The model

A monopolar response,
with a memory.

The model keeps the low-acceleration behaviour of MOND but changes what the medium answers to: the spherical field of the enclosed mass, not the local field. That single change removes the phantom disc and the Solar-System effects, and a formation memory decides which systems respond.

In plain words

Newton says a star's pull depends on how much mass sits inside your orbit. This model says the extra pull that galaxies show also depends on the mass inside your orbit, and on nothing else: not on the thin disc of stars you happen to be crossing, not on the star next door. That one choice removes the extra pull from the plane of the Milky Way, where measurements do not want it, and from the Solar System, where Cassini does not see it. A second rule says which objects respond at all: those born from diffuse gas (galaxies) do, those born from a dense cloud (star clusters, the Sun) do not. Unfamiliar words are in the glossary.

The law

What the medium answers to.

Extra acceleration = (ν(gsph/a0)−1) gsph, where gsph is the spherical field of the enclosed mass.

Here gsph = G Mb(<r)/r² is set by the mass a system encloses within radius r, and ν is the McGaugh transition. Because the response is spherical, there is no phantom disc in the plane, nothing around an isolated star, and no Solar-System multipole. The galactic rotation curves are those of MOND, because on the scale of a rotation curve the enclosed-mass field and the local field nearly coincide.

Geometry, not a postulate

Why monopolar is not arbitrary.

A medium made of smooth layers formed when the system collapsed, responding only to the field component normal to those layers, reproduces this monopolar law to within a few percent. The vertical field of the disc is tangent to the layers, so it is ignored. Monopolarity is a geometric projection onto the layers, not an extra ingredient.

What replaced the local models

An earlier quantum-interface Hamiltonian (LQ8) and a family of bounded local polarization models were explored first. They are local laws, and local laws are excluded by the vertical potential. The monopolar response is what survived.

The memory

The Sun, an atom trapped in an inclusion.

The medium is a glassy solid everywhere; in a diffuse galaxy most elements still have slack and respond in MOND. A star forms from a core that collapsed together with the medium bound to it. That medium was strained past its slack during the collapse and never recovers it: a pocket in which every element is engaged and the response is exactly Newtonian. The Sun sits inside such a pocket, like an atom trapped in an inclusion. This is why the Solar System is Newtonian and Cassini and Mercury are untouched. A hot galactic medium cannot be captured by a star at all (capture fraction 10−7); the pocket has to be the cold medium born with the core.

The same rule sorts the star clusters from the galaxies: Pal 14, born dense, is Newtonian; a dwarf galaxy, born diffuse, responds. It is the difference between two systems of the same size and mass whose only difference is their history.

A bracket from the data

The pocket has a size, set by the natal core.

The pocket is the medium that was bound to the natal core: 0.05 parsec for a solar-mass core, up to 0.4 parsec for a 100 M☉ clump. For the whole pocket to be engaged, the medium must lock at the density of a prestellar core, about 104 M☉/pc³ of gas; if it locked only at higher density, the outer, bound-but-slack medium would boost wide binaries by 20–150% at 5,000–20,000 AU, which Gaia DR3 (Banik et al. 2024) excludes. An earlier version of this site predicted a tip back toward MOND beyond 0.15–0.3 pc; no calculation supported it and it is withdrawn. The prediction is: Newtonian at every separation.

Does it fit?

The vertical potential, the rotation curves, and the Solar System, at zero free parameters.

The results