeyer and Anderson), it was taken to be Yukawa's, and for ten years it was called a "mesotron". It was not: it barely interacted with nuclei at all, which is precisely what Yukawa's particle could not do. The knot was cut in 1947 in Bristol, where Cecil Powell's group — with César Lattes and Giuseppe Occhialini — exposed nuclear emulsions at high altitude and found tracks in which one meson stopped and gave birth to a second. The primary was Yukawa's particle, the pion; the secondary was the impostor from 1937, the muon. What proved the case was not the existence of the second track but its length. In photograph after photograph the secondary meson ran the same distance before stopping — about $600\ \mu\mathrm{m}$ — and a constant range means a constant energy, which for a decay at rest means exactly two bodies in the final state: a muon and one unseen neutral particle. The kinematics of this article, read directly off a photographic plate. Powell took the 1950 Nobel Prize for the method and its result; the neutral particle whose existence the constant range implied was the neutrino, and it would not be seen directly for another nine years.

History

1914

— James Chadwick shows with a magnetic spectrometer that the beta spectrum is a continuum, not a line, contradicting what a two-body decay demands.

1927

— Ellis and Wooster's calorimeter measures the mean energy per beta decay of radium E and finds it far below the spectrum endpoint: the energy really is missing.

1930

— On 4 December, Pauli writes to the Tubingen conference proposing an unseen neutral particle as "a desperate remedy" for the continuous spectrum.

1934

— Fermi's theory of beta decay puts Pauli's particle into a quantitative framework and gives it the name neutrino.

1947

— Lattes, Occhialini and Powell find pion tracks in nuclear emulsion; the constant range of the secondary muon, about 600 micrometres, forces a two-body decay to a muon plus one neutral particle.

1956

— Cowan and Reines detect the neutrino directly at the Savannah River reactor, twenty-six years after Pauli inferred it from a spectrum's shape.

Going further

Texts. Griffiths, Introduction to Elementary Particles, 2nd ed., §3.4 · Rindler, Introduction to Special Relativity, 2nd ed., §31 · Perkins, Introduction to High Energy Physics, 4th ed., ch. 2 · Particle Data Group, Review of Particle Physics, "Kinematics" · Lattes, Occhialini & Powell, Nature 160, 453 (1947) · Pauli's letter of 4 December 1930, reproduced in Brown, Physics Today 31(9), 23 (1978).

Provenance. Signature $(+,-,-,-)$ ; $p\cdot p = +m^2c^2$ ; $s$ in energy² units, and in a decay $\sqrt{s} = Mc^2$ . verify_aq-sr-particle-decay-kinematics.py parses this article and re-derives every number it prints: $p^{*} = 29.79$ MeV/ $c$ , $E_\mu = 109.78$ MeV, $T_\mu = 4.12$ MeV, $E_\nu = 29.79$ MeV, $Q = 33.91$ MeV, the $88\%$ share, $\pi^0 \to \gamma\gamma$ at $67.49$ MeV each, the two-body $n \to p + e$ line at $781.6$ keV, the $Q$ -value $782.33$ keV, the $0.75$ keV proton recoil, and the boost-independent reconstruction $139.57039$ MeV at $\beta = 0,\ 0.5,\ 0.9,\ 0.999$ .

Scope. Rates — lifetimes, branching ratios, matrix elements — are dynamics and are not derived here; this article predicts where the products go, never how often the decay happens. Phase space beyond the two-body case is sketched, not developed. The $\pi^0 \to \gamma\gamma$ minimum-opening-angle result is stated but its derivation belongs with relativistic aberration.

Builds on. Centre-of-Momentum Frame · Conservation of Four-Momentum · Invariant Mass · Relativistic Four-Momentum · Massless Particles. Leads to. Threshold Energy · Relativistic Inelastic Collisions · Mandelstam Invariants · Binding Energy and Mass Defect.

Why is the daughter momentum fixed? Surely the decay could go any way it liked.

It can — but only in direction. Count the constraints: two daughters have six momentum components, four-momentum conservation supplies four equations, and the two mass shells supply two more. Six equations, six unknowns: the system is exactly determined, leaving only the three rotations that orient the decay axis. So the magnitudes — both energies and the common momentum — are functions of the three masses and nothing else. No coupling constant, no interaction, no lifetime can shift them. The dynamics decides how often the decay happens and which way it points; it has no say at all in how the energy is divided.

Does the monoenergetic result survive if the parent is moving?

Not for the individual energies, and this trips people up. A parent in flight produces daughters whose lab energies spread over $\gamma(E^{*} \pm \beta c\,p^{*})$ , depending on the rest-frame emission angle. So a spread of measured energies does not disprove a two-body decay — it may just mean moving parents. What survives every boost is the invariant reconstruction $M^2c^4 = (\sum E_i)^2 - |\sum\vec p_i|^2c^2$ , which returns the parent's mass exactly, at any speed. That is why Powell's evidence was decisive: his pions had stopped in the emulsion, so the rest-frame result applied directly, and the constant $600\ \mu\mathrm{m}$ range meant exactly what it looked like.

What is the Källén function actually doing?

It is measuring how far the decay is above threshold, in a symmetric way. $\lambda(x,y,z) = x^2+y^2+z^2-2xy-2yz-2zx$ factorises as $\lambda(M^2, m_1^2, m_2^2) = \big[M^2-(m_1+m_2)^2\big]\big[M^2-(m_1-m_2)^2\big]$ , and the first bracket is the one that matters: it vanishes exactly when $M = m_1+m_2$ , which is the threshold, and goes negative below it, making $p^{*}$ imaginary and the decay forbidden. The name "triangle function" comes from the fact that $\lambda < 0$ precisely when $\sqrt{x},\sqrt{y},\sqrt{z}$ can form a triangle. It is the same function that governs the two-body CM momentum of any system, with $\sqrt{s}$ in place of $Mc^2$ — Centre-of-Momentum Frame derives it there.

Why does the massless neutrino take most of the energy?

Because momentum, not energy, is what must balance — and a massless particle converts momentum into energy at the best possible exchange rate. Both daughters carry $p^{*} = 29.79\ \mathrm{MeV}/c$ . For the neutrino, $E = pc$ , so its entire $29.79\ \mathrm{MeV}$ is kinetic. For the muon, $E = \sqrt{(pc)^2 + (mc^2)^2} = 109.78\ \mathrm{MeV}$ , of which $105.66\ \mathrm{MeV}$ is just its rest energy — it is carrying a lot of energy but it did not gain much. Of the $33.91\ \mathrm{MeV}$ actually released, the neutrino takes $29.79$ and the muon $4.12$ . The heavier the daughter, the more energy it has and the less of the $Q$ it gets.

If the reconstruction is exact, why do invariant-mass peaks have width?

Two reasons, and neither is kinematics. The first is detector resolution: energies and angles are measured imperfectly, so the reconstructed number scatters around the true mass. The second is real physics — the parent's natural width $\Gamma = \hbar/\tau$ . A state that lives for a time $\tau$ does not have a single sharp mass; by the energy–time uncertainty relation its mass is smeared by $\Gamma$ . This is why a narrow peak is interesting: the $J/\psi$ 's width of $93\ \mathrm{keV}$ against a mass of $3097\ \mathrm{MeV}$ announced a state living anomalously long, which is what made 1974 a revolution. Because the kinematics contributes zero width, every bit of the observed width is either instrument or physics — and that is what makes the measurement worth making.

Could the continuous beta spectrum have meant energy is not conserved?

It could have, and Bohr seriously proposed exactly that — he was prepared to give up energy conservation in individual nuclear events rather than invent a new particle. Pauli's alternative was regarded at the time as the wilder of the two, which is why he called it "a desperate remedy" and declined to publish it for four years. What settled it was that the neutrino hypothesis made further predictions and they held: the spectrum's endpoint sits exactly where a massless third body puts it; the recoil of the daughter nucleus is measurable and matches a three-body decay; and in 1956 Cowan and Reines detected the particle directly. Energy conservation survived, and the episode is the standard example of a conservation law being defended by postulating something new rather than abandoned.