Yes, it would have to be set "horizontally", if you will, below the surface. Most models assume the poles are atop one another. I think there would be field problems not seen in the standard model, but it may explain some other phenomena equally well or better than the Cambridge model.