Pontallier has even hired Norman Foster to design new buildings at the chateau. At its centre will be a research and development facility that will house 40 vats, so that experimental crops can be vinified in larger quantities than present.

As the chateau says it wishes to share its research and be transparent about its findings, the walls of the test facility will be made out of glass.

Given the global influence of Bordeaux winemaking styles and the longevity of the wines, any changes could have a far-reaching impact. Pontallier cannot afford to make a mistake.

"The conclusions drawn today will have an impact for 20 years," he says.

The experiments started in the vineyards. A plot was set aside six years ago that would normally produce grapes for the estate's second wine, Pavillon Rouge, and its only white, Pavillon Blanc. Cabernet sauvignon and semillon grapes are grown using conventional, organic and biodynamic techniques, made into wine, aged in the bottle rather than the barrel and then tasted after a few years.

Recently, a new plot has been given over to experiments, one that could yield grapes for the first growth wine.

Pontallier seems to have already come to a partial decision on this front - the rest of the chateau's vineyards are now organic, but the experiments continue to see if fewer chemicals can be used.

Pontallier says, "We would like to know better how to control disease. The big problem is how to find a totally organic alternative to chemicals for mildew. We already use an organic treatment - copper sulphate. It's considered organic, it's a naturally occurring chemical.

"Alternatively, are there any totally natural products such as plant extracts that promote natural defences against viruses?"

Pontallier has hired a consultant to help implement biodynamic winemaking techniques, which can include the use of crystals in the vineyard, the burying of a bull's horn full of manure for a year and the consequent spraying of the grapes with the resulting goo, and the pressing of grapes according to the phases of the moon. A full moon is said to exert more gravitational pressure resulting in a better crush of the grapes.

Biodynamics can be more practical, and one step Chateau Margaux has taken is to reintroduce horses to the vineyards rather than tractors. Horses reportedly do not compact the soil as much as tractors, but this turned out to be an irrelevance on the estate's pebbly soil. There was, however, an unexpected benefit.

Horses guided by men are less likely to accidentally destroy or pull out the precious vines - plants that take 15 years to grow again.

"We found that with a good horse and the right man behind it we lost fewer vines, but it depends on the man behind the horse," he says.

Pontallier is in two minds about the biodynamic outlook and its adherents. One reason for that is that many fans seem to have had Damascene conversions, and the scientist in him is suspicious of an overnight change of mind. He explains that "it seems beyond science. That's why it looks a bit weird to me. I'm sceptical about the fundamentals. It looks a bit like religion, but it's impossible to write off religion, so I can't write off this solution either. I'm very sceptical."

In the Hong Kong blind tasting there was a just about equal preference for the organic and biodynamicwine, with a small number favouring the conventional, but the split was too insignificant to draw any major conclusions.

Pontallier says that if biodynamics proved to produce the best wine he would have to accept the conclusion. "Even if I don't understand it, even if it looks weird - fact is fact. Rationality is about accepting facts."

In fact, he explains that rejecting the science would also be a rejection of the traditions of the chateau.

"As far as tradition is concerned, it has always evolved depending on the progress of science and technology."

However, he also says that the best wine he can make is one in the Chateau Margaux style. The estate is noted for producing wines characterised as "soft and powerful" - having great structure and length but tannins so subtle they are barely noticeable.

While the research and development department has a brief to question everything, even to the point of annoying Pontallier, the area of greatest interest is the closure used on the bottle.

Pontallier agrees that losing corks would take some of the romance out of wine, which is one very good reason for not making a hasty decision.

"We still have 20 years to think about it," he says.

Corks have their drawbacks, too. If you're buying your wine to keep for several decades then you want to know that it will develop in the bottle in a fairly predictable way. You don't want to open your HK$1,330 bottle of Chateau Margaux 2005 to be greeted with the smell of mouse droppings that tell you the cork is tainted.

"The closure is the only thing we don't control," says Pontallier.

As a result he is experimenting with totally airtight screw caps and screw caps that allow for some air to penetrate. At the Hong Kong tasting it was wine stored in the latter that won the seal of approval.

Pontallier is a long way from giving up on corks, however. The technology exists to individually test corks for taint without destroying them. But he wants a Plan B in case the oak trees that produce cork become diseased.

Pontallier is also keeping a wary eye on global warming, without being in any hurry to draw conclusions.

"Everybody accepts it's happening, but nobody knows the consequences for sure."