Thermal Light
The blackbody laboratory reservoir shows how sung modons produce spectral free light when locked in a boundary that absorbs all frequencies, and emits at the Planck scale, the same curve as the CMB.
Interactive: open the thermal-light simulation — one mode between two walls at T, its boundaries primed and relaxed in the Boltzmann ratio, every capture, triggered release and breath drawn as it happens, the tally of who tipped the boundary against e^{-x}, \langle n\rangle against 1/(e^x-1) and g^{(2)}(0)\to2 (swap the wall for a pump and the same mode goes Poisson); the carbon speck — a cavity whose walls slide from mirror to black, a laser line fired in through the hole and re-priced or not, the spectrum stacked by who minted each coin, the leaky cavity’s \mu-distortion, and a crossings counter that never counts an exchange; a wall or a stretch — the Planck gas at last scattering stretched to a=1100 with no count touched, sliding onto Planck at T_0/a across the 3 THz floor where compact modons become windings, and a wall in between printing the distortion FIRAS does not see.
Introduction
Spectrum-Free Light shows three ways to make light — sung from an orbital ladder, shed from a driven boundary, and the reservoir: the ordinary blackbody, line-free like the shed family but not mechanically made. They have the distinctive and modeled temperature and cooling curve with the Planck spectrum from thermal radiation.
Take a box, heat it, and let the light inside come to equilibrium: the natural image is a gas — modons trapped in a pool, jostling, mixing. Since they cannot merge mid-flight, their interaction with atoms mediates their energy.
In the ocean, when two counter-rotating dipoles meet head-on and they exchange partners — each core pairs off with the other dipole’s opposite core, and two new dipoles leave at right angles to the way they came. They don’t merge - they bounce off. And unlike in the ocean, there is no friction with these vortices.
Light is linear to a precision that has no rival in physics: two laser beams cross without leaving a trace on each other, and quantum electrodynamics puts the photon–photon cross-section at fourth order in \alpha, through a loop of virtual pairs — about 10^{-70}\,\text{m}^2 at optical energies, some fifty orders of magnitude below the geometric cross-section of an atom. The Euler–Heisenberg action from Vacuum Birefringence shows the rare modon-to-modon coupling. From thespin statistics section, the photon is a boson, even-parity, its two cores cancelling, and it passes through the substrate transparently — no net polarization.
Planck pointed out that a cavity with perfectly reflecting walls, filled with radiation and nothing else, never reaches equilibrium — whatever spectrum you put in stays in, forever. To make the radiation thermalize he had to add a Kohlenstäubchen, a speck of carbon dust: a piece of matter that absorbs and re-emits at every frequency, and thereby shuffles energy between modes that cannot shuffle it themselves. The speck is not a technicality. It is the whole mechanism.
The modons in a blackbody cavity are trapped - taking turns - emitting, being absorbed, triggering another one. Every modon energy comes from a boundary, and the cavity has the geometry that makes each modon meet the wall many times before it finds the hole. The same Kirton–Keeling race in the photon-BEC chapter where thermalization and loss are in balance. A blackbody cavity is Planck’s carbon speck smeared over every surface, with a hole small enough that no modon escapes before being absorbed by the speck.
The Heat Wall
The wall that can absorb all modon frequencies eventually will blend them to the spectral free light. As in the laser chapter’s one coupling the receiving atom can borrow-and-return, capture, or trigger and release. The photon-BEC chapter showed that an uninverted boundary bath that runs a capture and re-shed closed cycle is a thermometer: each contact adjusts the modon’s frequency from the boundary’s state and deck of thermal potentials. The blackbody repeats this coupling to a detailed balance. The arithmetic is Einstein’s of 1917, which reads in substrate language without a single translation.
Take one mode of the cavity at frequency \nu, holding n modons, and the wall’s boundaries that resonate with it, \Delta E = h\nu. The wall is at temperature T because it sits in the universal bath, so its boundaries are primed (holding a coin of h\nu) and relaxed in the Boltzmann ratio N_2/N_1 = e^{-h\nu/k_BT}. Then:
- A relaxed boundary captures at a rate proportional to n — it can only be tipped upward by a modon that is there.
- A primed boundary releases at a rate proportional to n + 1: the n is triggered release into the wake of a modon already present, and the +1 is the lattice’s own breath — spontaneous emission is stimulated by the medium itself, one quantum of tickle per mode, A/B = 8\pi h\nu^3/c^3.
Balance, N_1\,n = N_2\,(n+1), gives
\boxed{\;n(\nu,T) \;=\; \frac{1}{e^{h\nu/k_BT}-1}\;}
and multiplying by the number of modes per volume, 8\pi\nu^2/c^3, and the coin’s denomination h\nu gives Planck’s law outright. Three ingredients went in, and only three:
- The coin is indivisible and its denomination is h\nu. A boundary hands over one whole modon or none — “a propagating breath cannot carry less than the standing one it is made of.” This is Planck’s quantum hypothesis, and is the modon’s existence condition.
- Capture is the time-reverse of release, B_{12}=B_{21} — one boundary geometry run in two directions, which is also Kirchhoff’s law of 1860: a boundary that captures a color well must release it equally well. Emissivity equals absorptivity because they are one matching condition read forwards and backwards.
- One breath per mode. The +1. Without it, balance would give n = \infty at every frequency; with it, every mode has a floor of exactly one quantum’s worth of tipping that does not depend on how many modons are present.
Notice what did not go in: the boundary’s ladder. Nothing in the three ingredients asks what \Delta E levels the wall has, only that at each \nu some boundary resonates. The ladder decides whether a color can be minted; the temperature decides how many coins. This makes the precise distinction between photons produced in a reservoir versus one sung by a precise orbital boundary. A sung source has a ladder with gaps and an open loop — the modon leaves and is never re-priced — so the ladder’s spacing is the spectrum. A black boundary is a ladder with no gaps: condensed matter, where atoms share boundaries in a rigid lattice of merged raceways, the levels smeared into bands and each level broadened further by fast contact with the bath — so every \nu finds a capture channel, and detailed balance prices every \nu by the same rule. The lines are not absent because there is no ladder; they are absent because the ladder is continuous and the loop is closed.
Kirchhoff’s three laws of spectroscopy are then the one coupling read three ways:
| Kirchhoff (1860) | Boundary | Loop | Spectrum |
|---|---|---|---|
| a hot, dense body | ladder with no gaps | closed — every modon re-priced | the full Planck curve |
| a hot, thin gas | ladder with gaps | open — sung once, never recaptured | emission lines |
| a continuum through a cool, thin gas | ladder with gaps, relaxed | capture at the rungs, re-shed elsewhere | absorption lines on the continuum |
A star’s spectrum shows the three rows stacked: the Planck envelope from the dense photosphere, the Fraunhofer lines from the thin cooler gas above it.
The star’s own first row gapless nature is shown in the Solar and Stellar Dynamics chapter - the H minus open boundary layer in the photosphere. The fire chapter sorts a candle the same way — the blue cone is row two, the yellow body is row one, thermal radiation from condensed-matter boundaries.
Who Tips the Boundary
Every primed boundary in the wall holds one coin, h\nu, and releases it only when something tips it. There are exactly two candidates. One is light already in the mode: each modon present adds one nudge, so a mode holding n modons nudges n times as hard. The other is the substrate itself, which supplies the “+1”: one nudge per mode whether or not any light is there. Every release is therefore either triggered or spontaneous, in the ratio n:1. The occupation n and the breath’s 1 compete for the job of tipping each primed boundary, and Planck’s curve is a map of which of them wins.
Reading the figure. The upper-left panel of Figure 1 draws that competition. Its horizontal axis is frequency in units of the wall’s temperature, h\nu/k_BT: how many thermal coins k_BT one light coin costs. Red light is to the left, blue to the right, and one drawing serves every temperature. The top plot is the Planck curve itself, the amount of light the wall emits at each frequency. The strip beneath it, on the same axis, is the tally of who tipped the boundary. At each frequency the teal band is the share of releases triggered by light already in the mode, n/(n+1), and the gold band is the share released spontaneously, 1/(n+1). Since n = 1/(e^{h\nu/k_BT}-1), the triggered share simplifies to e^{-h\nu/k_BT}, which is the teal edge of the strip. The dashed vertical line through both plots is the knee, where the two shares are equal.
- Below the knee, n>1: the mode already holds more than one modon, so most releases are triggered — light tips the boundary, and the wall’s coins are cloned into the wake of coins already there. Here the mode behaves classically, holding k_BT of energy regardless of \nu: this is the Rayleigh–Jeans regime, where equipartition was right.
- Above the knee, n<1: the mode is usually empty. Whatever is released is released spontaneously — the lattice tips the boundary, one quantum of tickle per mode, and the only question is how often a boundary at T can afford a coin of denomination h\nu. That is e^{-h\nu/k_BT}: the Wien tail. Blue light needs a hot wall not because blue modons are hard to make but because a boundary at T is rarely primed at h\nu \gg k_BT.
- The knee is n=1, where the light’s tickle and the lattice’s tickle are equal: h\nu = k_BT\ln 2. The spectral peak sits a little above it at h\nu = 2.82\,k_BT, where the mode count \nu^2 has finished pulling the maximum to the blue.
The figure shows something the equations hide: the teal wedge under the Planck curve is a sliver. Only about 1% of a blackbody’s energy lies to the red of the knee. The other 99%, including the whole peak, sits where the mode is usually empty. Most thermal light is not light begetting light. It is the substrate’s spontaneous breath, released one coin at a time by a wall that can only occasionally afford it. The two reference curves are the two halves of the history. The dashed Rayleigh–Jeans curve is the classical limit, where every mode is full enough that the +1 is negligible; it is right in the teal and wildly wrong beyond it. The dotted Wien curve keeps only the +1 and drops the triggering; it is right in the gold and too low in the teal. Planck’s curve is the one that counts both.
The ultraviolet catastrophe, in this language, was a bookkeeping error about the coin. Classical physics let a mode take any amount of energy, so equipartition handed every mode k_BT and the mode count \nu^2 did the rest. The substrate’s coin is indivisible: a boundary can mint a blue modon only by handing over a whole h\nu at once, and at h\nu \gg k_BT it almost never holds one. This sharpens the thermal-dynamics chapter’s one-line version, which credited the discrete Bessel spectrum of the modon’s interior. The interior spectrum sets what a modon is; what kills the catastrophe is that the coin comes in one piece.
Some numbers make the gold tail concrete. For a room at 300 K the knee sits at 4.3 THz, deep in the far infrared. Visible light at about 550 THz is h\nu \approx 88\,k_BT up the tail, so the chance that a wall boundary is primed with a visible coin is e^{-88}\approx 10^{-38}. That is why objects at room temperature do not glow. A tungsten filament at 2800 K brings visible light down to h\nu \approx 9\,k_BT and the odds up to about 10^{-4}. The light is still entirely spontaneous, but now it comes often enough to light a room.
| Source | T | knee h\nu = k_BT\ln2 | peak 2.82\,k_BT | Wien \lambda_\text{peak} | share of energy below the 3 THz floor |
|---|---|---|---|---|---|
| CMB | 2.7 K | 0.04 THz | 0.16 THz | 1.1 mm | all of it |
| liquid-nitrogen load | 77 K | 1.1 THz | 4.5 THz | 38\;\mum | 17\% |
| a room | 300 K | 4.3 THz | 18 THz | 9.7\;\mum | 0.5\% |
| tungsten filament | 2800 K | 40 THz | 165 THz | 1.0\;\mum | 10^{-5} |
| the Sun | 5772 K | 83 THz | 339 THz | 0.50\;\mum | 10^{-6} |
Wien’s displacement law is the same picture read across temperatures: the lattice’s breath is the same one quantum at every \nu and every T, while the boundary’s affordability e^{-h\nu/k_BT} slides with T — so the crossover where the light stops tipping and the lattice takes over slides with it, linearly. The shape never changes because the three ingredients never change; only where the knee falls.
Three Bookkeepings
The Light section has now met three ways to fill a mode, and the photon-BEC chapter’s table gains the column it was built to receive:
| Laser | Photon BEC | Blackbody | |
|---|---|---|---|
| population | inverted — pumped | Boltzmann, uninverted | Boltzmann, uninverted |
| state of the light | driven, far from equilibrium | equilibrium, \mu \neq 0 | equilibrium, \mu = 0 |
| who re-prices the modon | nobody — gain clamps to loss | the dye, returning the same coin | the wall, which also mints and melts coins |
| occupied mode chosen by | maximum gain | minimum energy | none — every mode, weighted by T |
| photon number | not conserved | conserved on average | set by T alone |
| statistics g^{(2)}(0) | 1 (Poissonian) | \to 2, flickering | 2 — bunched (Hanbury Brown & Twiss 1956) |
| switched off by | cutting the pump | nothing | cooling the wall |
The middle column and the right one differ by exactly one thing. The dye’s boundary takes and gives only the small change — the modon’s energy is re-priced but the modon itself comes back, so the number is conserved and light acquires a chemical potential. The wall of a furnace does more: its boundaries can take a whole coin into the bath and mint a whole coin out of it. That is the radiation channel and the conduction channel exchanging at the surface — a modon captured and converted to lattice weather, lattice weather converting back to a modon — and it is why blackbody photons have no conserved number and \mu=0: the wall is a reservoir of the coin, not just of its price.
The statistics row is the discriminator the photon-BEC chapter argued for, in its original form. Thermal light bunches — g^{(2)}(0)=2 — because each mode trades with a reservoir that creates and destroys, so its occupation fluctuates by its own mean; the flickering condensate was that same grand-canonical fact caught in a condensed mode. A furnace, a dye cavity, and a laser can all present a bright mode; only their number statistics say who the mode has been trading with, and on what terms.
A Wall or a Stretch
Now the other reservoir the paper leans on. The cosmic microwave background is the most perfect blackbody ever measured — Planck to 50 parts per million — and it has had no wall since last scattering at z\approx1100, nor anything able to re-thermalize it since a few years after the boil. If a blackbody is Planck maintained by contact, how does a photon gas with no contact stay Planck for 13.8 billion years?
Because the one operation the vacuum performs on it is the one operation that leaves the curve alone. The quiet-majority chapter worked out what expansion does to a modon in the framework’s own terms: a photon is a conserved quantum of circulation, its energy hc/\lambda set by the span the winding is stretched over, and as the lattice carrying the two ends of that span separates, the span grows with a and the energy falls as 1/a — “pulled longer by the medium it lives in.” Apply that to every mode at once. Every span is multiplied by the same a, every quantum’s energy divided by it, and the count per mode — the n that Planck’s law is really about — is untouched, because no boundary is there to change it. A Planck curve at T maps onto a Planck curve at T/a, mode by mode. The shape is a fixed point of stretching, and T\propto 1/a is not an extra assumption but the only thing a uniform stretch can do to n(\nu,T).
So a laboratory blackbody and the CMB wear the same curve for opposite reasons. The furnace is Planck maintained: a wall re-prices every modon, and Planck’s is the only spectrum a wall at T leaves alone. The CMB is Planck frozen: no wall touches it, and Planck’s is the only spectrum a stretch leaves alone. Planck’s curve is the unique fixed point of both operations, and that is why it is the universal attractor of light — a modon gas that touches a wall goes to it, and a modon gas that touches nothing stays there.
This also says what FIRAS’s 50 ppm is a measurement of. Any boundary contact after last scattering that did not run all the way to balance would have left a distortion — a chemical potential (\mu-type) or a Compton-heated tail (y-type) — and FIRAS bounds both at the 10^{-5} level. The framework has already cashed that bound twice: the floor crossing is adiabatic to one part in 10^{28}, and black-hole capture removes 10^{-18} of the background per Hubble time. Both are statements that the CMB has met no wall on the way here, and a Planck curve is the receipt.
The Floor Under the Reservoir
The framework’s one native structure in this domain is the modon floor at 3 THz, and this chapter has to say what a thermal source does across it. The answer follows from the three ingredients, and it is more conservative than the paper’s first statement of it.
Below the floor a single quantum is not a compact modon but a stretched winding spread over many cells — yet it is still one winding, still carrying h\nu, still handed over whole or not at all. Ingredient (i) holds. Capture and release remain one geometry run both ways, so (ii) holds. The mode count per volume is 8\pi\nu^2/c^3 on both sides, because the winding has no leading dispersion, so (iii) holds. Nothing in Planck’s law notices the floor. The thermal amplitude is Planck on both sides; what changes at 3 THz is the character of the quantum, not the amount of light.
And this is not a prediction so much as a description of laboratory practice. Every microwave radiometer on Earth is calibrated against a 300 K load and a 77 K load whose emission at gigahertz frequencies — two decades below the floor — is the Rayleigh–Jeans k_BT per mode, and the calibration holds to millikelvin. FIRAS’s own external calibrator was a local blackbody at 2.7 K, emitting at 60–600 GHz — sub-floor from end to end — and the sky was measured as its difference from that calibrator; the 50 ppm is the statement that a locally emitted sub-floor thermal spectrum and a cosmologically stretched one are the same curve. Locally emitted thermal light far below the floor is full-amplitude Planck; it has been, in every cold load, for seventy years.
That sharpens the thermal-dynamics chapter’s prediction of a downward deviation from Planck “at wavelengths approaching or exceeding \xi.” As an amplitude deficit at \lambda>\xi it is already excluded by the cold loads; what survives is the modon-floor chapter’s version — a change of quantization character confined to the 0.1–3 THz band, read in dispersion and in photon statistics, with the Planck amplitude continuous across it. The reservoir’s floor is a change of what kind of thing the wall is handing over, and a body colder than about 53 K — where the Planck peak falls below 3 THz — hands over mostly windings. Cryogenic thermal emission is, in the framework’s strict sense, mostly not made of photons; it is Planck anyway.
Honest Accounting
Four debts, in the house discipline.
First, every equation here is Planck’s, Einstein’s, or Kirchhoff’s. The occupation formula, the A/B ratio, the mode count, the T\propto1/a scaling and the FIRAS bounds are standard physics reproduced without alteration. The chapter’s contribution is identification: that the three ingredients of Planck’s law are three structures the framework already carries — the indivisible coin, the geometric B_{12}=B_{21}, the breath-per-mode — and that assembling them is the missing mechanism behind the reservoir column.
Second, modon non-interaction is asserted, not derived. The classical dipole the modon is modelled on does exchange partners on collision. The framework says the modon does not, and gives the reason as topological — even parity, elastic transit, a winding rather than a lump of fluid — but it has not exhibited the suppression from the Larichev–Reznik solution itself. Until it does, “modons do not merge” is a commitment the framework makes because optics demands it, with the Euler–Heisenberg reproduction in Vacuum Birefringence as the one place the residual coupling has been checked.
Third, why a ladder has no gaps is now answered structurally; how black a given boundary is, is not. This chapter originally said only that a condensed-matter boundary has bands rather than levels and fast bath contact — the standard solid-state picture, invoked rather than derived. The structural half is addressed in Solar and Stellar Dynamics § Three ways to fill a ladder: a ladder is a boundary-matching condition, and there are exactly three things that can happen to the seam to remove its rungs — merge it sideways with its neighbours (condensed matter, bands plus a shared-seam continuum), leave it open at the top so the state above is already continuum (the H^- ion, and with it the entire solar photosphere), or have no seam at all (free carriers, bremsstrahlung). All three hand every \nu a capture channel; detailed balance then does the rest wherever the loop is closed, by a wall or by optical depth. That reading also settles a live claim against the framework’s favour and then against the claimant’s: the dissident argument that a Planck curve proves condensed matter is right about the requirement and wrong that only a lattice meets it.
What is still owed is the quantitative half — the emissivity of a real boundary as a number. The framework has not derived it from boundary structure and cannot yet say from substrate parameters why graphite is black and silver is not. The seam picture names the two axes it should depend on (how completely the outer boundaries have dissolved, and how fast the merged seam exchanges with the bath); computing along them is Solar and Stellar Dynamics § Open Problem 5.
Place in the Framework
The Light section had read the substrate off its excitation as a frequency, a wavevector, an emission mechanism, a phase, and an occupation, and the spectrum-free chapter had left one column of its own table unexplained. This chapter fills it, and finds nothing new was needed: the reservoir is the one coupling run to balance against a boundary with no gaps in its ladder, and Planck’s curve is the record of who tipped each boundary — the light, below the knee; the lattice, above it. The picture of a pool of merging photons gives way to Planck’s carbon speck, which is what every wall is. And the CMB, the paper’s favourite reservoir, turns out to be Planck for the opposite reason from a furnace: nothing has touched it, and a stretch is the one thing that leaves the curve where it was. Wherever the substrate hands over coins to a boundary at temperature T, or pulls the spans of a sealed gas longer, it leaves the same curve — and the curve is the same because the three things that make it are the three things the framework was built from.