The resistance to aligning the domains is also what keeps them from spontaneously returning to their original random state when the magnetizing field is removed. This is true in so-called "hard" magnetic materials versus the "soft" materials, which retain very little of the impressed magnetic field.
From this explanation it is seen that the orientation of the poles is contained in the physical alignment of the domains in the original bar magnet. The direction of the poles will therefore be retained in sphere relative to its orientation in the unaltered bar.
The result would be quite different if the magnetic dipoles were composed instead of free monopoles, analogous to electric charges. If a conducting sphere were charged with an excess of electrons, their mutual repulsion would cause them to move and distribute evenly over the surface of the sphere, resulting in no electric "pole" at all.
Answered by: Scott Wilber, President, ComScire - Quantum World Corporation
Some people might be bothered by this description because they have been told that there are no "magnetic monopoles," i.e. there are no teeny little isolated magnetic field sources. This is true, but it turns out that if you take an electric field source (such as an electron) and run it around in a circle (say as if it were orbiting an atom), it then exhibits a magnetic moment, and looks a lot like a little magnet. These are the little magnets that I was referring to.
The next logical question might be, "well, if all it takes is electrons orbiting atoms to make up a little magnet, then why isn't everything made up of atoms magnetic?" The answer is that in a lot of cases, the teeny little magnets cancel each other out. One way they can do this is if you have two electrons orbiting a nucleus, only in opposite directions (i.e. one clockwise, one counterclockwise). Another way that magnetic fields cancel out is if the teeny bar magnets in a material are free to jiggle around in the material, and they become randomly oriented - then on average you have the same number pointed in one direction as in the opposite direction, and you get no overall effect. It's only in some special kinds of materials that you can get all the little bar magnets pointing in the same direction, and then "freeze" them there. These materials are then called magnetic materials, for obvious reasons.
Answered by: Gregory Ogin, Physics Undergraduate Student, UST, St. Paul, MN
'There must be no barriers for freedom of inquiry. There is no place for dogma in science. The scientist is free, and must be free to ask any question, to doubt any assertion, to seek for any evidence, to correct any errors.'
J. Robert Oppenheimer