# Distinguish a magnet from a non-magnet

Given two objects, one that is a permanent magnet and a "non-magnet" that is simply magnetizable, how would we determined which is which? Can this be done just by observing the motion they induce in one another, i.e. without using any external fields, magnets, or magnetizable objects?

The key distinction between the two objects is that the poles of the permanent magnet are fixed, but for the non-magnet the locations of its poles depends on the orientation of the two objects. However I can't seem to turn this fact into a procedure for identifying the magnet.

The question comes from the first section on the chapter on magnetism from a college physics text, so the answer should be expressible in very basic terms.

I have a spherical neodymium magnet and a Mond process ball that I stick to my fridge. The way I usually have people find out which is which is by dropping the down a piece of copper pipe. I guess that doesn't count because the pipe is being magnetized by Faraday's law as the magnet drops. But a test I just tried is to stick them together and try turning one of them. The Mond process ball, having no preferred orientation, is free to turn, but the permanent magnet wants one of its poles to be in contact with its partner, so when you try to turn it, it snaps back into position.

• I like the second solution! Do you think it generalizes to arbitrary geometries, not just spheres? – creillyucla Mar 26 at 3:20
• Spheres are great because a uniformly magnetized sphere has exactly a dipole field. I had hoped that I could sense the change in force as I rotated the magnet between my fingers because it should be something like $4\times$ as great when the pole is pointing at the other object as when it is pointing at right angles to the object. Hmm... doing it again just now I can tell the difference, but it's kinda subtle. An asymmetric object has a preferred orientation in a magnetic field. If you need instrumentation to tell them apart, you might just as well hang them by thread and see which is compass. – user5713492 Mar 26 at 3:50
• I do not understand, nickel is ferromagnetic anyway. "Nickel is one of four elements (the others are iron, cobalt, and gadolinium)[7] that are ferromagnetic at approximately room temperature." en.wikipedia.org/wiki/Nickel – anna v Mar 26 at 4:49
• @user5713492 that would be using an external field though! – creillyucla Mar 26 at 4:58
• @annav Yeah, I know and have all $4$ elements. But the rules specified "magnetizable" so I think that includes non-permanent ferromagnets, not forcing you to go to dysprosium (which I have) or terbium (I wish it weren't so expensive) for the second object. Take a piece of gadolinium out of the freezer and touch a Mond process ball to it and there is no force detected. – user5713492 Mar 26 at 5:33

Assume you had two bars, one a magnet, the other a magnetic material. If you also had superhuman strength, you could twist them individually into a horseshoe shape. The ends of the magnet would then attract each other, the magnetic material would feel no such force.

An alternative, though I haven't tried it out, could be as follows. Hold one bar (say Y) fixed on the y-axis, a little asymmetrically about the origin. Place the other bar (say X) along the x-axis, and hold it fixed. Now, holding the bar X fixed, let Y move. If Y is attracted end-first, Y is the magnet. If it is attracted middle first, then X is the magnet.

• Wouldn't the non-magnet also rotate when released, since doing so would increase its magnetic moment? – creillyucla Mar 26 at 3:09
• I am really not sure about this. The induced pole would be at the closest point at equilibrium, but beyond that, an experiment would help here. – NewUser Mar 26 at 3:14

If you only have these two objects, then take two bars . The permanent magnet one will have field lines . The magnetizable one none, or only acquired ones from the field of the permanent.

The permanent magnet one will stick much harder to the "non permanent" if the two bars are brought together in a T shape at their center, because the true magnet has strong lines at the pole, but the pole induced to the "non permanent" will be much weaker, due to the few lines.

• good suggestion but I think (not certain) that the rules are no other magnetizable objects are allowed – creillyucla Mar 26 at 4:55
• @creillyucla thanks,I edited , I took away the third party suggestion. – anna v Mar 26 at 6:23
• I tried this with a nail and four fridge magnet stuck end-to-end and it seems to agree with your procedure. That is, the nail will "snap" to the poles when moved around the magnet, but the magnet will stick anywhere on the nail. How do I know this is not an accident of the particular geometry of the nail and magnet? The induced dipole moment will be strong for different nail orientations, so how do we know that a geometry geometry will not lead to "snapping" of the magnet to the locations which optimize the induced dipole moment in the non-magnet? – creillyucla Mar 26 at 15:01
• @creillyucla it is the density of field lines , optically, could be done mathematicaly too, weaker fields. Induced quantities cannot be as strong as the original ones supplying the energy. – anna v Mar 27 at 8:40