Can you put a magnetic ball into a hollow magnetic sphere? if all magnets have to have two poles(one north one south), is it possible to construct a hollow sphere where the inside face of the sphere was one pole, and the outside face another pole?
is it also possible to make a magnetic sphere (not hollow) where the outside face is one pole?
If both of these ideas are possible, what would happen if we put a small sphere (non hollow) inside of a hollow sphere, where the polarity of the hollower sphere's inside face was the same as the polarity of the smaller ball? the ball would levitate theoretically, but what if the magnets were so strong that they pushed the ball more and more into the center until it had nowhere else to move?
 A: The answer is that you can build such a sphere.
Assuming the sphere is built perfectly, then the magnetic field will cancel everywhere. 
A: Conventional magnetic fields are observed to be loops. Where a magnetic field exists, it is impossible to carve out a volume of space such that more magnetic lines enter that region, than leave.  Although a bar magnet has identifiable north and south poles, the imaginary lines of the field actually continue inside the magnet, forming complete circuits.
If you had a sphere which was all north all over, there would be no way for the lines emanating from that pole to meet with the interior south pole. That sphere would effectively be a magnetic monopole, as far as what we can tell from its exterior. If we contain that sphere in a region of space, that region of space has magnetic field lines emanating all over, without any such lines re-entering it.
http://en.wikipedia.org/wiki/Magnetic_monopole
If the search for magnetic monopoles ever turns up something, then it will confirm that such a sphere can exist.
http://en.wikipedia.org/wiki/Magnetic_monopole#Searches_for_magnetic_monopoles
Now of course could have a sphere such that most of its surface is a magnetic pole, like north, but such that it has some small region (or regions) where the magnetic lines emanating from this north re-enter into the sphere.  These regions would then be de-facto south poles, though, making this different from a sphere which is north all over.  If you make an object by tiling a sphere with magnets, I suspect you will get this kind of animal. The inner field lines will "escape" somewhere.
A: For the most part, See Above. 
All I have to add is that from a simple geometric perspective, you would be forced to leave gaps in between magnets on the outside of the sphere. Magnetic poles can be thought of as straight lines, and so the number of ends on the inside of your sphere would have to equal the number of ends on the outside. The outside of a hollow sphere has more surface area of course, and so you would always have gaps in between. These would be the de-facto south poles, as noted above. 
A: Possibly by using a nonmagnetic sphere as your center and pressing a neodynium cobalt alloy around it sealing the outside sphere shut. I have thought about it for a bit but have no opportunity to attempt such thing. 
A: I believe if you start with a flat magnet then press it into a semi sphere, I say semi because you want a hole in it, the hole you will role in on itself to form an inner lip creating a third semi sphere effect.  As long as you have that hole I figure it will weaken the inner pole enough to give you a mono pole effect.  The effect would be sufficient enough to make a mono magnet wheel to spin a generator. 
