A wide variety of animals are able to sense small variations in an ambient magnetic field. Magnetotactic bacteria, for example, living in the impenetrable gloom of mud on the ocean floor use a magnetic sense to help them tell up from down. In the absence of visual cues, honey bees rely on a sense of magnetism to build their combs in a north-south orientation. And it has proved possible to train tuna fish to swim through a frame in response to magnetic cues. Some species of bird, including robins and homing pigeons, rely on a magnetic detection system to help them navigate when other cues, such as the position of the Sun and stars, aren't available. Perhaps most intriguing of all, whales appear to use the magnetic stripes that were produced during the spreading of the ocean floors as direction indicators to guide them on their long-distance travels.

 

Do we humans, too, have a sense of magnetism? To test this possibility, Robin Baker, of the University of Manchester, drove groups of blindfolded student volunteers along twisting roads up to 30 miles away from their campus. Then he asked them to indicate the way back to their starting point. Interestingly, the students tended to give good estimates of the direction home while still blindfolded, but became disoriented when the blindfolds were removed, as if the overpowering sensation of sight masked a more subtle one of magnetism.

 

In a second experiment, Baker took a bus-load of blindfolded schoolchildren to an unknown destination three miles southwest of their school. Half of the children had bar magnets attached to their heads, whereas the others, although they believed they were wearing magnets, actually had dummy magnets instead. When they arrived at their destination, the children were asked to point in which direction they thought their school lay. Baker found that while the children wearing the dummy magnets generally gave good estimates of the school's bearing, those with the real magnets tended to give estimates that were 90 degrees anticlockwise of the true bearing.

 

If humans do have an unconscious ability to detect magnetism then this offers one possible explanation of how dowsing works. In 1970, Duane Chadwick of Utah State University carried out a series of experiments on 150 dowsers and found that the dowser's rod movements were related to minute variations of the Earth's magnetic field. In 1986, Gene Simmons, a geophysicist at the Massachusetts Institute of Technology, came to a similar conclusion having conducted surveys of magnetic fields around two wells discovered by dowsers near Boston. Unlike most other wells drilled in the crystalline rocks of the district, the dowsed wells yielded large quantities of water – upward of 30,000 gallons per hour. Simmons found that the dowsers had sited both holes within a narrow magnetic anomaly only a few yards wide. The anomaly resulted from a fracture zone that was channeling the flow of ground water and so giving rise to the observed high flow rate. A magnetic sense could also explain how dowsers are able to locate veins of metallic ore. These veins are usually associated with faults or fracture zones, which often produce magnetic anomalies, as, in some cases, do the ores themselves.

 

In plying their trade, dowsers tend to use either a Y-shaped stick, traditionally of hazelwood, or, more commonly these days, two L-shaped metal rods, the short branches of which are held loosely upright in the hands. Even the slightest inward rotation of the arms will cause the Y-shaped stick to bend down or the L-shaped rods to cross. The general supposition, the, is that when the dowser subconsciously senses the presence of the sought-after material, her arm muscles involuntarily react, causing the diving tool to respond.

 

But how could human beings detect magnetism in the first place? Chains of small crystals of magnetite – magnetic iron oxide – have been found in many species of animal known to possess a magnetic sense. These chains often seem able to rotate so that they can act like microscopic compass needles. A similar concentration of magnetic materials has been found by Richard Baker and his colleagues in the very thin bones of the human sinuses, in the front of the skull, suggesting that we may indeed have an anatomical adaptation allowing us to detect subtle variations in our magnetic environment. Just how acute this sense could be is hinted at by experiments on a rather different variety of vertebrate, the yellowfin tuna, carried out by Michael Walker of the University of Hawaii and colleagues at the California Institute of Technology and the National Marine Fisheries Service in La Jolla. In the yellowfin tuna, a sensor consisting of an estimated 85 million microscopic crystals of magnetite allows magnetic field changes as small as one nanotesla, or less than one twenty-thousandth of the Earth's field strength, to be detected. Such a sensitive system, it it existed in humans, would be perfectly capable of accounting for many of the positive results claimed – and, in some cases, experimentally verified – for the phenomenon of dowsing.

 

Of course, this remains a highly contentious subject. Some researchers who have investigated it have failed to find anything definite at all. But the spectacular practical success of dowsers demands a serious scientific response – and not just because the phenomenon might be intrinsically fascinating. When an estimated 80 percent of all diseases in the Third World are caused by polluted water, and at least two billion people have inadequate water supplies, the benefits that could come from a more effective way of locating deep aquifers are obvious.