The more a government emphasises its commitment to defence, the less it seems to care about the survival of its people. Perhaps it is because its attention may be focused on more distant prospects: the establishment and maintenance of empire, for example, or the dynastic succession of its leaders. Whatever the explanation for the neglect of their security may be, the people of America have discovered that casual is the precursor of casualty.
But while we should be asking what George Bush and his cabinet knew and failed to respond to before September 11, we should also be exploring another, related, question: what do they know now and yet still refuse to act upon? Another way of asking the question is this: whatever happened to the anthrax investigation?

After five letters containing anthrax spores had been posted, in the autumn, to addresses in the United States, the Federal Bureau of Investigation promised that it would examine "every bit of information [and] every bit of evidence". But now the investigation appears to have stalled. Microbiologists in the US are beginning to wonder aloud whether the FBI's problem is not that it knows too little, but that it knows too much.

Reducing the number of suspects would not, one might have imagined, have been too much to ask of the biggest domestic detective agency on earth. While some of the anthrax the terrorist sent was spoiled during delivery, one sample appears to have come through intact. The letter received by Senator Tom Daschle contained one trillion anthrax spores per gram: a concentration which only a very few US government scientists, using a secret and strictly controlled technique, know how to achieve. It must, moreover, have been developed in a professional laboratory, containing rare and sophisticated "weaponisation" equipment. There is only a tiny number of facilities - all of them in the US - in which it could have been produced.

The anthrax the terrorist sent belongs to the "Ames" strain of the bacterium, which was extracted from an infected cow in Texas in 1981. In December, the Washington Post reported that genetic tests showed that the variety used by the terrorist was a sub-strain cultivated by scientists at the US army's medical research institute for infectious diseases (USAMRIID) at Fort Detrick, Maryland. That finding was publicly confirmed two weeks ago, when the test results were published in the journal Science. New Scientist magazine notes that the anthrax the terrorist used appears to have emerged from Fort Detrick only recently, as the researchers found that samples which have been separated from each other for three years acquire "substantial genetic differences".

The Ames strain was distributed by USAMRIID to around 20 other laboratories in the US. Of these, according to research conducted by Barbara Hatch Rosenberg, who runs the Federation of American Scientists' biological weapons monitoring programme, only four possess the equipment and expertise required for the weaponisation of the anthrax sent to Senator Daschle. Three of them are US military laboratories, the fourth is a government contractor. While security in all these places has been lax, the terrorist could not have stolen all the anthrax (around 10 grams) which found its way into the postal system. He must have used the equipment to manufacture it.

Barbara Hatch Rosenberg has produced a profile of the likely perpetrator. He is an American working within the US biodefense industry, with a doctoral degree in the relevant branch of microbiology. He is skilled and experienced at handling the weapon without contaminating his surroundings. He has full security clearance and access to classified information. He is among the tiny number of Americans who had received anthrax vaccinations before September 2001. Only a handful of people fit this description. Rosenberg has told the internet magazine Salon.com that three senior scientists have identified the same man - a former USAMRIID scientist - as the likely suspect. She, and they, have told the FBI, but it seems that all the bureau has done in response is to denounce her.

Instead, it has launched the kind of "investigation" which might have been appropriate for the unwitnessed hit and run killing of a person with no known enemies. Rather than homing in on the likely suspects, in other words, it appears to have cast a net full of holes over the entire population.

In January, three months after the first anthrax attack and at least a month after it knew that the sub-strain used by the attacker came from Fort Detrick, the FBI announced a reward of $2.5m for information leading to his capture. It circulated 500,000 fliers, and sent letters to all 40,000 members of the American Society for Microbiology, asking them whether they knew someone who might have done it.

Yet, while it trawled the empty waters, the bureau failed to cast its hook into the only ponds in which the perpetrator could have been lurking. In February, the Wall Street Journal revealed that the FBI had yet to subpoena the personnel records of the labs which had been working with the Ames strain. Four months after the investigation began, in other words, it had not bothered to find out who had been working in the places from which the anthrax must have come. It was not until March, after Barbara Hatch Rosenberg had released her findings, that the bureau started asking laboratories for samples of their anthrax and the records relating to them.

To date, it appears to have analysed only those specimens which already happened to be in the hands of its researchers or which had been offered, without compulsion, by laboratories. A fortnight ago, the New York Times reported that "government experts investigating the anthrax strikes are still at sea". The FBI claimed that the problem "is a lack of advisers skilled in the subtleties of germ weapons".

Last week, I phoned the FBI. Why, I asked, when the evidence was so abundant, did the trail appear to have gone cold? "The investigation is continuing," the spokesman replied. "Has it gone cold because it has led you to a government office?" I asked. He put down the phone.

Had he stayed on the line, I would have asked him about a few other offences the FBI might wish to consider. The army's development of weaponised anthrax, for example, directly contravenes both the biological weapons convention and domestic law. So does its plan to test live microbes in "aerosol chambers" at the Edgewood Chemical Biological Center, also in Maryland. So does its development of a genetically modified fungus for attacking coca crops in Colombia, and GM bacteria for destroying materials belonging to enemy forces. These, as the research group Project Sunshine has discovered, appear to be just a tiny sample of the illegal offensive biological research programmes which the US government has secretly funded. Several prominent scientists have suggested that the FBI's investigation is being pursued with less than the rigour we might have expected because the federal authorities have something to hide.

The FBI has dismissed them as conspiracy theorists. But there is surely a point after which incompetence becomes an insufficient explanation for failure.