Hawking’s cosmology logically leads to an observed multiverse. This article argues it is a superposition of at least three 3-dimensional universes in a 4-dimensional space, of which two dimensions overlap with our universe. Nothing that could disturb the superposition exists outside it. This explains why dark matter causes a linear decrease in gravity with distance to visible mass at large radii in galaxies. To support this, the visible matter distribution in the disks and bulges, calculated by the SPARC team, and the observed rotation velocities have been used. Lelli and Mistele showed that the common way to project dark matter halos around galaxies cannot be valid. Since General Relativity would need these halos too, it must be modified with additional terms, or an added wire-like mass must be modelled in galaxies with the Levi-Civita metric. Bekenstein and the paper in hand respectively do this. Using TeVeS, the decay of the contribution of dark matter to gravity with the expansion of space is confirmed. This explains the rapid development of large galaxies in the early universe as reported by Labbé. A new prediction method for rotation velocities, that works at all radii in galaxies, is 19 to 27 % more accurate than MOND and TeVeS. In galaxy clusters the improvement of the predicted velocity dispersions is 44 to 57 % over a huge range of cluster masses. It gives a logical explanation of the meaning of Milgrom's contant and the Tully-Fisher relationship does directly follow from the hypothesis.