Back in 1986, a Patriot missile managed to shoot a ballistic missile out of the sky. It was during an experiment in the New Mexico desert, and the American missile's developers took pride in its wonderful performance and said the Patriot was the best missile defense system ever built. A year later, a Patriot shot another missile out of the sky. This time it was another Patriot, made to simulate a Soviet surface-to-surface SS-23.

Three years later, Patriot developers continued to launch it against test missiles and each time, the results were impressive; altogether, the Patriot was tried more than 10 times and each time it knocked its target out of the sky. In 1990, the missile system was operational and deployed by the U.S. Army to protect its troops from ballistic missiles.

Then in 1991, the Gulf War broke out and the Patriot was tried for the first time off the testing grounds. Altogether 158 Patriots were launched during the war, from batteries positioned in Israel and Saudi Arabia, trying to head off a total of 47 Iraqi El Hussein missiles. The Patriots did not manage to take down a single missile.

It is very important to remember the Patriot's dismal operational performance before getting carried away in the enthusiasm of the developers of the Arrow, which last week managed to knock down a Scud in a test in the U.S. There's no minimizing the technological achievement, but from there to the axiomatic "closing of the breach that had made the rear vulnerable to long-range missiles launched from distant sites," as Maj. Gen. (res.) Yitzhak Ben-Yisrael claimed, is very far.

The unsuccessful Patriot launches show that weapons systems that work well during their tests can fail in war. While experiments are conducted in a controlled environment, during a war, developments occur that the engineers and strategists didn't take into account, and the enemy weapons perform differently than in the scenarios on which the experiments and tests are based. That's what happened to the Patriot, which was designed for use against certain types of ballistic missiles, which suddenly behaved in unexpected ways. As opposed to what was planned, the Iraqi missiles broke up, their warheads took unexpected turns, and the result was a total failure of the Patriots' attempts to down the Scuds.

The successful test of the Arrow last week in the U.S. also took place in a controled experimental environment. The target missile, a Scud, was a lone missile. It came from a known direction, at a known angle, at a speed set in advance, on a predetermined trajectory and at a precise speed - without any attempts to jam or deceive predators deployed against it. The Arrow was launched toward it according to a preplanned schedule and the experiment did not include any surprises and changes in the target's performance that were not anticipated.

In real time, in a real war, the Arrow operators won't have any data about the attacking missiles. The missiles will come by surprise and possibly in large numbers, in one barrage. It is reasonable to assume that in the not very distant future the ballistic missiles of the region will be equipped with decoys, as well as other mechanisms meant to disrupt Arrow's radars. Some will be equipped with submunitions, meaning 100 or more small containers of chemical or biological materials that will fill the warheads of each missile. These containers will be released shortly after launch, far from the Arrow's range, and then continue the ballistic trajectory to their target area. Clearly, the Arrow won't be able to fire hundreds of missiles to try to bring down the submunitions of even one missile.

Other missiles will have maneuverability upon reentry to the atmosphere. It is doubtful the Arrow will be useful against all that. And since there is no intention - and in some cases, it is impossible - to test the Arrow against such eventualities, the system could encounter them for the first time in war conditions.

Those in the defense establishment and army who are responsible for development and operation of the Arrow are exploiting public ignorance, and particularly public fears, to disseminate illusions. They know very well that success in one experiment, against one missile, does not mean all the technological obstacles are out of the Arrow's way. They certainly remember the experience of the Gulf War.

Therefore, when they promise that the Arrow will provide the best defense against ballistic threats expected in the future, and when they talk about "closing" the skies to hostile missiles, it is somewhat irresponsible. Their participation in the media celebrations that accompanied the Arrow test in the U.S. says something about the immunity from criticism they enjoy. They know very well that when this balloon bursts, none of them will be required to account for themselves. It's our custom never to make anyone pay a price for spreading baseless promises. But the frightened public should listen with considerable skepticism to those responsible for security, and be sure not to forget that the Arrow was preceded by the Patriot.