But now a collaboration between the Keays lab at the Institute of Molecular Pathology (IMP) in Vienna and researchers at The University of Western Australia's Centre for Microscopy, Characterisation and Analysis (CMCA) has added some important pieces to the puzzle.
Their work, published today in the online version of Current Biology, reports the discovery of iron balls in sensory neurons. These neural cells, called hair cells, are found in the ear and are responsible for detecting sound and gravity. Remarkably, each cell has a single iron ball, and it's in the same place in every cell.
"It's very exciting. We find these iron balls in every bird, whether it's a pigeon or an ostrich, but not in humans," said Mattias Lauwers, the IMP researcher who discovered the balls.
CMCA research associate Dr Jeremy Shaw, who has studied iron in a range of animals from molluscs to humans, said it was an astonishing finding.
"Despite decades of research, these conspicuous balls of iron had never been observed previously," Dr Shaw said. "Nature keeps surprising us with the various ways iron can be utilised by animals."
Dr Shaw and fellow CMCA researcher Professor Martin Saunders, both of whom specialise in the use of analytical electron microscopy, helped to analyse the new iron structures and to describe them in the paper.
The finding builds on previous work by the Keays lab and the CMCA which last year showed that iron-rich cells in the beak of pigeons - previously believed to be the magnetic sensors - were really just blood cells.
"These cells are much better candidates, because they're definitely neurons," said IMP researcher and group leader, Dr David Keays. "But we're a long way off understanding how magnetic sensing works - we still don't know what these mysterious iron balls are doing.
"Who knows? Perhaps they are the elusive magnetoreceptors. Only time will tell."
Not so new...
The theorised magnetic guidance system goes back to at least the late 1940s. Proof of the ability, in several species, was shown in the following decades.
In 1975 Richard P. Blakemore, then a graduate student at the University of Massachusetts in Amherst, astonished the world of biology with the announcement that some bacteria, the lowliest of all cells, also had a magnetic sense. Blakemore made the discovery when, studying the salt marshes of Cape Cod, he noticed that one type of bacterium always oriented itself north-south on his microscope slides. Soon he found mag- netotactic bacteria (those reactive to magnetism) near Cambridge, Mas- sachusetts, where he set about studying them with Richard Frankel of MIT's magnet lab. The direction to magnetic north points through the earth somewhat down from the horizon, and the scientists became con- vinced that the bacteria were using the field to guide themselves ever downward to the mud where they throve, since they were too small to sink through the random molecular motion of the water around them.
Blakemore's electron micrographs soon revealed a surprising structure. Each bacterium contained within it, like a chain of cut jet stones, a straight line of magnetite microcrystals. Surrounded by a thin mem- brane, each of these particles was a single domain, the smallest piece of the mineral that could still be a magnet.
In 1983, using magnetic measurements in selective-shielding experi- ments, Baker and his co-workers reported locating magnetic deposits close to the pineal and pituitary glands in the sinuses of the human ethmoid bone, the spongy bone in the center of the head behind the nose and between the eyes. It's interesting to note that selective-shielding studies done in the early 1970s, by Czech emigre biophysicist Zaboj Harvalik, an adviser to the U.S. Army Advanced Material Concepts Agency, pointed to this same spot as one of two areas—the other was the adrenal glands—where the dowsing ability resided. In 1984 a group headed by zoologist Michael Walker of the Univer- sity of Hawaii in Honolulu isolated single-domain magnetite crystals from a sinus of the same bone in the yellowfin tuna and Chinook salmon. The crystals were of a shape normally shown only by magnetite synthesized by living things rather than geological processes. Abundant nerve endings entered the magnetic tissue, and the crystals were orga- nized in chains much like those in magnetotactic bacteria. Each crystal was apparently fixed in place but free to rotate slightly in response to external magnetic forces. Calculations showed that such chains would be able to sense the earth's magnetic field with an accuracy of a few seconds of arc, or a few hundred feet of surface position. This result correlated perfectly with earlier homing studies on live tuna by the same group.