On the internet it is already a spreading legend: did the mass stranding and deaths of whales and dolphins on an Australian beach signal the advent of the earthquake that caused the Boxing Day tsunami?

And did an Indian professor, as a result of the first event, warn of the second?

You might think it's a pretty wacky idea. But it's got currency. Yet is it true? What is true is that on December 4, three weeks before the earthquake off Indonesia, an Indian academic, Dr Arunachalam Kumar, professor of anatomy at Kasturba Medical College at Mangalore in Karnataka, posted a note about a recent whale-stranding in Tasmania, and its possible implications, on a "listserve", an e-mail distributor, hosted by Princeton University.

About 120 whales stranded and died in Australia at the end of November and 50 pilot whales died on a Coromandel beach at the same time.

Kumar is a well-known figure in India. An amateur naturalist of some repute and a prolific author, he is a larger-than-life character, frequently in the press.

"It is my observation, confirmed over the years, that mass suicides of whales and dolphins that occur sporadically all over the world, are in some way related to change and disturbances in the electromagnetic field co-ordinates and possible realignments of geotectonic plates thereof," he wrote.

"Tracking the data and plotting the locales of tremors and earthquakes, I am reasonably certain that major earthquakes usually follow within a week or two of mass breaching of cetacians [sic]. I have noted with alarm, the last week report of such mass deaths of marine mammals in an Australian beachside. I will not be surprised if within a few days a massive quake hits some part of the globe. The interrelationship between the unusual 'death-wish' of pods of whales and its inevitable aftermath, the earthquake, may need a further impassioned and unbiased looking into."

There's no doubt that he posted his note on December 4. To read it in the listserve itself, go to new-lists.princeton.edu/listserv/nathistory-india.html and click on "December 2004". In reading it, many are likely to experience a rising of the hair on the back of the neck.

But the story hasn't remained there. It has been widely reported across India and the net. And, in the telling, the story has grown. On January 10, it surfaced on the discussion board of the electronic version of the British Medical Journal. There, a letter from one Jairaj Kumar Chinthamani, a research fellow in Mangalore, said the professor had predicted the earthquake "almost to the day". He actually said "within a week or two" and "within a few days". The quake took place three weeks later.

According to Chinthamani, the professor "wrote that he had made a five-year record of dates and locales of whale strandings, plotted their locales, and correlated them to occurrences of upheavals on land or undersea, and had observed a remarkable connection between the events. In fact, Kumar never mentioned anything as precise as a five-year record.

But never mind. It would not be surprising if the legend continued to grow until eventually Kumar was regarded as having signalled the breaching on shore of the tsunami itself to the very minute.

His original message, though, is intriguing enough. Yet does it have any substance? The answer is that it may have some.

Scientists are aware of the possible connection between the behaviour of cetaceans (whales and dolphins) and the Earth's magnetic field.

"There is thought to be a correlation between some whale strandings and geomagnetic anomalies," says Dr Simon Northridge, of the Sea Mammal Research Unit at St Andrews University, Britain's principal whale and dolphin research centre. "It's certainly out there as a hypothesis."

In fact, the idea was put forward in the late 1980s by Dr Margaret Klinowska, from Cambridge University. Klinowska argued whales navigated partly by following geomagnetic contours, and that in certain circumstance, such as when the contours ran at right angles to the coastline (rather than parallel to it) they could run themselves aground.

The theory is still discussed, and it is a respectable one. But are stranded whales precursors of earthquakes? You won't find a lot of backing for that.

Mark Simmonds is director of science at Britain's Whale and Dolphin Conservation Society (WDCS).

He has studied strandings in detail because one of the questions most frequently asked of him and the WDCS is why they happen. The main answer is, he says, that most whales are intensely social animals, and act together. If one heads into the beach, the others follow. It may be an accident; sometimes human agency may be partly to blame; sometimes the Earth's magnetic field may play a role.

"But nobody has shown any correlation between whale strandings and earthquakes. If you're saying there is, you would have to present the data to prove your case."

Over to Professor Kumar. His original email strongly implies that he is in possession of such data. But, contacted at his office in Mangalore, he was unable to provide any.

Did he have a list of the correlations between previous whale strandings and earthquakes? The correlations in which he had tracked the data and plotted the locales? "I don't have a lot of these things," he said. "I'm just an avid reader. I watch with particular interest.

"As a science man, I don't want to put these things on paper," he replied. "It would take me a long time to put it right."

So Kumar appears to have no evidence at all for backing up his core assertion that cetacean strandings and earthquakes are linked.

Yet he undoubtedly did post his solemn warning just three weeks before the biggest earthquake of the past 40 years: "I will not be surprised if within a few days a massive quake hits some part of the globe." Chance? Luck? Science? Make of it what you will. Plenty of others are.