A glob of Neo magnet spheres is a wonderful toy
2004 W. Beaty

The price on 8mm or .25" Neodymium supermagnet spheres (/amateur/neosphere.html) has come down enough that now we can afford to play with huge blobs of them. They behave like giant atoms. ...giant atoms existing in the everyday world! Play with huge atoms while observing the strange behavior of the chemical bonding force, see 3D molecular structures, and spontaneous self-assembly, and various biological phenomena.         

Bead-array Bar Magnet

If a single large supermagnet approaches a color CRT, the colors will distort. But wave a blob of bead magnets near the CRT and nothing happens. They naturally organize themselves into small closed loops with nearly zero field outside the loops. In a blob, the beads act like random domains, and the blob of beads acts like unmagnetized iron.

Make a bar magnet & compass
more billb videos

So let's "magnetize" a blob of sphere-magnets by aligning all the spheres in the same direction. First, make a long chain of magnet beads. Long chains of magnets always align, and they have a strong North end and South end. Dangle the chain vertically, then break it in the center and bring the lower chain up parallel to the top half. Push the two chains together. (The ends, being alike poles, will repel, but the center zips together.) Now repeat: hold the dual chain vertically, break it in the center, then bring the bottom half up and push the two chains together. Fold the four chains lengthwise to form a rod. This rod is a powerful bar-magnet with opposite poles on its ends. Wave it a foot or two away from a video monitor and observe the distortion and colors. (WARNING! KEEP THE MAGNET AT LEAST 10" AWAY FROM YOUR VIDEO SCREEN!) Now smash the rod and scramble it up into a glob again. The external field goes away.

WARNING: if you mess with TWO large neodymium supermagnets, you are bound to pinch your fingers if you let them come together. This REALLY HURTS, it's like pliers. Also, you should wear safety glasses, since when two NIB magnets slam together, they tend to launch tiny chips at high velocities. I've had 3/4" magnets acidentally fall together and be completely shattered. The things are brittle! The ones larger than 1/4" are NOT FOR UNSUPERVISED CHILDREN!

NEW DANGER: swallowed magnets can cause lethal stomach problems. See


Using 60 sphere magnets, make twelve 5-magnet pentagons, and assemble them into a 60-atom Buckyball (but not exactly Carbon Fullerene.) The one shown in this diagram has patterns of pentagons, but tip-to-tip bonding between pentagons is unstable, so hexagon facets are unstable. The pentagons rotate to form side-to-side bonding with squares instead of hexagons. Magnets bond as dipoles like sulfur does, not as quadrupoles like carbon atoms do. Sulfur chains could probably condense from vapor to form this sort of sulfur fullerene "S60 Buckyball."

It's actually a kind of puzzle. Here's the secret: for this to work, all the circular direction of north/south poles arranged around the pentagon-rings must be the same.

[drawing of 60-bead sphere array] More realistic buckyball: Once you're successful in making Buckyballs, try duplicating the way they form in nature. Build two 30-atom hemispheres like so: first form two pentagons, place each one against a circle of ten magnet spheres to form a little cup, then place a circle of 15 spheres on the lip of each cup. This is how buckyballs actually form, and if the two cups come together, they merge seamlessly to form a 60-magnet hollow sphere. AH! THAT'S how Fullerene molecules end up with metal ions trapped inside. It's like a clamshell closing on a stone. (NOTE: this two-hemisphere technique is not trivial! The polarity of the chain-circles are important, an they must be placed so they merge with the magnet spheres adjacent, NOT with the magnets moving into the clefts of the adjacent chains. Also, you must "knit" the lips of the hemispheres together in a zigzag pattern, otherwise you'll get something that looks like a clam. It's a puzzle needing some thought and experiments to solve.)

Buckyball Sphere
more billb videos

Build a "decorated" Fullerene: add 12 extra magnets to the centers of the twelve pentagons. Looks like a cold virus. Build a C-180 Fullerene: form twelve pentagons, but add 10 sphere magnets to the rim of each, then assemble the super-size buckyball. Also note: pentagons can connect to form a "fabric" in square array.

  Travelling Defects

Make a big circle from 25 magnets or more, then squeeze it together to form two chains side by side. Now place the chain-pair on the table and bend the tip. It goes "snap" and forms a bend ...a kink...

2-Dimensional Amoebas
more billb videos

...but on both sides of the kind is the same sort of chain pattern. YOU CAN MOVE THE KINK ALONG. Push it, and the beads click into place as the kink/defect moves through the chain. Form two opposite kinks, then push them around. If they come together, they cancel out and vanish.

Or form many kinks in the same direction, then bring the ends together and make a crude circle with two parallel strands. It's like a tiny amoeba with a crystalline bi-layer cell membrane, with lots of movable kinks allowing changes of shape when you apply pressure to the little creature.


Mysterious Energy Source

On a smooth but not polished surface, throw two sphere magnets together so they connect. They will spin for a very, very long time. A VERY long time. Something weird is going on. Why doesn't friction slow them down? Have you discovered a new source of energy? No, but you've discovered a gravity motor. This is similar to the physics demonstration called "precessing spinning pipe," where the "pipe" very slowly lays down as it spins, and the Potential Energy of gravity *powers* the spinning. Friction apparently doesn't slow it down as it normally would, since there's an energy source keeping it running. A gravity-powered motor! Bring a thick conductive object near the spinning pair (or move an aluminum plate up from below the table) and the spinning halts rapidly. Observe the spinning bead-pair closely and you'll see two glowing circles which slowly come together as the raised bead lays down. The circles are caused by overhead lights; they're from the reflected "highlights" in the two silver spheres. Since one sphere is tilted upwards, one circle hovers above the other. And since the raised sphere slowly settles, the two glowing circles will slowly merge just before the spinning spheres grind to a halt. (I heard about this trick from Mr. George at Eng Concepts magnet store:

Compass, Coupled Rotary Oscillators

Get fourteen magnet spheres. Make two rings of six magnets, then push one extra magnet sphere into the center of each ring, and the center one slightly sticks out, and the chunk of 7 becomes a compass. Lay each "compass" down so the center magnet sticks out the bottom slightly. These "compasses" can easily rotate. Place two "compasses" a few inches apart, and if one is forced to turn, the other turns too. Spin one, and watch it stop while the other spins (coupled oscillators.) Make a chain of three compasses in a row a few inches apart, turn one, and the motion propagates. But if spacing is wrong, the motions won't spread to other compasses (the domain-wall won't shift.) Use black beads to mark one spot on the compass rim...
I heard about this trick from the guys at:


Wind long chains of magnet spheres around your wrists. Use at least 30 in a chain. Hundreds of magnets work even better. Touch the two bracelets together. If the pair is wound in one direction, they merge to form handcuffs, but if one of them is wound in the other direction, they just stick together and peel apart again without disrupting the bracelet. Note that they still exhibit the same effect even if you flip one of them over, (just as screw threads behave.) To change behavior, you have to actually disassemble one bracelet and re-wind it with a reverse screw-wise "chirality." (Or pass it through a matter-transporting mirror, as in "Through the Looking Glass.")

SPIRAL Nanotubes:

Using thirty or more magnets, start a "winding-order seed" by winding a chain of magnet beads on a pencil. Wind a couple of turns, then remove the pencil and keep winding, and the chain will self-assemble into a nanotube. This is much easier than building hexagonal rings and then stacking them up!

Make a very long chain of a hundred or more magnets, then start two "seeds" of slightly different winding order on either end, then wind them up equally. You'll end up with a two-region nanotube with a defect between them in the center. This defect can move! Twist only one end of the nanotube, and the defect clicks along, changing the tube diameter as it goes. The narrow tube can consume the wide one, making the whole assembly longer. Or vice versa. Start three seeds, creating two defects, then twist the center segment of the nanotube and the defects move in the same direction (the center segment moves along.)

If you can figure out a special "bent seed," then successively wound spiral layers also have a bend, and you'll end up with a cone of stacked up rings, with rings having increasing length of 4,5,6, etc. The resulting nanotube becomes a nano-funnel or a Tibetan monk's hat.

Useless Machine

Make a tube by winding a chain of sphere magnets around a pencil. Use about 25 magnets (more if you want a long tube.) Leave the ends long, then hook the ends together to form a loop. If you pull on one end of the loop, the tube unwinds at one end and winds up at the other, while the beads flow constantly through the loop.

"Magnet-eating Amoeba"

Make a chain of three magnets, then bend it to form a triangle. Lay it down and roll a fourth bead to it. It will "eat" the extra bead, becoming a square. Roll a fifth bead and the square eats the extra bead, becoming a pentagon. The pentagon can become a hexagon. But any more magnet spheres than six will just stick to the outside and not merge with the ring unless the ring is put under stress. This is an analogy for surface-energy in curved fluids. Well, it works backwards from the way fluids do! So...

Stick a magnet to the side of a long chain, and it doesn't merge. But put your chain of spheres under tension, and any spheres touching the side will be eaten.

2-State Memory

Make three squares of four magnets each, and stack them up. They won't remain stable, but instead form a pair of hexagons side by side. Try to make them back into squares by squeezing the vertices of the hexagons together, and SNAP!, it becomes two hexagons again, but rotated 90deg.

BIOLOGY (self-assembly)

Make a buckyball from 60 magnet spheres, or just collect a huge bunch of beads into a glob. Touch it with one end of a chain of magnets, then pull, and the buckyball will unravel and become a long chain. As each succeeding "carbon atom" joins the chain, its powerful dangling bonds will immediately grab any neighboring "atom." Long-chain self-assembly tends to happen spontaneously! Ah, THAT'S why human beings appeared from nothing: because chains of atoms are also able to appear from nothing. Supply a cloud of carbon atoms, and the rest is just details.

Plasmid gene transport: a 60-sphere loop of bead-chain can be bent, necked-down, and a tiny ring pulled off like the DNA plasmid from bacterial chromosome. Move the tiny plasmid to another spot on the chain, then merge it by reversing the process. Or make two large rings which represent two separate bacteria, then perform "gene transfer" by moving a ring between them (for effect, make the large rings out of random black and silver beads.) Or form one chain into an "S" shape, touch the lobes of the "S" together, then split the touching parts 90deg differently, and the chain will be whole again, but with a segment in reverse order. Now make a "chromosome" by wrapping the long chain into a tight square-section nanotube. With enough magnets you could wrap this nanotube into a spiral, to form a thick hollow tube, then wrap THAT tube into a spiral to form a bigger tube, etc.


Shouldn't iron atoms repel each other? After all, if you place their alike-poles together side by side, they repel. Why doesn't solid iron spontaneously evaporate? Or why don't the atoms all rotate so Iron is non-magnetic? Let's see. Make some bead-chains, then merge them side by side with their strongly-repelling alike-pole ends together. The ends repel, yet the main body of the chains attract strongly and "zipper" together to form crystalline "permanent magnets." Add more and more chains to the side to build entire single-domain magnetized crystals. See, iron atoms DO stick together to form large magnets. Now make some "compass" clumps (see above,) and use them as field sensors: test at what distance your "single domain" crystal fragment affects them.


Chains of beads can merge with north ends near south ends, where the overall field is zero (it forms rings.) Or, if you just gather a wad of beads together, it will have no large external field, since the beads have a strong tendency to form internal rings. Use rotating "compass" clumps to detect the field, and you'll find that a ring of magnets has almost no effect on them.

Two rings become one:

Make a long chain using an even number of magnets (60 magnets is good.) Connect the chain to form a circle, then squash the circle to form two parallel rows having squared-off ends. Lift the ends up and join them, forming a 2-row bracelet. If you now carefully peel the two rows apart by a bit, then suddenly pull them wide, the whole thing snaps to become a single large circle!

Where to buy 'em?



  • WARNING: if you mess with TWO large neodymium supermagnets, you are bound to pinch your fingers if you let them come together. This REALLY HURTS, it's like pliers. Also, you should wear safety glasses, since when two NIB magnets slam together, they tend to launch tiny chips at high velocities. I've had 3/4" magnets acidentally fall together and be completely shattered. The things are brittle! The ones larger than 1/4" are NOT FOR UNSUPERVISED CHILDREN!

    NEW DANGER: swallowed magnets can cause lethal stomach problems. See

    Created and maintained by Bill Beaty. Mail me at: .