Re: Waht happens when you connect the poles of the same magnet?

James,

This is an entertaining question. Please don't worry, you won't open a 
rift in the fabric of space-time by trying to connect the poles of the 
same magnet. Although a magnetic monopole would be quite interesting, to 
my knowledge the existence of magnetic monopoles has yet to be discovered. 
(Besides I think you would need things on the quantum scale for such 
things.)

So, let's get down to the problem. This is a feasable question so it 
requires a feasable solution. First lets consider the bar magnet as a 
dipole. This is a simple model which adequately describes the bar 
magnet's "bulk" properties. The magnet has a North end and a South end. 
This is sufficient to describe the bar magnet's interaction with other 
magnets. The problem is, "You only have one magnet!!". We either need to 
reject the dipole model, or maybe modify it slightly for our needs. Lets 
choose the later by digging deeper in to the structure of a magnet. The 
bar magnet has a microscopic crystalline structre which consists of many 
fused together "domains". It is the allignment of these magnetic domains 
that gives the magnet its large scale magnetic properties. (try taking an 
old bar magnet and whacking it with a hammer, the process will typically 
ruin a perfectly good magnet) Now back to the dipole model. Let's 
redescribe the bar magnet as a collection of these magnetic domains and 
replace the tiny grains that make up the magnetic with small dipole 
magnets. When all these tiny dipoles allign north to south, the net effect 
is a large bar magnet. (You are probably guessing where we are going by 
now) Now bend this collection of dipoles into circle and connect the ends. 
We have just created a ring (or annulus) of alligned dipoles. The net 
effect of all these little dipoles is a donut (or bagel) shaped magnet 
with magnetic flux lines shooting through the center. North and South are 
labled along the axis running through the center of the hole. (this new 
magnet must be considered in three dimensions!!)

I hope this answer is as helpful. There are some excellent resources on 
the Web for investigating magnets. Try starting here:
 http://www.execpc.com/~
rhoadley/magfield.htm

That page also has images of the field around a bar magnet and a ring 
shaped magnet. 

Also if you are feeling very magnetic there is some !free! software that 
will simulate magnetic fields around objects you design. I tested it out 
and the package works well, only it takes some time getting the hang of 
it. The documentation is well written.
 http://members.aol.com/GMagnetics
/

Good luck in your studies. 

Keith Heberlein
Center for Magnetic Resonance Research
Department of Biomedical Engineering
University of Minnesota Medical School
Minneapolis MN USA