PulseCore

Chapter 3 · Section 14

Galaxies: the Wells and the Stars Around Them

Where do the stars of a universe gather, and why do the gatherings take the shapes they do? In ours, nearly every star lives in a galaxy, a system of millions to trillions of stars held by the gravity of its dark matter, with a well at its centre. Galaxies come in a small number of shapes, and what decides the shape is known: spin, collisions, feedback and surroundings. This part sets out the types, the causes, the wells at their centres and the brightest things in the universe, the quasars, and reads each in Binary Pulse Theory.

The Hubble Sequence

Hubble sorted the galaxies he could resolve into a sequence that still frames the subject (Hubble, 1926; Hubble, 1936), which de Vaucouleurs extended (de Vaucouleurs, 1959):

  • Ellipticals (E0 to E7): smooth, round to elongated, of old red stars with little gas; their stars orbit every way at once rather than in a disk. The giants sit at the centres of clusters of galaxies.
  • Lenticulars (S0): a disk with no spiral arms and little gas, between the ellipticals and the spirals.
  • Spirals (Sa to Sd): a turning disk with arms, from Sa, with a large central bulge and tight arms, to Sd, with almost no bulge and loose arms. About two thirds carry a bar of stars across their centre (SB, and SAB for a weak bar); our own galaxy is a barred spiral.
  • Irregulars (Irr, Im): no settled shape, rich in gas and forming stars, often small or disturbed.

Beyond the Sequence

  • Dwarfs, by far the most numerous: dwarf ellipticals, dwarf spheroidals, the faintest of all and almost entirely dark matter, dwarf irregulars and blue compact dwarfs.
  • Ultra-diffuse galaxies: as wide as ours but a hundred times fainter (van Dokkum et al., 2015).
  • Starbursts: forming stars a hundred times faster than usual, most often after a collision.
  • Ring and peculiar galaxies: shaped by encounters, collected in Arp's atlas (Arp, 1966); a ring forms where a small galaxy has passed straight through a disk.
  • Active galaxies: galaxies whose central well is accreting, below.

What Sets the Type

Spin. Every clump of a young universe is spun up by the tidal pull of its neighbours (Part 3.12). Its spin is measured by a pure number, the spin parameter, which simulations and observations find spread about 0.035 (Peebles, 1969; Bullock et al., 2001).

Halo Spin Parameter G

λ⟲ = J⟲ · E⊖^(1/2) / (G · M^(5/2))

How fast a clump turns, against how fast it would have to turn to hold itself up

Where:

  • λ⟲ [∅] — the spin parameter (the Spin Parameter of Part 3.12, written there squared)
  • J⟲ [𝕄·𝕃²·𝕋⁻¹] — the clump's angular momentum
  • E⊖ [𝕄·𝕃²·𝕋⁻²] — the size of its total energy, which for a bound clump is negative
  • G [𝕄⁻¹·𝕃³·𝕋⁻²] — the gravitational constant
  • M [𝕄] — its mass

➢ A clump of λ⟲ = 1 would be held up by its turning alone; real ones turn at a few hundredths of that. It is small, but it decides everything after, because gas that cools keeps its spin.

Gas that cools in a spinning clump sinks until its spin holds it up, and settles into a disk whose size its spin sets (Fall & Efstathiou, 1980; Mo, Mao & White, 1998): the Disk Size of Part 3.12, its clump the halo.

Disk Size G

R⊘ = λ⟲ · R⊡ / √2

Dimensional analysis:
Where (from the book): R⊘ [𝕃], λ⟲ [∅], R⊡ [𝕃]
[𝕃]
[∅] × [𝕃] ÷ √[∅] = [𝕃] ÷ √[∅] = [𝕃] ÷ [∅] = [𝕃]
[𝕃] = [𝕃] ✓ Valid PulseCore Verified ✓

A disk is as wide as its halo's spin allows

Where:

  • R⊘ [𝕃] — the scale length of the disk
  • λ⟲ [∅] — the halo's spin parameter
  • R⊡ [𝕃] — the radius of the halo it settles in

➢ With λ⟲ ≈ 0.035 a disk is a few hundredths of its halo across, which is why disks are thin and small inside their great dark halos.

Collisions. When two disks of similar mass merge, their orderly orbits are scrambled and the result is an elliptical (Toomre & Toomre, 1972); a small galaxy absorbed by a large one thickens its disk and grows its bulge. Feedback. Supernovae and an accreting central well heat and blow out gas: a galaxy that keeps its gas goes on forming young blue stars, one that loses it turns red and dead. Surroundings. Crowded clusters are full of ellipticals and lenticulars and the open field of spirals, the morphology–density relation (Dressler, 1980).

How the Type Is Read

The shape of a galaxy is the share of its stars' motion that is orderly rotation. A disk turns together; an elliptical's stars swarm. The kinematic measure is the fraction of the stars' kinetic energy in rotation about the axis (Sales et al., 2012).

Rotation Fraction G

κ_rot = K_φ / K

Dimensional analysis:
Where (from the book): κ_rot [∅], K_φ [𝕄·𝕃²·𝕋⁻²], K [𝕄·𝕃²·𝕋⁻²]
[∅]
[𝕄·𝕃²·𝕋⁻²] ÷ [𝕄·𝕃²·𝕋⁻²] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

A disk turns together; a spheroid swarms

Where:

  • κ_rot [∅] — the fraction of the stars' kinetic energy in ordered rotation
  • K_φ [𝕄·𝕃²·𝕋⁻²] — the kinetic energy of their motion round the axis
  • K [𝕄·𝕃²·𝕋⁻²] — their whole kinetic energy

➢ Above about one half a galaxy is disk-dominated, a spiral or a lenticular; below it, spheroid-dominated, an elliptical. The game reads each galaxy's type this way, from the stars round its well: κ_rot and the share of young blue stars together give spiral (rotating, forming stars), lenticular (rotating, quiet), elliptical (swarming), irregular (half-settled, κ_rot 0.3 to 0.5, or rich in gas, forming stars at the spiral rate), dwarf (small) and starburst (forming stars fast). Its field is far coarser than any disk, so the game cannot measure a galaxy's turning from its matter, and it reads the type by the measure it can trust, its star formation, as the sky divides: the blue cloud of galaxies still forming stars are disks, spirals, or starbursts when they form fast; the red sequence of quenched ones are spheroids, ellipticals above the knee of the galaxy masses and lenticulars below it (Strateva et al., 2001; Wuyts et al., 2011). And a disk drawn at its true size would be a speck on that field, so each galaxy is drawn enlarged, nearly halfway to its nearest neighbour, its arms, its turning and its well at the centre of it in view.

What separates forming stars from quiet is measured. A galaxy's specific star formation rate, its new stars a year over the stars it holds, falls into two groups, and the line between them sits near 10⁻¹¹ a year: below it a galaxy is quenched (Wetzel, Tinker & Conroy, 2012). A disk forming stars as today's spirals do runs near 10⁻¹⁰ a year (Speagle et al., 2014), and a starburst forms them at four times that or more (Rodighiero et al., 2011). The game cannot time a year, so it reads the same rate from the stars themselves: a star of two Suns lives about 1.8 billion years (lifetime ∝ M^−2.5, Part 6.10), so the share of a galaxy's stars still above two Suns, set against that share in a newborn population (2.5% under the Kroupa mass function), is the share of its stars formed within one such lifetime. Multiplying each rate by that lifetime gives its line: below about 2% of the newborn share a galaxy is quiet, near 10% it forms stars as a spiral does, and above about 45% it is a starburst. A long-lived disk like ours holds only about one star in six hundred above two Suns, about 6% of the newborn share, and reads as forming stars, as it should.

A dwarf is small against the galaxies round it, not against the whole universe. Galaxy masses follow the Schechter function, φ(M) ∝ (M/M*)^α e^(−M/M*), many small galaxies and a knee M* above which they grow rare (Schechter, 1976); measured, the knee sits near 4.6 × 10¹⁰ Suns of stars and dwarfs below about 10⁹, about a fiftieth of it (Baldry et al., 2012). The game reads its own knee from its galaxies: for α ≈ −1.2 the mass-weighted mean galaxy is (α + 2) M* = 0.8 M*, so M* is 1.25 times that mean, and a galaxy below a fiftieth of M* is a dwarf. Each galaxy's stars are the bodies its well pulls hardest, so neighbours' stars are not counted in it.

How fast does a disk turn? Its clump collapsed to about two hundred times the mean density of its universe (178 in the spherical collapse of Gunn & Gott, 1972; 200 as simulations count it, Navarro, Frenk & White, 1996), so an orbit at the clump's edge turns at Ω = √(G·M/R³) = √(Δ/2)·H, about ten times the expansion rate H when it formed. The disk settles inside at R_d = λ⟲·R/√2 (Mo, Mao & White, 1998), on a rotation curve that stays flat (Rubin, Ford & Thonnard, 1980), so at twice its scale length it turns faster by R/(2R_d):

Disk Turn Rate G

Ω_d = √(Δᵥ / 2) · H_f / (√2 · λ⟲)

Dimensional analysis:
Where (from the book): Ω_d [𝕋⁻¹], Δᵥ [∅], H_f [𝕋⁻¹], λ⟲ [∅]
[𝕋⁻¹]
√([∅] ÷ [∅]) × [𝕋⁻¹] ÷ (√[∅] × [∅]) = √[∅] × [𝕋⁻¹] ÷ (√[∅] × [∅]) = [∅] × [𝕋⁻¹] ÷ (√[∅] × [∅]) = [𝕋⁻¹] ÷ (√[∅] × [∅]) = [𝕋⁻¹] ÷ ([∅] × [∅]) = [𝕋⁻¹] ÷ [∅] = [𝕋⁻¹]
[𝕋⁻¹] = [𝕋⁻¹] ✓ Valid PulseCore Verified ✓

How fast a galaxy's disk turns, set by when it formed

Where:

  • Ω_d [𝕋⁻¹] — the disk's angular speed at twice its scale length
  • Δᵥ [∅] — the density its clump collapsed to, over the universe's mean, about 200
  • H_f [𝕋⁻¹] — the expansion rate when the clump formed
  • λ⟲ [∅] — its spin parameter, about 0.035 (Halo Spin Parameter)

➢ Ω_d ≈ 202·H_f. Ours: our disk formed when the expansion ran about three times faster than today, so Ω_d ≈ 600·H₀ ≈ 1.4 × 10⁻¹⁵ per second, against the measured 235 km/s at twice its 2.6 kpc scale length, 1.5 × 10⁻¹⁵ (Bland-Hawthorn & Gerhard, 2016): one turn in about 150 million years, some sixty since our disk formed. A disk turns as often as its universe gives it, whatever the universe's size: with H = 2/(3t), Ω_d·t_f ≈ 135, about 21 turns in the first doubling of its age after it forms. Inside a disk each star turns at its own rate on the flat curve, Ω(r) = Ω_d·(2R_d/r), faster within, so the disk shears round its well, while the arms' pattern turns as one (Spiral Arms). The game cannot read this turning from its field, whose cells are far coarser than any disk: measured there, at level 133, the largest disk turned a thirtieth of a turn in its whole life. So it turns each galaxy by this law, from the moment its stars first gathered: each star on its own orbit, and the young stars of the arms with their pattern.

Spiral Arms

No one places a galaxy's arms. A disk turns faster near its centre than at its edge, so any clump in it is sheared into a trailing streak; while it shears, the disk's own gravity pulls more stars and gas into it, and for a moment the streak grows many times over before it winds away. This is swing amplification (Goldreich & Lynden-Bell, 1965; Julian & Toomre, 1966; Toomre, 1981), and it raises arms from the smallest unevenness, over and over, in every simulated disk cold and well resolved enough to hold them (Sellwood & Carlberg, 1984; Fujii et al., 2011; D'Onghia, Vogelsberger & Hernquist, 2013). A long-lived pattern can also run as a density wave (Lin & Shu, 1964). Gas passing through an arm is squeezed, shocks and forms stars there (Roberts, 1969), and the bright blue stars die within a few tens of millions of years, before they drift far, so the arms are lit by the young stars they make. Three numbers, all set by how the disk turns and how heavy it is, decide what the arms look like.

Shear Rate G

Γ = (R / Ω) · (Ω_drop / R_step)

Dimensional analysis:
Where (from the book): Γ [∅], R [𝕃], Ω [𝕋⁻¹], Ω_drop [𝕋⁻¹], R_step [𝕃]
[∅]
([𝕃] ÷ [𝕋⁻¹]) × ([𝕋⁻¹] ÷ [𝕃]) = [𝕃·𝕋] × ([𝕋⁻¹] ÷ [𝕃]) = [𝕃·𝕋] × [𝕃⁻¹·𝕋⁻¹] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How fast a disk's turning falls off outward

Where:

  • Γ [∅] — the shear rate: 0 for a disk turning as one, 1 for a flat rotation curve, 1.5 round a lone central mass
  • R [𝕃] — the distance from the centre
  • Ω [𝕋⁻¹] — the angular speed of the disk there
  • Ω_drop [𝕋⁻¹] — how much the angular speed falls over a short step outward
  • R_step [𝕃] — that step

➢ The negative logarithmic slope of the angular speed. It sets how fast a clump is sheared and so how tightly the arms wind.

Epicyclic Frequency G

κ_ep = Ω · √(2 · (2 − Γ))

Dimensional analysis:
Where (from the book): κ_ep [𝕋⁻¹], Ω [𝕋⁻¹], Γ [∅]
[𝕋⁻¹]
[𝕋⁻¹] × √([∅] × ([∅] - [∅])) = [𝕋⁻¹] × √([∅] × [∅]) = [𝕋⁻¹] × √[∅] = [𝕋⁻¹] × [∅] = [𝕋⁻¹]
[𝕋⁻¹] = [𝕋⁻¹] ✓ Valid PulseCore Verified ✓

How fast a nudged star swings back

Where:

  • κ_ep [𝕋⁻¹] — the epicyclic frequency, the rate at which a star pushed off its circle oscillates about it
  • Ω [𝕋⁻¹] — the angular speed of the disk
  • Γ [∅] — the shear rate

➢ Binney & Tremaine (2008). It is √2 Ω for a flat rotation curve and Ω round a lone central mass: the disk's resistance to being gathered into arms.

Arm Number G

m_a = κ_ep² · R / (4 · π · G · S_d)

Dimensional analysis:
Where (from the book): m_a [∅], κ_ep [𝕋⁻¹], R [𝕃], G [𝕃³·𝕄⁻¹·𝕋⁻²], S_d [𝕄·𝕃⁻²]
[∅]
[𝕋⁻²] × [𝕃] ÷ ([∅] × [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄·𝕃⁻²]) = [𝕃·𝕋⁻²] ÷ ([∅] × [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄·𝕃⁻²]) = [𝕃·𝕋⁻²] ÷ ([∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄·𝕃⁻²]) = [𝕃·𝕋⁻²] ÷ ([𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄·𝕃⁻²]) = [𝕃·𝕋⁻²] ÷ [𝕃·𝕋⁻²] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

Heavy disks grow few arms, light disks many

Where:

  • m_a [∅] — the number of arms swing amplification raises most
  • κ_ep [𝕋⁻¹] — the epicyclic frequency
  • R [𝕃] — the radius where the arms are read, two disk scale lengths
  • G [𝕃³·𝕄⁻¹·𝕋⁻²] — the gravitational constant
  • S_d [𝕄·𝕃⁻²] — the disk's surface density there

➢ Amplification is strongest at a wavelength twice the disk's critical one, λ_crit = 4π²G·S_d/κ_ep² (Toomre, 1981); fitting that wavelength round the circumference gives the count (Carlberg & Freedman, 1985; D'Onghia et al., 2013). A heavy, self-gravitating disk grows two or three grand arms; a light disk in a heavy halo grows many short ones, flocculent.

Arm Pitch Angle G

iₚ = arctan((2 / 7) · √(4 − 2 · Γ) / Γ)

Dimensional analysis:
Where (from the book): iₚ [∅], Γ [∅]
[∅]
arctan(([∅] ÷ [∅]) × √([∅] - [∅] × [∅]) ÷ [∅]) = arctan([∅] × √([∅] - [∅] × [∅]) ÷ [∅]) = arctan([∅] × √([∅] - [∅]) ÷ [∅]) = arctan([∅] × √[∅] ÷ [∅]) = arctan([∅] × [∅] ÷ [∅]) = arctan([∅] ÷ [∅]) = arctan([∅]) = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

The shear sets how tightly the arms wind

Where:

  • iₚ [∅] — the pitch angle, between an arm and the circle it crosses
  • Γ [∅] — the shear rate

➢ Michikoshi & Kokubo (2014), from swing amplification and N-body disks: about 22° for a flat rotation curve, about 11° round a dominant central mass, wider in a disk that turns nearly as one. Observed spirals run from about 5° to 40°, and a bigger central well goes with tighter arms (Seigar et al., 2008), as a steeper fall in Ω, a larger Γ, requires.

A disk only grows arms while it is cold enough to respond and resolved finely enough to hold them: simulated disks need a hundred thousand particles or more for arms to rise and last (Fujii et al., 2011).

A Well at Every Centre

Every large galaxy has a well at its centre, millions to billions of Suns, and the well's mass follows the swarm of the bulge around it so closely that the two must have grown together (Ferrarese & Merritt, 2000; Gebhardt et al., 2000; Kormendy & Ho, 2013).

Well–Bulge Relation G

M∅ = M◇ · (σ / σ◇)^(438 / 100)

Dimensional analysis:
Where (from the book): M◇ [𝕄], σ [𝕃·𝕋⁻¹], σ◇ [𝕃·𝕋⁻¹]
[𝕄]
[𝕄] × ([𝕃·𝕋⁻¹] ÷ [𝕃·𝕋⁻¹])^([∅] ÷ [∅]) = [𝕄] × ([𝕃·𝕋⁻¹] ÷ [𝕃·𝕋⁻¹])^([∅]) = [𝕄] × [∅]^([∅]) = [𝕄] × [∅] = [𝕄]
[𝕄] = [𝕄] ✓ Valid PulseCore Verified ✓

The central well grows with the swarm of stars round it

Where:

  • M∅ [𝕄] — the mass of the central well
  • M◇ [𝕄] — the reference mass, 3.1 × 10⁸ Suns
  • σ [𝕃·𝕋⁻¹] — the spread of speeds of the bulge's stars
  • σ◇ [𝕃·𝕋⁻¹] — the reference spread, 200 km/s

➢ Kormendy & Ho (2013). In Binary Pulse Theory the galaxy is the basin of its central well, the stars it gathers and the Data Matter that holds them (Part 3.13), so the well and its galaxy are one structure, grown together by the same infall.

Quasars and Active Galaxies

When the central well accretes fast, the gas spiralling in heats until it outshines the whole galaxy: a quasar, the brightest steady light in the universe, first recognised in 1963 (Schmidt, 1963) and explained as an accreting well (Salpeter, 1964; Lynden-Bell, 1969). Gas falling in turns about a tenth of its rest energy into light (Shakura & Sunyaev, 1973), and a spinning well throws two jets out along its spin axis (Blandford & Znajek, 1977).

Accretion Luminosity G

L = η · M′ · c²

Dimensional analysis:
Where (from the book): L [𝕄·𝕃²·𝕋⁻³], η [∅], M′ [𝕄·𝕋⁻¹], c [𝕃·𝕋⁻¹]
[𝕄·𝕃²·𝕋⁻³]
[∅] × [𝕄·𝕋⁻¹] × [𝕃²·𝕋⁻²] = [𝕄·𝕋⁻¹] × [𝕃²·𝕋⁻²] = [𝕄·𝕃²·𝕋⁻³]
[𝕄·𝕃²·𝕋⁻³] = [𝕄·𝕃²·𝕋⁻³] ✓ Valid PulseCore Verified ✓

An accreting well turns a tenth of what falls in into light

Where:

  • L [𝕄·𝕃²·𝕋⁻³] — the light given off
  • η [∅] — the share of rest energy radiated, about 0.1
  • M′ [𝕄·𝕋⁻¹] — the mass falling in each unit of time
  • c [𝕃·𝕋⁻¹] — light speed

➢ Ten times more efficient than nuclear fusion. A quasar taking in a few Suns a year outshines a trillion stars. In Binary Pulse Theory this light is matter's own, made outside any horizon, and so it is not the vacuum's (What Returns to the Vacuum, Part 3.13): a quasar's blaze goes out into its universe and comes home only as its photons do, through the wells or at the close.

Eddington Luminosity G

L⥣ = 4 · π · G · M · mₚ · c / σ_T

Dimensional analysis:
Where (from the book): L⥣ [𝕄·𝕃²·𝕋⁻³], G [𝕄⁻¹·𝕃³·𝕋⁻²], M [𝕄], c [𝕃·𝕋⁻¹], σ_T [𝕃²]
[𝕄·𝕃²·𝕋⁻³]
[∅] × [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄] × [𝕄] × [𝕃·𝕋⁻¹] ÷ [𝕃²] = [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄] × [𝕄] × [𝕃·𝕋⁻¹] ÷ [𝕃²] = [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄] × [𝕄] × [𝕃·𝕋⁻¹] ÷ [𝕃²] = [𝕃³·𝕋⁻²] × [𝕄] × [𝕃·𝕋⁻¹] ÷ [𝕃²] = [𝕄·𝕃³·𝕋⁻²] × [𝕃·𝕋⁻¹] ÷ [𝕃²] = [𝕄·𝕃⁴·𝕋⁻³] ÷ [𝕃²] = [𝕄·𝕃²·𝕋⁻³]
[𝕄·𝕃²·𝕋⁻³] = [𝕄·𝕃²·𝕋⁻³] ✓ Valid PulseCore Verified ✓

The most a well can shine before its light holds back its food

Where:

  • L⥣ [𝕄·𝕃²·𝕋⁻³] — the Eddington luminosity
  • G [𝕄⁻¹·𝕃³·𝕋⁻²] — the gravitational constant
  • M [𝕄] — the well's mass
  • mₚ [𝕄] — the proton mass
  • c [𝕃·𝕋⁻¹] — light speed
  • σ_T [𝕃²] — the Thomson cross-section of an electron

➢ Eddington (1926). A well accreting near this limit is a quasar; one accreting at a small part of it is a Seyfert galaxy (Seyfert, 1943), a bright core in an ordinary spiral; one whose jets reach far beyond its stars into great lobes is a radio galaxy. Collisions pour gas inward, so mergers light quasars, and the light and jets that follow are feedback that ends the burst.

What Supplies a Well: Torques in the Disk

A well shines only from gas: data and Data Matter fall in dark, and a merger adds mass without a disk to heat. The gas cannot fall straight in, because it orbits; it must first lose its spin, and in a galaxy what takes it away are the gravitational torques of the disk's own lopsided and barred patterns. How fast those torques bring gas in sets how fast a well is supplied (Hopkins & Quataert, 2011):

Torque-Limited Accretion G

Ṁ_T = α_T · f_d^(5/2) · m₈^(1/6) · m₉ · r_100^(−3/2) · Ṁ₁ / (1 + f₀ / f_gas)

Dimensional analysis:
Where (from the book): Ṁ_T [𝕄·𝕋⁻¹], α_T [∅], f_d [∅], m₈ [∅], m₉ [∅], r_100 [∅], Ṁ₁ [𝕄·𝕋⁻¹], f_gas [∅]
[𝕄·𝕋⁻¹]
[∅] × [∅]^([∅] ÷ [∅]) × [∅]^([∅] ÷ [∅]) × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅]) × [∅]^([∅] ÷ [∅]) × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] × [∅]^([∅] ÷ [∅]) × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅] ÷ [∅]) × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅]) × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅] ÷ [∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅]) × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [𝕄·𝕋⁻¹] ÷ ([∅] + [∅] ÷ [∅]) = [𝕄·𝕋⁻¹] ÷ ([∅] + [∅]) = [𝕄·𝕋⁻¹] ÷ [∅] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

The gas a galaxy's disk torques bring in toward its well

Where:

  • Ṁ_T [𝕄·𝕋⁻¹] — the inflow
  • α_T [∅] — the normalization, about 5
  • f_d [∅] — the disk's share of the mass inside R₀
  • m₈ [∅] — the well's mass over 10⁸ Suns
  • m₉ [∅] — the disk's mass inside R₀ over 10⁹ Suns
  • r_100 [∅] — R₀ over 100 parsecs
  • Ṁ₁ [𝕄·𝕋⁻¹] — one Sun a year
  • f₀ [∅] — the gas fraction below which the inflow falls
  • f_gas [∅] — the gas fraction of the disk inside R₀

➢ Hopkins & Quataert (2011), in the form and normalization of Anglés-Alcázar, Özel & Davé (2013). It rises steeply with the disk share and depends only weakly on the well, so wells grow with their galaxies without being tuned to.

Torque Gas Scale G

f₀ = 0.31 · f_d² · m₉^(−1/3)

Dimensional analysis:
Where (from the book): f_d [∅], m₉ [∅]
[∅]
[∅] × [∅] × [∅]^([∅] ÷ [∅]) = [∅] × [∅]^([∅] ÷ [∅]) = [∅] × [∅]^([∅]) = [∅] × [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

Below this gas fraction the torques supply less

Where:

  • f₀ [∅] — the gas fraction scale
  • f_d [∅] — the disk share
  • m₉ [∅] — the disk's mass over 10⁹ Suns

➢ Anglés-Alcázar, Özel & Davé (2013). Below f₀ the inflow falls with the gas fraction as 1/(1 + f₀/f_gas): a gas-poor disk cuts off its well's supply however it turns, which is one reason the quasar era ends as galaxies use up their gas.

Well Supply G

Ṁ_w = εₘ · Ṁ_T

Dimensional analysis:
Where (from the book): Ṁ_w [𝕄·𝕋⁻¹], εₘ [∅], Ṁ_T [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹]
[∅] × [𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

The share of the inflow that reaches the well

Where:

  • Ṁ_w [𝕄·𝕋⁻¹] — the well's accretion
  • εₘ [∅] — the share retained, about 5 %; the rest forms stars or is blown out on the way
  • Ṁ_T [𝕄·𝕋⁻¹] — the torque-limited inflow

➢ Anglés-Alcázar, Özel & Davé (2013); with it, Null Well and galaxy masses follow each other in simulations without feedback tuned to make them (Anglés-Alcázar et al., 2017). Capped at the Eddington rate, a well of 10⁸ Suns taking 2.2 Suns a year at a tenth of its light.

Measured against what the well could take, the ratio that decides a Seyfert (from a hundredth) or a quasar (from a tenth), this reads, for the reference disk of 10⁹ Suns within 100 parsecs:

Torque Eddington Ratio G

λ_E = 0.113 · f_d^(5/2) · (10 · M_w / M_d)^(−5/6) / (1 + f₀ / f_gas)

Dimensional analysis:
Where (from the book): λ_E [∅], f_d [∅], M_w [𝕄], M_d [𝕄], f_gas [∅]
[∅]
[∅] × [∅]^([∅] ÷ [∅]) × ([∅] × [𝕄] ÷ [𝕄])^([∅] ÷ [∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅]) × ([∅] × [𝕄] ÷ [𝕄])^([∅] ÷ [∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] × ([∅] × [𝕄] ÷ [𝕄])^([∅] ÷ [∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × ([∅] × [𝕄] ÷ [𝕄])^([∅] ÷ [∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × ([∅] × [𝕄] ÷ [𝕄])^([∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × ([𝕄] ÷ [𝕄])^([∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅]^([∅]) ÷ ([∅] + [∅] ÷ [∅]) = [∅] × [∅] ÷ ([∅] + [∅] ÷ [∅]) = [∅] ÷ ([∅] + [∅] ÷ [∅]) = [∅] ÷ ([∅] + [∅]) = [∅] ÷ [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How brightly a well supplied by its disk shines

Where:

  • λ_E [∅] — the Eddington ratio, the accretion over the Eddington rate
  • f_d [∅] — the disk share
  • M_w [𝕄] — the well's mass
  • M_d [𝕄] — the disk's mass inside R₀
  • f₀ [∅] — the gas fraction scale
  • f_gas [∅] — the disk's gas fraction

➢ 0.113 = ε_m · α_T / 2.22, the Eddington accretion of a 10⁸-Sun well being 2.22 Suns a year. A small well in a gas-rich disk blazes; a well grown heavy against its disk, or a disk that has used its gas, goes quiet.

What the sky measures. The torques that bring gas to a well are the same torques that gather it into stars, so a well's growth follows its galaxy's star formation. Stacking the X-rays of thousands of star-forming galaxies, a central well grows at about a thousandth of its galaxy's star-formation rate, the same at every stellar mass and unchanged back to redshift 2, which is just what keeps it at about a thousandth of the stars it lives among (Mullaney et al., 2012); summed over the sky, the history of Null Well growth follows the history of star formation (Madau & Dickinson, 2014).

Wells Grow With Their Stars G

Ṁ_w = ε_★ · Ṁ_★

Dimensional analysis:
Where (from the book): Ṁ_w [𝕄·𝕋⁻¹], ε_★ [∅], Ṁ_★ [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹]
[∅] × [𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

A central well grows at about a thousandth of its galaxy's star formation

Where:

  • Ṁ_w [𝕄·𝕋⁻¹] — the well's growth by gas
  • ε_★ [∅] — about 10⁻³, measured
  • Ṁ_★ [𝕄·𝕋⁻¹] — the galaxy's star-formation rate

➢ Mullaney et al. (2012). An average over galaxies: a single well flickers above and below it as gas comes and goes. In the game this sets each well's average supply from its galaxy's star formation every four Pulses, and the Eddington Ratio Distribution puts the flicker on it.

Eddington Ratio of a Well G

λ_E = Ṁ_w · t_S / M_w

Dimensional analysis:
Where (from the book): λ_E [∅], Ṁ_w [𝕄·𝕋⁻¹], t_S [𝕋], M_w [𝕄]
[∅]
[𝕄·𝕋⁻¹] × [𝕋] ÷ [𝕄] = [𝕄] ÷ [𝕄] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How brightly a well shines against what it could take

Where:

  • λ_E [∅] — the Eddington ratio: a Seyfert from a hundredth, a quasar from a tenth
  • Ṁ_w [𝕄·𝕋⁻¹] — its growth by gas
  • t_S [𝕋] — the Salpeter time, about 45 million years at a tenth of its light
  • M_w [𝕄] — its mass

➢ A well on the Well–Bulge Relation, a thousandth of its stars, shines at about its galaxy's specific star formation times the Salpeter time: near 2 × 10⁻⁹ a year at cosmic noon, a tenth, the quasar era; near 3 × 10⁻¹¹ a year for the Milky Way today, about a thousandth, quiet.

What the data adds. The data that falls into a well stays in it, held inside the horizon all through the life, until it comes out on the other side as the next universe's Pulsecore (Return to Bifurcation). But a physical well's data is Data Matter, and Data Matter is collisionless: like the dark matter it is, it cannot shed its spin by rubbing or radiating, so it reaches the horizon only on orbits that pass within 4GM/(c·v) of the well:

Capture of the Data G

Ṁ◍ = 16 · π · G² · M_w² · ρ◍ / (c² · v)

Dimensional analysis:
Where (from the book): Ṁ◍ [𝕄·𝕋⁻¹], G [𝕃³·𝕄⁻¹·𝕋⁻²], M_w [𝕄], c [𝕃·𝕋⁻¹], v [𝕃·𝕋⁻¹]
[𝕄·𝕋⁻¹]
[∅] × [∅] × [𝕄⁻²·𝕃⁶·𝕋⁻⁴] × [𝕄²] × [𝕄·𝕃⁻³] ÷ ([𝕃²·𝕋⁻²] × [𝕃·𝕋⁻¹]) = [∅] × [𝕄⁻²·𝕃⁶·𝕋⁻⁴] × [𝕄²] × [𝕄·𝕃⁻³] ÷ ([𝕃²·𝕋⁻²] × [𝕃·𝕋⁻¹]) = [𝕄⁻²·𝕃⁶·𝕋⁻⁴] × [𝕄²] × [𝕄·𝕃⁻³] ÷ ([𝕃²·𝕋⁻²] × [𝕃·𝕋⁻¹]) = [𝕃⁶·𝕋⁻⁴] × [𝕄·𝕃⁻³] ÷ ([𝕃²·𝕋⁻²] × [𝕃·𝕋⁻¹]) = [𝕄·𝕃³·𝕋⁻⁴] ÷ ([𝕃²·𝕋⁻²] × [𝕃·𝕋⁻¹]) = [𝕄·𝕃³·𝕋⁻⁴] ÷ [𝕃³·𝕋⁻³] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

How fast a well takes in the collisionless Data Matter round it

Where:

  • Ṁ◍ [𝕄·𝕋⁻¹] — the Data Matter the well captures
  • G [𝕃³·𝕄⁻¹·𝕋⁻²] — the gravitational constant
  • M_w [𝕄] — the well's mass
  • ρ◍ [𝕄·𝕃⁻³] — the density of Data Matter round it
  • c [𝕃·𝕋⁻¹] — light speed
  • v [𝕃·𝕋⁻¹] — the speed of the Data Matter, its halo's

➢ The capture cross-section of a Null Well for slow particles, 16π(GM/c²)²(c/v)² (Shapiro & Teukolsky, 1983). Against the gas, which can lose its spin, it is millions of times slower, which is why the wells we weigh, data and all, come to about a thousandth of their stars.

Wells flicker. The supply above is an average. A single well is not supplied steadily: the gas reaches it in spells, and between them it is quiet. Measured across twenty-five thousand galaxies, the chance of finding a well at each Eddington ratio is one power law, the same at every stellar mass, falling as λ_E^(−0.65) up to the Eddington limit (Aird et al., 2012), which breaks there (Aird et al., 2013):

Eddington Ratio Distribution G

p(λ_E) = p₀ · λ_E^(−0.65)

Dimensional analysis:
Where (from the book): p(λ_E) [∅], λ_E [∅], p₀ [∅]
[∅]
[∅] × [∅]^([∅]) = [∅] × [∅]^([∅]) = [∅] × [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How often a well is found at each brightness

Where:

  • p(λ_E) [∅] — the chance of an Eddington ratio, per interval of its logarithm
  • λ_E [∅] — the Eddington ratio, from about 10⁻⁴ to 1
  • p₀ [∅] — the normalization, the active share of wells

➢ Aird et al. (2012), from 242 X-ray active nuclei among about 25,000 galaxies at 0.2 < z < 1. Per decade of λ_E it rises toward the faint end, so a well spends most of its time far below its peaks, and its mean lies over two decades beneath them.

Well Flicker G

λ_E = ⟨λ_E⟩ · ξ

Dimensional analysis:
Where (from the book): λ_E [∅], ⟨λ_E⟩ [∅], ξ [∅]
[∅]
[∅] × [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

A well's brightness at a moment: its average, times the measured flicker

Where:

  • λ_E [∅] — its Eddington ratio at that moment, at most 1
  • ⟨λ_E⟩ [∅] — its average, from its galaxy's star formation
  • ξ [∅] — the flicker, drawn from the Eddington Ratio Distribution over four decades, 0.022 to 222, its mean one

➢ The shape is the sky's, the same for every well; its place is the well's own average. At ⟨λ_E⟩ = 0.005 a well shines as a Seyfert or brighter about 5 % of the time and as a quasar about 1 %; at cosmic noon, ⟨λ_E⟩ near 0.1, about 38 % and 8 %; in a quiet galaxy today, ⟨λ_E⟩ near 0.001, about 1.6 % and 0.2 %: why quasars are rare and brilliant, and were commonest at cosmic noon.

The game's reading. A galaxy in the game is too coarse to show its inner disk turning, where the torques work: read from the game's own orbits, the Torque-Limited law supplied wells about ten thousand times too little, and read at its reference disk with the game's whole disk it made wells as heavy as their stars. So the game supplies each well as the sky measures: every four Pulses it reads how much of its galaxy's gas became stars, and the well takes a thousandth of that from the disk's gas, never faster than its Eddington rate; its Eddington ratio, and with it whether it is a Seyfert or a quasar, is read from that supply and from matter it takes in, never from data falling in dark. That supply is spent as the sky spends it: every four Pulses each well's Eddington ratio is drawn by Well Flicker, its average times the measured flicker, so its mass grows on average by its galaxy's thousandth while its light comes and goes. A Proto Well, in the data layer, takes in the data that reaches it; from the physical layer a well takes in Data Matter by Capture of the Data, and the data streaming past it goes home through it uncounted in its mass. Measured at level 25: wells a hundredth of their stars early, falling toward the sky's thousandth, and their supply just below a Seyfert's.

Feedback: What Keeps Galaxies Blue

Without feedback a galaxy forms its stars too fast, too early, and too many: gas cools, falls in and turns into stars until none is left, and every galaxy ends red and dead. Early simulations made exactly this mistake, the overcooling problem (White & Frenk, 1991). Two sources of energy push back, and with them the galaxies come out as observed.

Supernovae. Each massive star ends by throwing 10⁵¹ erg into the gas around it. Most of that energy radiates away, but the shell it drives carries a nearly fixed momentum out through the gas, heating it, stirring it, and in a small galaxy blowing much of it out altogether (Dekel & Silk, 1986).

Supernova Momentum G

p_sn = p◇ · (n_H / n◇)^(−2 / 17)

Dimensional analysis:
Where (from the book): p_sn [𝕄·𝕃·𝕋⁻¹], p◇ [𝕄·𝕃·𝕋⁻¹], n_H [𝕃⁻³], n◇ [𝕃⁻³]
[𝕄·𝕃·𝕋⁻¹]
[𝕄·𝕃·𝕋⁻¹] × ([𝕃⁻³] ÷ [𝕃⁻³])^([∅] ÷ [∅]) = [𝕄·𝕃·𝕋⁻¹] × ([𝕃⁻³] ÷ [𝕃⁻³])^([∅]) = [𝕄·𝕃·𝕋⁻¹] × [∅]^([∅]) = [𝕄·𝕃·𝕋⁻¹] × [∅] = [𝕄·𝕃·𝕋⁻¹]
[𝕄·𝕃·𝕋⁻¹] = [𝕄·𝕃·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

The momentum one supernova drives into the gas around it

Where:

  • p_sn [𝕄·𝕃·𝕋⁻¹] — the momentum the remnant's shell carries when it stops sweeping up gas
  • p◇ [𝕄·𝕃·𝕋⁻¹] — the reference momentum, about 2.8 × 10⁵ Suns at 1 km/s
  • n_H [𝕃⁻³] — the density of the gas it explodes in
  • n◇ [𝕃⁻³] — the reference density, one atom per cubic centimetre

➢ Kim & Ostriker (2015), from simulations of single remnants: the momentum hardly depends on the gas, falling only as the −2/17 power of its density, so every supernova delivers about the same push. It is enough to move a hundred thousand Suns of gas at a few kilometres a second, and a galaxy's supernovae together hold its gas up against its own gravity.

Quasars. An accreting well gives off a tenth of the rest energy of what it takes in as light (Accretion Luminosity, above), and a few per cent of that couples to the galaxy's gas as winds and heat. When the well is supplied fast, in a quasar, the wind clears the gas from the galaxy's centre and cuts off its own supply; when it is supplied slowly, its jets keep the gas round a massive galaxy too hot to fall in and form stars. This is what turns the largest galaxies red and keeps them so (Di Matteo, Springel & Hernquist, 2005; Croton et al., 2006; Bower et al., 2006), and it is why a well's mass follows its bulge: the well grows until its own feedback stops the gas that supplies both (Silk & Rees, 1998).

Quasar Feedback Energy G

E_fb = ε_f · L · Δt

Dimensional analysis:
Where (from the book): E_fb [𝕄·𝕃²·𝕋⁻²], ε_f [∅], L [𝕄·𝕃²·𝕋⁻³]
[𝕄·𝕃²·𝕋⁻²]
[∅] × [𝕄·𝕃²·𝕋⁻³] × [𝕋] = [𝕄·𝕃²·𝕋⁻³] × [𝕋] = [𝕄·𝕃²·𝕋⁻²]
[𝕄·𝕃²·𝕋⁻²] = [𝕄·𝕃²·𝕋⁻²] ✓ Valid PulseCore Verified ✓

The share of an accreting well's light that stirs its galaxy

Where:

  • E_fb [𝕄·𝕃²·𝕋⁻²] — the energy given to the galaxy's gas
  • ε_f [∅] — the share of the light that couples to the gas, about 0.05
  • L [𝕄·𝕃²·𝕋⁻³] — the well's accretion luminosity
  • Δt [𝕋] — the time it accretes

➢ Di Matteo, Springel & Hernquist (2005) found that ε_f ≈ 0.05 reproduces the Well–Bulge Relation in simulated mergers, the quasar shining until its wind stops its supply.

In Binary Pulse Theory feedback is the data of a star or a well's intake returned to the field and pushed back out before the wells take what is left: the supernova's shell carries its star's matter and the data it holds back into the gas (Part 2.9), and the quasar's wind sends what its well could not take in back out across its galaxy. The simulation gives each supernova an outward push on the gas round it and keeps that gas too hot to form stars for a while, and each well accreting past a hundredth of its Eddington rate a push and heating in proportion; the reach of both is scaled to what the device can compute, so their effect is seen in the universes deep enough to have stars, from about level 23 (the Star Existence Condition, Part 2.10).

Wells per Galaxy

The progeny law gives the number of wells a universe forms in its life, 2^(L/3) (Part 3.10), and most of a universe's wells are the small ones its dying stars leave (Part 2.9), scattered through its galaxies, with one great well at each centre.

Wells per Galaxy G

n∅ = 2^(L / 3) / N_g

Dimensional analysis:
Where (from the book): n∅ [∅], L [∅], N_g [∅]
[∅]
[∅]^([∅] ÷ [∅]) ÷ [∅] = [∅]^([∅]) ÷ [∅] = [∅] ÷ [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

The wells a galaxy holds, from the progeny law

Where:

  • n∅ [∅] — the wells in a typical galaxy
  • L [∅] — the level, 202 for ours
  • N_g [∅] — the galaxies the universe holds

➢ At our level 2^(202/3) ≈ 1.9 × 10²⁰ wells. With 2 × 10¹² galaxies in the observable universe (Conselice et al., 2016), or the few hundred billion that measurements of the dark sky's light now favor (Lauer et al., 2021), that is 10⁸ to 10⁹ wells a galaxy, and our own galaxy is estimated to hold about a hundred million Null Wells from dead stars (Elbert, Bullock & Kaplinghat, 2018). The progeny law, derived for a universe's children, lands on the count of its stellar wells.

Proto Wells: Where Wells Will Form

No Null Well has ever been seen to form except by collapse: a star's core falling in, with a supernova or without one, as a red supergiant of about 25 Suns in the galaxy NGC 6946 brightened in 2009 and then vanished (Gerke, Kochanek & Stanek, 2015; Adams et al., 2017), or, in the first few hundred million years, a whole cloud of gas falling in at once. Wells that would form from the early universe's own density, the primordial black holes (Hawking, 1971; Carr & Hawking, 1974), are allowed in a few windows of mass and have never been found (Carr & Kühnel, 2020). So a Proto Well is not a well. Formed in the data before the physical layer, it is the deepest place of the Data Matter, the seed of a halo (Part 3.13), and at the Physical Nova it becomes a site: where the gas will gather, where the first stars will light, and where a well will be born. It is the data's prediction of where the wells of matter will form, as the peaks of the early density field are where halos collapse in standard cosmology (Bardeen, Bond, Kaiser & Szalay, 1986). A well is born at a site in one of two ways.

By direct collapse, early. A site makes its well at once when three things hold together (reviewed by Woods et al., 2019). It is deep enough that gas falling in is heated to where atomic hydrogen radiates, T_vir ≥ 10⁴ K by the Virial Temperature (Part 3.12), a circular speed of about 17 km/s. Its gas is pristine, its metals below about 10⁻⁵ of the Sun's, so neither metal lines nor dust can cool it further (Omukai, Schneider & Haiman, 2008), which only the first gas is, before supernovae have enriched it. And a neighbour already shines, its ultraviolet breaking the molecular hydrogen that would otherwise cool the gas to a few hundred kelvin (Shang, Bryan & Haiman, 2010; the synchronised pairs of Visbal, Haiman & Bryan, 2014). The gas then holds at the atomic line, about 8,000 K, and falls in at its sound speed cubed over G, the inflow of an isothermal sphere (Shu, 1977).

Sound Speed of Warm Gas G

cₛ = √(k_B · T / (μ · mₚ))

Dimensional analysis:
Where (from the book): cₛ [𝕃·𝕋⁻¹], T [Θ], μ [∅]
[𝕃·𝕋⁻¹]
√([𝕄·𝕃²·𝕋⁻²·Θ⁻¹] × [Θ] ÷ ([∅] × [𝕄])) = √([𝕄·𝕃²·𝕋⁻²] ÷ ([∅] × [𝕄])) = √([𝕄·𝕃²·𝕋⁻²] ÷ [𝕄]) = √[𝕃²·𝕋⁻²] = [𝕃·𝕋⁻¹]
[𝕃·𝕋⁻¹] = [𝕃·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

How fast a pressure wave crosses gas at temperature T

Where:

  • cₛ [𝕃·𝕋⁻¹] — the sound speed
  • k_B [𝕄·𝕃²·𝕋⁻²·Θ⁻¹] — Boltzmann's constant
  • T [Θ] — the gas temperature, about 8,000 K held by atomic hydrogen
  • μ [∅] — the mean mass per particle in proton masses, about 1.2 for neutral primordial gas
  • mₚ [𝕄] — the proton mass

➢ At 8,000 K, c_s ≈ 7.4 km/s; at 200 K, where molecular hydrogen cools the first gas, 1.2 km/s. It enters the Collapse Inflow Rate cubed, so the inflow rises steeply with the gas's temperature: atomic gas at 8,000 K supplies a site about two hundred times faster than molecular gas at 200 K.

Collapse Inflow Rate G

Ṁ_in = 0.975 · c_s³ / G

Dimensional analysis:
Where (from the book): Ṁ_in [𝕄·𝕋⁻¹], c_s [𝕃·𝕋⁻¹], G [𝕄⁻¹·𝕃³·𝕋⁻²]
[𝕄·𝕋⁻¹]
[∅] × [𝕃³·𝕋⁻³] ÷ [𝕄⁻¹·𝕃³·𝕋⁻²] = [𝕃³·𝕋⁻³] ÷ [𝕄⁻¹·𝕃³·𝕋⁻²] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

The gas a site draws in while it collapses

Where:

  • Ṁ_in [𝕄·𝕋⁻¹] — the inflow into the site's centre
  • cₛ [𝕃·𝕋⁻¹] — the warm gas's sound speed, or in a deeper site the site's v_c / √2
  • G [𝕄⁻¹·𝕃³·𝕋⁻²] — the gravitational constant

➢ Shu (1977). At 8,000 K about 0.1 Suns a year, and more in a deeper site, as simulations of atomic-cooling halos find (Latif et al., 2013); in gas at 200 K about 0.0004, which makes the first ordinary stars. Above about 0.04 Suns a year the gas arrives faster than a protostar can contract, and it swells into a supermassive star instead (Hosokawa et al., 2013; Haemmerlé et al., 2018). The ratio Ṁ_in / 0.04 Suns a year is how far above the line a site stands.

Direct Collapse Seed G

M_dc = Ṁ_in · t_★

Dimensional analysis:
Where (from the book): M_dc [𝕄], Ṁ_in [𝕄·𝕋⁻¹], t_★ [𝕋]
[𝕄]
[𝕄·𝕋⁻¹] × [𝕋] = [𝕄]
[𝕄] = [𝕄] ✓ Valid PulseCore Verified ✓

The well a supermassive star leaves when it collapses

Where:

  • M_dc [𝕄] — the seed well, born at the site's centre
  • Ṁ_in [𝕄·𝕋⁻¹] — the Collapse Inflow Rate
  • t_★ [𝕋] — the supermassive star's life, a million years or two

➢ About 10⁵ to 10⁶ Suns, with no supernova (Woods et al., 2019), never more than the site's gas can give. Such a seed was seen in UHZ1, a well of about 4 × 10⁷ Suns shining 470 million years after the big bang, as heavy as all its galaxy's stars (Bogdán et al., 2024).

By its first dying star. Everywhere else the gas cools, breaks into stars, and the site waits for them: its well is born when its first heavy star dies and the remnant sinks to its centre (Madau & Rees, 2001; Volonteri, 2010). Most sites wait, and many never fill: nearly every large galaxy holds a well at its centre, while among the dwarfs only some do (Greene, Strader & Ho, 2020).

The central well grows from light. A central well is one well. It is supplied by its galaxy (Wells Grow With Their Stars, flickering by the Eddington Ratio Distribution), but never faster than it can grow, e-folding once in a Salpeter time, t_S ≈ 45 million years (Salpeter, 1964):

Eddington Growth of a Central Well G

Ṁ_E = Mₕ / t_S

Dimensional analysis:
Where (from the book): Ṁ_E [𝕄·𝕋⁻¹], Mₕ [𝕄], t_S [𝕋]
[𝕄·𝕋⁻¹]
[𝕄] ÷ [𝕋] = [𝕄·𝕋⁻¹]
[𝕄·𝕋⁻¹] = [𝕄·𝕋⁻¹] ✓ Valid PulseCore Verified ✓

The fastest a central well can grow

Where:

  • Ṁ_E [𝕄·𝕋⁻¹] — its growth at the Eddington limit
  • Mₕ [𝕄] — the central well's own mass
  • t_S [𝕋] — the Salpeter time, about 45 million years

➢ A central well grows at the lesser of what its galaxy supplies it and Ṁ_E. Born light, a remnant against a whole galaxy, it is held to the limit and blazes as a quasar until it has grown to what its galaxy supplies; then it follows its stars, a Seyfert at times, quiet most of its life. A seed of a hundred Suns needs about sixteen e-folds, 0.7 billion years, to reach the 8 × 10⁸ Suns of the quasar seen 690 million years after the big bang (Bañados et al., 2018): light seeds alone barely make the earliest giants, and the heavy seeds of direct collapse make them with room.

The prediction scored. If the Proto Wells are where wells will form, wells are born inside sites far more often than chance:

Site Lift G

Λ_s = N_in / (N_w · A_s)

Dimensional analysis:
Where (from the book): Λ_s [∅], N_in [∅], N_w [∅], A_s [∅]
[∅]
[∅] ÷ ([∅] × [∅]) = [∅] ÷ [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How much more often wells are born inside sites than by chance

Where:

  • Λ_s [∅] — the lift: 1 is chance
  • N_in [∅] — wells born within a site's core, the radius holding half its matter
  • N_w [∅] — all wells born
  • A_s [∅] — the share of the surface within the sites' cores

➢ Read in every universe the game plays, with the share of sites holding their well over its life. A lift well above 1 is the data's prediction holding: the wells of matter form where the scaffold's Proto Wells stood, as the peaks of the early density field are where halos collapse in standard cosmology; a lift near 1 would mean the sites foretold nothing.

The Count, and the Wells Unseen

The progeny law asks for one well for every 65 stars at every level: a universe forms 925·(m_P/m_p) stars (Part 2.10) against 2^(L/3) wells, since m_P/m_p = 2^(L/3)/14.3. A Kroupa population ends about one star in six hundred as a stellar well (Part 2.9), and the models of our sky about one in 500: 4 × 10¹⁹ stellar Null Wells (Sicilia et al., 2022) against about 2 × 10²² stars. The law calls out wells the sky has not counted, and the sky has reasons to have missed them. A Null Well alone is dark: of all those in our galaxy, one isolated one has been found, by the starlight it bent (Sahu et al., 2022). And the stars seen to explode reach only about 17 Suns, though stars up to 25 or 30 should explode the same way, the red supergiant problem (Smartt, 2009 and 2015), as if many heavier stars collapse without exploding, as the one in NGC 6946 seems to have done. At one well for every 65 stars our galaxy holds 1.5 to 6 billion wells, 15 to 60 times the hundred million the models give (Elbert, Bullock & Kaplinghat, 2018): a count the microlensing survey of the Nancy Grace Roman Space Telescope can make. The game counts its wells as the sky would and lets the gap show: a well born of a dying star counts for the stars its drawn star stands for, and a central well, or a clump of gas collapsing whole, counts as one well, since a cloud the game cannot resolve is one collapse and not a share of the law. It reports what it counted against what the law asks, and the difference is the wells unseen. At level 202, ours, it counts about 10¹⁹, nearly all of them stellar, where the law asks 1.9 × 10²⁰ and the sky's models give 4 × 10¹⁹.

Most sites are not a galaxy's centre. A galaxy is a halo, and one halo gathers many: the deepest at its centre, the others satellites sinking toward it by dynamical friction (Chandrasekhar, 1943), with their stars, gas and wells, and many small halos never light a galaxy at all (Sawala et al., 2016). And a well is seen by what it does, never by itself: one accreting gas blazes as a quasar or a Seyfert, a quiet one is known only by the stars that orbit it, as ours was found from the orbits of the stars round it (Ghez et al., 2008; Genzel, Eisenhauer & Gillessen, 2010), and a stellar well alone is dark.

Ascendant Wells

Every well seeds a child, and what the child is born holding is its well's information (Return to Bifurcation): a well records one bit for every 4·ln2 Planck areas of its horizon (Bekenstein, 1973; Hawking, 1975), so its information grows with the square of its mass.

Well Information G

I_w = 4 · π · G · M² / (ħ · c · ln(2))

Dimensional analysis:
Where (from the book): I_w [∅], G [𝕃³·𝕄⁻¹·𝕋⁻²], M [𝕄], ħ [𝕄·𝕃²·𝕋⁻¹], c [𝕃·𝕋⁻¹]
[∅]
[∅] × [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄²] ÷ ([𝕄·𝕃²·𝕋⁻¹] × [𝕃·𝕋⁻¹] × [∅]) = [∅] × [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄²] ÷ ([𝕄·𝕃²·𝕋⁻¹] × [𝕃·𝕋⁻¹] × [∅]) = [𝕄⁻¹·𝕃³·𝕋⁻²] × [𝕄²] ÷ ([𝕄·𝕃²·𝕋⁻¹] × [𝕃·𝕋⁻¹] × [∅]) = [𝕄·𝕃³·𝕋⁻²] ÷ ([𝕄·𝕃²·𝕋⁻¹] × [𝕃·𝕋⁻¹] × [∅]) = [𝕄·𝕃³·𝕋⁻²] ÷ ([𝕄·𝕃³·𝕋⁻²] × [∅]) = [𝕄·𝕃³·𝕋⁻²] ÷ [𝕄·𝕃³·𝕋⁻²] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

The bits a well holds, from the area of its horizon

Where:

  • I_w [∅] — the information of the well, in bits
  • G [𝕃³·𝕄⁻¹·𝕋⁻²] — the gravitational constant
  • M [𝕄] — the well's mass
  • ħ [𝕄·𝕃²·𝕋⁻¹] — the reduced Planck constant
  • c [𝕃·𝕋⁻¹] — light speed

➢ About 1.5 × 10⁷⁷ bits for a well of one Sun, 1.5 × 10⁷⁹ for a stellar well of ten. It is also the count a well gives back by its glow once it has evaporated (Part 3.17): a well returns exactly its horizon's bits, and its child's Pulsecore is born holding them.

A child's level is (log₂ I − 4)/2 (Part 2.10), so the level it is born with is set by its well's mass against one mass, the mass of a well whose child is exactly as deep as its parent:

Ascendant Mass G

M≡ = √(ln(2) / (4 · π)) · 2^(L + 2) · m_P

Dimensional analysis:
Where (from the book): M≡ [𝕄], L⊛ [∅], m_P [𝕄]
[𝕄]
√([∅] ÷ ([∅] × [∅])) × [∅]^([∅] + [∅]) × [𝕄] = √([∅] ÷ [∅]) × [∅]^([∅] + [∅]) × [𝕄] = √[∅] × [∅]^([∅] + [∅]) × [𝕄] = [∅] × [∅]^([∅] + [∅]) × [𝕄] = [∅] × [∅]^([∅]) × [𝕄] = [∅] × [∅] × [𝕄] = [∅] × [𝕄] = [𝕄]
[𝕄] = [𝕄] ✓ Valid PulseCore Verified ✓

The well whose child is as deep as its parent

Where:

  • M≡ [𝕄] — the Ascendant Mass
  • L [∅] — the parent universe's level
  • m_P [𝕄] — the Planck mass

➢ In the universe's own Suns, 0.54·m_P·(m_P/m_p)² each with m_P/m_p = 2^(L/3)/14.3 (Parts 2.10 and 6.10), M≡ ≈ 356 × 2^(L/3) Suns: about 356 Suns for every well its universe forms, so always about an eighth of the universe's whole mass, Data Matter with it, at every level. At ours, about 6 × 10²² Suns, some 14 % of the matter of the observable universe.

Child Level G

L_c = L + log₂(M / M≡)

Dimensional analysis:
Where (from the book): L⊛_c [∅], L⊛ [∅], M [𝕄], M≡ [𝕄]
[∅]
[∅] + log₂([𝕄] ÷ [𝕄]) = [∅] + log₂([∅]) = [∅] + [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

Each doubling of a well's mass makes its child one level deeper

Where:

  • L_c [∅] — the child's level
  • L [∅] — the parent's
  • M [𝕄] — the mass of the well that seeds it
  • M≡ [𝕄] — the Ascendant Mass

➢ A well at the Ascendant Mass gives a child as deep as its parent; each doubling above it, one level deeper; each halving below it, one level shallower. At our level a stellar well of ten Suns gives a child at about level 129, 73 levels down, full of stars all the same; at level 25 the same well gives 11.5, below stable matter.

Ascendant Wells. A well at the Ascendant Mass or above is an Ascendant Well: its child is as deep as its parent or deeper, and it alone carries a lineage on undiminished. No well grows so large in a universe's life: a well supplied by a disk of gas stops growing near 5 × 10¹⁰ Suns, where its disk can no longer form outside its horizon (King, 2016), and the heaviest measured, in the Cosmic Horseshoe, holds about 3.6 × 10¹⁰ (Melo-Carneiro et al., 2025), about a two-trillionth of our Ascendant Mass; its child would be about 41 levels shallower than ours. Ascendant Wells grow in the decay, when the wells take in all that is left (Part 3.10). So a universe ends with a few Ascendant Wells and countless ordinary ones: the ordinary wells seed smaller, shallower universes, many of them starless, whose lines soon end, and the Ascendant Wells carry the unisphere on. This is the selection by which the unisphere grows from the first universe deep enough for prolonged progeny.

How many can a universe have? Each holds at least the Ascendant Mass, so together they can hold no more than the universe itself:

Ascendant Ceiling G

N≡ ≤ M_U / M≡

Dimensional analysis:
Where (from the book): N≡ [∅], M_U [𝕄], M≡ [𝕄]
[∅] ≤ [𝕄] ÷ [𝕄] = [∅] ≤ [∅] ✓ Valid PulseCore Verified ✓

A universe holds at most eight Ascendant Wells

Where:

  • N≡ [∅] — the Ascendant Wells a universe forms
  • M_U [𝕄] — its whole mass, matter and Data Matter
  • M≡ [𝕄] — the Ascendant Mass

➢ From the book's laws the universe's matter is about 3,500 · (m_P/m_p)⁴ proton masses, its stars holding 6 % of it (Part 2.10), and its Data Matter 27/5 times that (Part 3.13), so M_U ≈ 2.23 × 10⁴ · (m_P/m_p)⁴ · m_p; the Ascendant Mass is √(ln 2 / 4π) · 2^(L + 2) · m_P ≈ 2.75 × 10³ · (m_P/m_p)⁴ · m_p. Their ratio is 8.1 at every level: at most eight Ascendant Wells, and eight only if the whole universe ended in eight equal wells. Counting the matter alone, without its Data Matter, it is 1.3: at most one. At ours M_U ≈ 4.6 × 10²³ Suns, the measured matter of the observable universe.

Read against the universe it is in, a well's child stands by the share of everything the well holds:

Level from Share G

L_c = L + log₂((M_U / M≡) · s)

Dimensional analysis:
Where (from the book): L_c [∅], L [∅], M_U [𝕄], M≡ [𝕄], s [∅]
[∅]
[∅] + log₂(([𝕄] ÷ [𝕄]) × [∅]) = [∅] + log₂([∅] × [∅]) = [∅] + log₂([∅]) = [∅] + [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

A child's level from its well's share of its universe

Where:

  • L_c [∅] — the child's level
  • L [∅] — its parent's
  • M_U [𝕄] — the parent universe's whole mass, matter and Data Matter
  • M≡ [𝕄] — its Ascendant Mass
  • s [∅] — the share of that whole mass the seeding well holds, M / M_U

➢ With M_U / M≡ = 8.1 at every level (Ascendant Ceiling), L_c = L + log₂(8.1 · s): a well holding an eighth of everything seeds a child at its own level, one holding 3 % a child about two levels down, one holding a thousandth a child seven down. In our sky the heaviest well measured, about 3.6 × 10¹⁰ Suns against our 4.6 × 10²³, holds s ≈ 8 × 10⁻¹⁴, a child about 40 levels below ours. The game reads s for every well. In the 20 universes it has played to the end of their outward expansion, at levels 20 to 28, the heaviest held 1.8 % to 7.4 % of everything (half of them under 3.3 %), so even the heir began 0.7 to 2.8 levels down and none was Ascendant; in the one universe read again after its decay, its heaviest held 36.5 %, an Ascendant Well whose child stands 1.6 levels up, as the decay should make them.

Ascendant Share G

f≡ = N≡ / 2^(L / 3)

Dimensional analysis:
Where (from the book): f≡ [∅], N≡ [∅], L [∅]
[∅]
[∅] ÷ [∅]^([∅] ÷ [∅]) = [∅] ÷ [∅]^([∅]) = [∅] ÷ [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

The share of a universe's wells that are Ascendant

Where:

  • f≡ [∅] — the Ascendant Wells among all the wells it forms
  • N≡ [∅] — its Ascendant Wells, eight at most
  • L [∅] — its level

➢ At most 8 · 2^(−L/3): one well in 40 at level 25, one in about 5 × 10⁴ at level 56, and at ours one in about 2 × 10¹⁹, against some 1.9 × 10²⁰ wells. A deep universe makes countless children and only a handful of heirs.

Level and Generation

A universe's generation counts its bifurcations: how many stand between it and the first, the Zinf. Its level is not that count. Its space, its time and its vacuum are set by its data: the physical layer is built by the underlying data (Part 3.6), and what differs between a rich seed and a poor one is not how many bits it grows to, since every Pulsecore's count grows by the pulse to its capacity, but how much information they hold. So the level L in every physical law is read from that information: its places, its life of 2^L Tempos, its Level Λ Law, its matter, its stars and its wells. The level is read from a universe's physics; its generation counts the bifurcations behind it, which nothing inside it measures (Reading a Lineage). The two are independent: a universe of the eightieth generation can stand at level 25, and ours, at level 202, may be of the four-thousandth. A child of a small well is a whole universe on a smaller scale, its space, vacuum and matter scaled together.

Every child is one generation on, whatever its level:

Generation Step G

g_c = g + 1

Dimensional analysis:
Where (from the book): g_c [∅], g [∅]
[∅]
[∅] + [∅] = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

Every child is one generation beyond its parent

Where:

  • g_c [∅] — the child's generation
  • g [∅] — its parent's, counted in bifurcations from the Zinf

➢ The step never varies; the level does. A child's level is set by its well (Level from Mass), so a lineage counts up one generation at a time while its level rises, holds or falls.

With level written so, the Child Level law loses its parent: a well's mass alone sets its child's level, counted in doublings of one mass, the mass whose horizon holds the sixteen bits of a Zinf-level Pulsecore (χ∂ = 16 at level 0):

Seed Mass G

M₀ = √(16 · ln(2) / (4 · π)) · m_P

Dimensional analysis:
Where (from the book): M₀ [𝕄], m_P [𝕄]
[𝕄]
√([∅] × [∅] ÷ ([∅] × [∅])) × [𝕄] = √([∅] ÷ ([∅] × [∅])) × [𝕄] = √([∅] ÷ [∅]) × [𝕄] = √[∅] × [𝕄] = [∅] × [𝕄] = [𝕄]
[𝕄] = [𝕄] ✓ Valid PulseCore Verified ✓

The well whose horizon holds a Zinf-level seed

Where:

  • M₀ [𝕄] — the seed mass
  • m_P [𝕄] — the Planck mass of the universe the well is in

➢ M₀ ≈ 0.94 m_P, about 2 × 10⁻⁸ kg: the smallest well physics allows, a Null Well of about one Planck mass, holds 4π/ln 2 ≈ 18 bits, a Zinf-level seed and two bits more.

Level from Mass G

L_c = log₂(M / M₀)

Dimensional analysis:
Where (from the book): L⊛_c [∅], M [𝕄], M₀ [𝕄]
[∅]
log₂([𝕄] ÷ [𝕄]) = log₂([∅]) = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

A child's level is its well's mass in doublings of the seed mass

Where:

  • L_c [∅] — the child's level
  • M [𝕄] — the mass of the well that seeds it
  • M₀ [𝕄] — the seed mass

➢ The same law as Child Level, since M≡ = M₀ · 2^L; the parent drops out. Read forward it gives a child's level from the well that makes it; read backward, Route 3 of Part 1.16 asks it for our own level, and waits on the mass of the well that made us.

The ladder of wells. By Level from Mass, each well's child takes its place on one ladder, the same in every universe in its own Planck units:

  • About one Planck mass, level 0: a Zinf-level universe, the minimum viable universe. The smallest well physics allows seeds exactly the smallest universe.
  • About seven Planck masses, level 3: the First Fork, the first universe that can hold two wells.
  • About 0.08 grams, level 12: below it no stable matter, a universe of light and wells (Part 2.10).
  • About a tenth of a kilogram, level 22: below it no stars, so no stellar wells: its only wells are those its matter makes by collapsing whole (the Star Existence Condition, Part 2.10).
  • Ten Suns, level about 129: a stellar well's child, full of stars.
  • About 6 × 10²² Suns, level 202: an Ascendant Well of ours, a child at our own level.

So a tenth of a kilogram divides the wells whose lines can go on from those that end. Every Null Well yet seen is a few Suns or more and seeds a universe full of stars; the wells below the line could only be primordial, from a universe's first instants, and none survive, since a well's evaporation time grows with the cube of its mass (Hawking, 1975): a tenth of a kilogram lasts about 10⁻¹⁹ seconds, and every primordial well under about 5 × 10¹¹ kilograms has already gone home in our 13.8 billion years (Carr et al., 2010). The dead ends of our universe flashed and returned in its first moments.

The Zinf remains the smallest unit. The smallest well carries just two bits beyond the Zinf level's sixteen: too little for anything beyond a Zinf-level universe, so nothing smaller than the Zinf can be seeded, and the Zinf remains the smallest unit. And because the smallest well in every universe, at every level, holds the same seed in its own Planck units, every Zinf-level universe carries the same base DNA, as the theory holds; what varies is the outcome, and how large the wells grow. Two cautions belong with it. The two laws that meet here are both written in Planck units, so they were bound to meet near the Planck mass; what was not bound is how closely, the book's sixteen against the horizon's 4π/ln 2 ≈ 18, the same small factor the Level Λ Law leaves to the base multiplier and the conversion from nats to bits. And Hawking's law is semiclassical and strains at the Planck mass itself, where whether a Null Well can stand, or a relic is left when one evaporates, is still open (MacGibbon, 1987).

Climbing back. A child of a well below the Ascendant Mass is smaller than its parent, but its lineage is not bound there: each generation its own Ascendant Wells can raise the next by up to log₂ 8.1 ≈ 3 levels, so a line begun from a small well can, given enough generations, grow past the universe it came from. From a stellar well of ours, 73 levels down, that is some 25 generations of universes ending almost wholly in one well each. Whether wells can also take in the Null Vacuum as they grow in the decay, which would lift the Ascendant Ceiling to about twenty and the climb to about four and a half levels a generation, is still open: a true cosmological constant supplies no Null Well, since the growth goes as ρ + p and that vanishes when w = −1 (Babichev, Dokuchaev & Eroshenko, 2004), but the Null Vacuum changes as Data Matter returns (Vacuum Return, Part 3.13).

Reading a Lineage

A universe's level is read from its own physics: ours is 202 because our life to the End of the Outward Expansion is 2²⁰² Planck times (Part 1.D). Its generation, how many bifurcations stand between it and the Zinf, is another number, and nothing we measure gives it. The laws bound it from below, since a line climbs by at most log₂ 8.1 ≈ 3 levels a generation:

Minimum Generation G

g ≥ L / log₂(M_U / M≡)

A universe at level L stands at least this many generations from the Zinf

Where:

  • g [∅] — its generation
  • L [∅] — its level
  • M_U / M≡ [∅] — the Ascendant Ceiling, 8.1 at every level

➢ Each generation climbs at most log₂ 8.1 ≈ 3.02 levels, so ours, at level 202, is at least the 67th generation, and a level 28 universe at least the 10th. It is a floor, not a count: a line that sinks through ordinary wells and climbs again stands many generations further out, so ours may be of the thousandth generation or beyond, and nothing we measure tells which.

And a line's level is the sum of its steps, each set by the share its seeding well held (Level from Share):

Level Walk G

L_g = L₀ + Σ log₂((M_U / M≡) · sᵢ)

A lineage's level is the sum of its steps

Where:

  • L_g [∅] — the level after g generations
  • L₀ [∅] — the level the line began at, 0 at the Zinf
  • sᵢ [∅] — the share the well that seeded generation i held of its universe, the sum running over i = 1 … g

➢ A line of heirs holding a few per cent each sinks about two levels a generation; only Ascendant Wells, an eighth of everything or more, carry it up, and by at most three levels each. From the Zinf to our 202, then, at least 67 of our ancestors were seeded by Ascendant Wells.

What the laws do give exactly is a reading of our parent. The well that seeded us held 2²⁰² seed masses, about 6 × 10⁶⁰ Planck masses of our parent, and since a child is at most log₂ 8.1 deeper than the universe it came from:

Parent Level Bound G

Lₚ ≥ L − log₂(M_U / M≡)

Dimensional analysis:
Where (from the book): L⊛_p [∅], L⊛ [∅], M_U [𝕄], M≡ [𝕄]
[∅] ≥ [∅] - log₂([𝕄] ÷ [𝕄]) = [∅] ≥ [∅] - log₂([∅]) = [∅] ≥ [∅] - [∅] = [∅] ≥ [∅] ✓ Valid PulseCore Verified ✓

A universe's parent was at most about three levels shallower

Where:

  • Lₚ [∅] — the parent's level
  • L [∅] — the child's
  • M_U [𝕄] — the parent universe's whole mass
  • M≡ [𝕄] — its Ascendant Mass

➢ With M_U / M≡ = 8.1 (Ascendant Ceiling), our parent stood at level 199 or higher. Level alone sets no upper bound: our seed held all of a parent at 199, half of one at 200, an eighth of one at 202, under 2 % of one at 205 and about 0.05 % at 210.

The seed was a merger, and its masses were unlikely to be equal. The book's description of our origin took two wells of equal mass, spinning near the Thorne limit, as its default (Part 6.9); equal masses were never measured. Merging Null Wells come in a spread: the gravitational-wave population fits the mass ratio as about q^1.1, leaning to equal but with a long tail (Abbott et al., 2023), as in GW190412, about one to four (Abbott et al., 2020a), and GW190814, about one to nine (Abbott et al., 2020b); galaxies' central wells grow mostly by minor mergers. A seed holding as large a share as ours most likely grew as a dominant well absorbing smaller ones. A merger adds levels by the mass it joins, less what it gives the space layer:

Merger Level Gain G

ΔL = log₂((1 + q) · (1 − ε_GW))

Dimensional analysis:
Where (from the book): ΔL⊛ [∅], q [∅], ε_GW [∅]
[∅]
log₂(([∅] + [∅]) × ([∅] - [∅])) = log₂([∅] × ([∅] - [∅])) = log₂([∅] × [∅]) = log₂([∅]) = [∅]
[∅] = [∅] ✓ Valid PulseCore Verified ✓

How many levels a merger adds to its well's child

Where:

  • ΔL [∅] — the levels the merger adds
  • q [∅] — the mass ratio, lighter over heavier
  • ε_GW [∅] — the share of the pair's mass radiated, by the numerical-relativity fit (Gravitational-Wave Power)

➢ About 0.93 levels for equal wells, 0.53 at one to two, 0.29 at one to four, 0.13 at one to ten. Equal mergers climb fastest; a line of lopsided ones needs more generations, and a seed from one was nearly Ascendant before its merger.

The heritage ratio as a record. Every merger in a lineage multiplies its compression into the heritage, f^L = f₀^L × ∏⟪C⟫ (Part 6.8), so (f / f₀)^202 would be the product of every compression in our ancestry, and with each merger's ⟪C⟫ from its spins it would count the mergers behind us. It does so through the clock: written by level and lineage, ①⥂⧗♁ = ℨ · 2^L · H (Part 2.8), our level gives the doubling and our measured Planck time gives the Zinf time times our lineage's heritage, H = f₀^g · ∏⟪C⟫. The book's value, (1.018)²⁰² ≈ 36.7, is the reading of one generation a level with one merger, a calibration, and the Zinf time 2.28 × 10⁻¹⁰⁶ s follows from it; so H, and with it the mergers behind us, is not yet read. The game reads it: every played child records its parent's clock and its own, each generation's step h, the heritage of whole lines, and f₀ from lines that never merged. An estimate of H for lines that reach our level gives the Zinf time, and with it our own H; with f₀ measured, our H then bounds how many generations and how many mergers stand behind us.

The game as an instrument. A universe's whole history cannot be seen from inside it, but the laws can be run forward many times and their results compared with ours. The game records, for every merger, both masses and their ratio, both spins and the angle between them, each well's rank and share, what it gave the space layer and when in the life it happened; for every child, the well or merger it came from and everything it is born with, its level, generation, spin, heritage, compression and Planck time; and when a child is played, its own record links back to that inheritance. Each play adds a pair, the signature in and the universe out. Gathered over many universes, the pairs map how mass ratio, spins and angle become a child's constants, and asking which signatures give constants like ours reads our parent's merger as a measured range, not an assumption. The pairs are pieces; each universe played adds one.

The Galaxy in Binary Pulse Theory

A galaxy is a well's basin gathering stars. Its Data Matter, laid on the scaffold before any matter (Part 3.13), is its dark halo; the Proto Well at the halo's densest place is its site, where its central well is born, from its first gas collapsing whole or from its first dying star; its spin is the spin its clump gathered, carried down the lineage from the Pulsecore's own (Parts 3.12 and Z.11); its stars are the gas that cooled and kept that spin; and its type is how much of that order survived collisions and feedback. The wells it leaves as its stars die are the 10⁸ to 10⁹ of the count above, and its central well is the one whose child, at the End of the Outward Expansion, inherits the galaxy's data (the Return to Bifurcation).

The Game's Reading

The simulation draws each well at its true horizon, a dark point too small to see from afar, and shows it by the stars gathered round it; a tap anywhere in a galaxy selects its well. Stars are drawn by type: blue-white giants bright and few in young clusters, red giants and supergiants large and orange, red dwarfs small and dim, brown dwarfs fainter still, white dwarfs as small sharp points, neutron stars blinking as pulsars; supernova remnants and planetary nebulae as expanding shells. A central well that accretes fast lights as a quasar, a blazing core with two jets along its spin axis, standing up from the torus's surface, and a weaker one as a Seyfert glow. A selected galaxy is named by its type, read from its stars' rotation fraction and their colour. A galaxy in the game holds tens of bodies, not the hundred thousand arms need. Sharpening their pull on each other does not help: a system of N bodies is stirred by close passes within about N / (8 ln N) crossings (the two-body relaxation time; Binney & Tremaine, 2008), about one crossing for thirty-five, so measured, letting the bodies pull each other directly up close heated the disks into swarms (disk-dominated galaxies fell from 104 to 73 of 274). That is why simulations soften gravity at low particle counts, and the grid's own smoothing does it here. So, as cosmological simulations do, its inner structure is drawn by the known laws from what the simulation measures in it: from the mass round its well, its rotation curve, and from that its shear rate, epicyclic frequency, disk surface density, its number of arms (Arm Number) and their pitch (Arm Pitch Angle); the disk itself turns by the Disk Turn Rate on a flat curve, each star on its own orbit, the arms' pattern with it, in the galaxy's own sense. Its matter does not arrive from nowhere: galaxies grow by cold streams falling in along the filaments of the cosmic web, which run between neighboring halos (Kereš et al., 2005; Dekel et al., 2009). So each of the largest galaxies draws two or three streams of stars, aimed along the filaments toward its nearest neighboring galaxies. Their stars fall inward from about halfway to each neighbour, curl into the galaxy's turning as they near it (keeping their angular momentum, they turn faster as they fall), and join the disk at its edge, while new stars start far out on the filament. Its stars are its own population, by type; the young blue ones lie on the arms, where they form, and the old ones in a smooth disk and bulge. So a spiral is an exponential disk with as many trailing logarithmic arms as its disk can raise, round a warm bulge; a lenticular as a smooth disk with a larger bulge; an elliptical as a round warm swarm; an irregular or a starburst as clumps.

The game's reading of the wells. Every well is counted and kept, the stellar wells, the central wells born at sites and those from collapsing clumps, and every Proto Well is kept as a site, and numbered by rank, heaviest first: Well #1, the universe's Alpha Well, marked through its whole life. At the End of the Outward Expansion a census reads each well's share of the universe, its mass taken with the matter it will draw in, and its child's level by Child Level; when the full decay has taken the matter in, with the Data Matter alongside, the census is read again and every child takes the level its well then holds. A child plays at that level, one generation on (Level and Generation). Ascendant Wells are named on screen and Well #1 is marked with a gold hoop; a well is drawn only when it can be seen, accreting, at a galaxy's centre, Ascendant or first, and the rest keep their place and their lives without being drawn. Each well's life, its kind, rank, peak mass, share and child's level, and the census, go on the universe's record. In a universe of level 25 the heaviest well holds one or two hundredths of everything, so its child lies just under the star line: the children of the shallowest starry universes are starless, and a lineage keeps its stars only from about level 26 deeper. A site pulls and gathers its galaxy but is no well: it is not counted, drawn, ranked or shining, and it takes in nothing. At each reading of its galaxy it tests the conditions of direct collapse (deep enough, its gas pristine, a shining neighbour, and no stars of its own yet), and one that passes makes its central well, M_dc, never more than a tenth of its gas; otherwise its first stellar well born within its reach sinks to it (drawn at once) and the two merge into its central well, or a clump collapsing whole within it makes the central well, held to the Direct Collapse Seed. A merger that carries any well past an eighth of its universe is announced as an Ascendant Well, and its tag floats with it from then on. A central well carries its own mass, one well's, beside the drawn mass that moves it, as simulations carry their Null Wells (Springel, Di Matteo & Hernquist, 2005): in a deep universe, where a drawn star stands for billions, it is born at a sliver of its galaxy and blazes as a quasar; at level 28, where a galaxy holds about a hundred stars, a single remnant outweighs what it is supplied and no quasar lights, as so small a universe would have none. A central well grows by the gas it takes in, solved exactly between readings, at the limit and then with the supply; a drawn body it takes in joins its galaxy's centre and not its own mass, since deep a drawn body is a cluster of billions of stars. Every well born is tested for a site's core, and the Site Lift and the share of sites holding their well over the life go on the universe's record with each site's conditions and how and when its well was born. Measured in one universe at each depth: at level 28 about one site in nine holds its well by the end, a few by direct collapse and most from their first dying star, its central wells of 10 to 20 Suns flicker as Seyferts and none is a quasar; at level 133 the central wells reach a few ten-thousandths to a few thousandths of their galaxies, as in our sky, and a quasar era comes of itself, up to fifteen at once between a tenth and a fifth of the life while the central wells are born and catch up, fading by half the life, more of them at once than our sky's few in a hundred; and at both, wells are born inside sites' cores about 1.4 times as often as chance.

No band preferred. The torus has no favoured band, as the cosmological principle has no favoured place: the first ripples and the Data Matter keep every local lump but no row-wide contrast, and the Data Nova lays its data to the pinch's own row, the hole being closed. Before, the data's empty ring at the pinch and its banded share pulled the matter into one band, more than half of it into a fifth of the surface.

Testable Predictions

  • Wells where the data said: wells, and the galaxies round them, should form where the Proto Wells stood, far more often than chance (Site Lift), and a galaxy's central well should be born either early, heavy and metal-free, or later from its first dying star.
  • The wells unseen: one well for every 65 stars, so 1.5 to 6 billion in our galaxy; the Roman Space Telescope's microlensing survey should find isolated Null Wells at 15 to 60 times the rate today's models expect.
  • Wells per galaxy: the count of stellar Null Wells in a galaxy like ours, measured by gravitational microlensing and gravitational waves, should come to 10⁸ to 10⁹, as 2^(L/3) divided by the number of galaxies gives.
  • Spin inherited: the spins of galaxies across the sky should carry a faint common alignment along the pinch's axis, the same preferred axis predicted for the microwave background (Part 3.6).
  • Disks and Data Matter together: a galaxy's disk size should follow its Data Matter halo's spin, R⊘ = λ⟲ R⊡ / √2, as standard theory has it, since Data Matter is the dark matter.

Links to Binary Pulse Theory

  • Part 2.9: the stars, their ends, and the Stellar Null Wells a galaxy fills with.
  • Part 3.10: the progeny law that counts the wells; the central well's child inherits the galaxy's data.
  • Part 3.12: the tidal spin every galaxy keeps, and the cooling that turns it into a disk.
  • Part 3.13: Data Matter, the dark halo every galaxy sits in; the Proto Well that marks its site.