Feeds:
Posts
If $N$ is a four-manifold and $M$ is a hypersurface representing the physical universe, a metric on $M$ can represent gravity and a 1-form $A$ on $N$ could represent electromagnetism.  Rather that take the approach where components of a metric on $N$ contain data for electromagnetism, I think it is better to consider the Ricci curvature equation for $M$ to be the gravitational field equations and then seen the relation between gravity and electromagnetism equating second fundamental form terms with the field strength components of $F=dA$.
In principle, the geometry of $N$ could be arbitrary for the issue of coherent unification of electromagnetism and gravity, i.e. perhaps for many different possible geometries of $N$, locally one could produce a metric and a 1-form that are related by the Ricci curvature equation for submanifolds on $M$.  The evidence that $N$ has a specific geometry — that of a scaled 4-sphere — could be thought to be independent of this unification scheme.  Then we can say that $N = S^4(1/h)$ falls in a class of manifolds where the unification scheme works.