In most countries the word "metro" is used for an underground railway in a large city. In Switzerland it means an underground railway for the whole country.

Imagine being propelled right across the country in 75 minutes, from Geneva in the south west to St Gallen in the north east, without seeing a car, a lake or even a blade of grass. If Swissmetro is ever built, that's exactly what you'll be able to do - and, say its supporters, your ticket will cost you no more than it would for a normal train.

The scheme envisages the construction of a total of 680 km (400 miles) of tunnel linking the main cities. The first line is supposed to have stations at Geneva, Lausanne, Bern, Lucerne, Zurich and St Gallen; the second will start at Basel, cross the first at Lucerne and go on to Bellinzona, and subsequent lines will run from Lausanne to Sion and St Gallen to Chur. Eventually there could be extensions across the Swiss frontier... The trains will not come out of their tunnel, and only the above-named towns are to be served.

The idea was first put forward back in 1974, by Rodolphe Nieth, a civil engineer from Lausanne, who aroused the interest of the Federal Institute of Technology there. For the last 10 years it has been promoted by a private company, Swissmetro AG. It has nothing to do with the SBB, although its promoters have said they would be happy to see the SBB running it.

Supporters of Swissmetro point to its many advantages. It will cut travelling times between the main cities considerably. (That 75 minute trip from Geneva to St Gallen takes 4 hours 8 minutes by the SBB's intercity express.) It will not be affected by even the most severe weather conditions, and it will also be much safer. Delays caused by bad weather and accidents are not only a nuisance, they are also expensive. Time wasted in motorway hold-ups now cost the economy several hundred million francs a year, the Swissmetro supporters say, so the underground railway will bring cost benefits too. And then the system is ecologically friendly. No more countryside would disappear under concrete for the construction of the track and stations. There would be no noise or any other kind of pollution. And not only would it be environmentally friendly itself, but it would attract customers who would otherwise be using road or air, and thus help cut down the pollution caused by those means of transport.

As for the technology, the tunnels and the trains complement each other.

The idea is that there will be two tunnels, one in each direction, each with a diameter of 5 meters (16.4 ft), built at depths of between 60 and 300 meters (200 and 980 ft), depending on the topography. Air pressure in the tunnels will be reduced by means of electrical pumps, which means that the trains will move faster and consume less energy because there will be less air resistance.

Propulsion of the trains will be by linear electric motors, either on board, or in the tunnel. Beneath the trains will be a magnetic levitation and guidance system.

The trains will be about 200 meters (650 ft) long, and able to take some 800 passengers. They will board and leave the trains from underground stations which will be connected to the overground transport systems.

However, such technology does not yet exist and it is far from clear whether it is feasible.

The government has in general shown itself attracted to the idea, but wary. In a communiqué in 1999, responding to Swissmetro's application for a concession to build a Geneva-Lausanne pilot line, it described the project as "promising." "The proposed technology could be a quantum leap for a durable transport policy," it said. "But the weak point is still the financing of the proposed pilot stretch." It turned down flat Swissmetro's request for three billion francs (2.1 billion dollars at the exchange rate of the time) from the public sector and refused to grant the concession until the financing was clear. (Swissmetro had proposed raising another 700-800 million - 490-560 million dollars - from private investors.)

Unfortunately, Swissmetro's plans for whisking passengers between the major cities would compete with the Rail 2000 project. Rail 2000's total costs - the first stage is budgeted at 7.4 billion francs (5.1 billion dollars) - pale into insignificance in comparison with Swissmetro, but it would still be a lot of money to spend on a project condemned to become obsolete within a few years. Large sums have also been allocated to the NEAT construction, which cannot be diverted.

Despite a sneaking fascination for the boldness of the project, there is widespread scepticism as to whether it will ever be built. It was dealt what could be a death blow in November 2002, when Moritz Leuenberger told a newspaper that as transport minister he no longer believed in the project - though he admitted that the research was still interesting. On the other hand, many grandiose projects - not least the Channel Tunnel linking Britain and France - were planned, and even technically possible, many decades before they were actually built. Swissmetro might yet run, even if not in the lifetime of many of us alive today.