PUH Theorem 391 — What the Rebound's Handedness Can Paint on the Microwave Sky: an Axis Alone Gives Only Odd Multipoles, So the Uniform Twist Needs a Pseudoscalar, and Today's Turning Is 10⁶ Too Slow

Photonic Universe Hypothesis (PUH) — Theorem paper. THE QUESTION. A uniform twist of the microwave background's polarization, about a third of a degree, has been reported since 2020, and a new test (Lonappan, Keating and Arnold 2026) finds the detector-angle pattern behind Planck's value consistent (chi-squared 8.18 for 7 degrees of freedom, probability to exceed 0.317). T354 had said that a rebound axis is a dipole and a global handedness the monopole, without deriving either. What can the rebound's handedness paint on the sky, and from which of its objects? RESULT 391.1 — THE MIRROR RULE. A twist reverses under parity. If light's own laws are mirror-symmetric, a pattern made by axes alone (axial vectors, like the rebound's spin) has only odd multipoles: dipole, octupole and so on, and no uniform twist. Every even multipole, the uniform twist included, needs a pseudoscalar: a handedness that points nowhere, or a direction of travel joined to an axis as a screw. Checked numerically to rounding error; Carroll, Field and Jackiw's background vector is the standard instance (timelike part: uniform twist; spacelike part: dipole). RESULT 391.2 — THE REBOUND'S THREE HANDED OBJECTS. The rebound's spin axis (T136, T216) is axial and paints odd multipoles only. The chirality the rebound selected (T212, T352) is a pseudoscalar, the same in every direction, as beta decay shows to experimental precision: it paints the uniform twist. Our jet's travel along the axis (V8) makes, with the spin, a screw: a pseudoscalar of the archive's own, opposite in the two jets and absent in the equatorial throat. T354's monopole belongs to the chirality or the screw, not to the axis, which gives none; with all three on one axis and no other direction entering, the twist is symmetric about it. RESULT 391.3 — BY TURNING ALONE, TODAY'S SPIN CANNOT PAINT A MEASURABLE TWIST. The background's polarization is made at last scattering (z = 1090; 13.864 Gpc comoving; 13.790 Gyr and 4.228 Gpc of light path). With the vorticity falling as a^-2 and the CMB's limit (omega/H)_0 < 5.2e-11 (Saadeh et al. 2016), a rotation turns a polarization plane by at most 5.62e-9 rad (3.22e-7 deg), 9.93e5 times less than the reported dipole; T136's single quantum, 5.45e-141 rad/s as T368 corrected it, gives 2.7e-121 rad. Drag by a turning medium is the fraction 1 - 1/(n n_g) of the medium's own turn: zero in vacuum and in a cold plasma, and within the same bound for n n_g >= 1/2. A measurable twist needs a coupling of the spin or the handedness to light, which the archive does not yet have. RESULT 391.4 — THE MEASUREMENTS. The uniform twist: 0.35 +- 0.14 deg (Minami and Komatsu 2020), 0.30 +- 0.11 (Diego-Palazuelos et al. 2022), 0.342 +0.094 -0.091 (Eskilt and Komatsu 2022), 0.215 +- 0.074 from Diego-Palazuelos and Komatsu's analysis of ACT DR6 with an optics-model calibration, and 0.265 +- 0.058 (4.6 sigma) combined with WMAP and Planck. Its shape is the one a pseudoscalar handedness paints: it fits the chirality or the screw, with no amplitude derived. Accrued evenly it is 2.88e-29 to 4.95e-29 rad per metre of the light's path, and it is consistent with no change across frequency (n = -0.35 +0.48 -0.47; a nu^2 rotation lies 4.9 upper errors above, Faraday's nu^-2 3.5 lower errors below). The full-sky dipole, 0.32 +- 0.10 (stat) +- 0.08 (syst) deg toward (l, b) = (295, 17) (Contreras, Boubel and Scott 2017), marginal and in part dust, lies 49.8 deg from the quadrupole-octupole axis T136 names, within about two of its direction errors (chi-squared 4.67 to 5.80 for two degrees of freedom): the comparison cannot decide yet, OPEN. RESULT 391.5 — T354'S TESTS REVISED. Its proposed test on ACT's map cannot be run: a field rebuilt from L = 15 holds no dipole, and the map covers 0.228 of the sky; the full-sky dipole replaces it. The uniform twist was subtracted in one analysis and measured in another. The axion template still does not apply to a pattern symmetric about one axis. NOT CLAIMED: an amplitude for any multipole; that the archive's handedness couples to light; that no coupling of rotation to light exists in nature, only that the archive has none; that the screw is the selected chirality; that the uniform twist or the dipole is cosmological, or evidence for this framework; the direction of the rebound axis; the sign of any twist; anything about axions. KILL-CONDITIONS: light's propagation through the substrate shown to break mirror symmetry by itself; the selected chirality shown to point along an axis; a cosmological twist shown not to be symmetric about one axis while all the archive's handedness lies on the rebound axis and no other direction enters; a rotation shown to turn polarization, by turning alone, by more than the angle it turns through; the uniform twist shown to be instrumental. OPEN: the coupling of the selected handedness, or of the spin, to light, nearly achromatic and a few times 1e-29 rad per metre of path; whether the chirality is the screw, and where the complementary chirality lies; the axis's direction and the contested galaxy-spin axes; the full-sky dipole, the dust or a mechanism not yet found; the sign of the uniform twist.