Magnets are remarkable exemplars of fairness—each north pole is invariably accompanied by a counterbalancing south pole. Split a magnet in two, and the result is a pair of perfectly neutral magnets, each with its own north and south.

For decades researchers have sought the exception to this rule of fairness and balance: the magnetic monopole. Magnetism's answer to electricity's negatively charged electron, a monopole would be a free-floating carrier of either magnetic north or magnetic south—a yin unbound from its yang.

A pair of papers published online this week in Science offer experimental evidence that such monopoles do in fact exist, albeit not as electron-like elementary particles, a caveat that one self-professed purist says disqualifies them from genuine monopole status.

Both studies examine the magnetic behavior of a family of rare-earth materials known as spin ices—one group using holmium titanate and the other dysprosium titanate. The man-made spin ices take their name from their similarity to water ice—at the molecular level their internal magnetic structure is analogous to the arrangement of protons in ice.

Claudio Castelnovo, a postdoctoral physicist at the University of Oxford who co-authored one of the Science papers and also co-wrote a paper in Nature last year describing how monopoles might be realized in spin ices, explains that the compounds offer a peculiar combination of order and freedom that facilitates the dissociation of the poles.

At low temperatures, there is still some magnetic wiggle room in the spin ice's lattice structure, but not a lot—the magnetic freedom of the system is frustrated, so to speak. "As a result, this is a substance that has degrees of freedom that look the same, microscopically, as you would see in a fridge magnet," Castelnovo says. "But a fridge magnet is able to order so as to act as a fridge magnet and stick to metals, while this one is not able to achieve this level of ordering in spite of having this magnetic structure inside, because of this frustration."

Internally, the tiny magnetic components arrange themselves head to tail in strings, like chains of bar magnets stretching across a table in different directions. In a very cold, clean sample, those strings form closed loops. But excitation induced by a rise in temperature can introduce tiny defects in these chains, Castelnovo says—in the bar-magnet analogue, one of the magnets is flipped, breaking the head-to-tail continuity. "You have your path that is north–south–north–south, and at a certain point one of the needles actually twists 180 degrees and points the wrong way," he explains.

On either side of that defect, all of a sudden, is a concentration of magnetic charge—two norths at one end, two souths at the other. Those concentrations can float free along the string, acting as—voilà—magnetic monopoles.

"The beauty of spin ice is that the remaining degree of disorder in this low-temperature phase makes these two points independent of each other, apart from the fact that they attract each other from a magnetic point of view because one is a north and one is a south," Castelnovo says. "But they are otherwise free to move around."

Of course, this method of synthesizing monopoles cannot bring a north into existence without also generating a south—the key is their dissociation. "They always have to come in pairs," Castelnovo says, "but they don't have to be anywhere specifically in relation to one another."

But Kimball Milton, a University of Oklahoma physicist who wrote a 2006 review article summarizing the status of monopole searches, is not convinced. "These are not magnetic monopoles," he says.

"I might object to [the researchers] saying 'genuine magnetic monopoles', because when you say genuine, that implies to me it's a point particle, and it's not," Milton says. "It's an effective excitation that at some level looks like a monopole, but it's not really fundamentally a monopole."

He also says it is "completely wrong" to describe, as the research teams do, the chain of magnetism within spin ices as a Dirac string, a hypothetical invisible tether with a monopole at its end that was envisioned in the 1930s by English physicist Paul Dirac. "But that's just because I'm a purist," Milton says.

By his assessment, the magnetic strings in the spin ice do not fit the Dirac definition because they are, in fact, observable and merely carry flux between two opposing so-called monopoles. "Real monopoles, if they existed, would be isolated, and the string would run off to infinity," he says.

"I'm not trying to put down the experiment or the work in any way," Milton says. "I'm sure [the new studies] are important in the field of condensed matter physics. They're not important from a fundamental point of view."