"About this big." High in the United Nations building in New York, a U.N. official is holding his arms out in a circle, like a man gripping a beach ball. "About a yard across, weighing about a ton."

This is the Iraqi bomb—slightly smaller than the one dropped on Hiroshima, but nearly twice as powerful—packing an explosive force of at least 20,000 tons of TNT. The official is dramatizing a drawing he has made in his notebook, based on documents seized in Iraq. He is sure that the bomb, if built to the specifications in the drawing, will work.

At the bomb's center is an explosive ball of weapon-grade uranium. Around this is a layer of natural uranium to boost the yield and a second layer of hardened iron to keep the core from blowing apart prematurely. If the bomb is to detonate properly, these parts must have just the right dimensions, and there must be a firing circuit accurate to billionths of a second. Documents in the United Nations' possession show that the Iraqis have all the right dimensions and the necessary firing circuit.

This is the bomb that, according to U.N. estimates, Saddam Hussein was 18 to 24 months from building when the gulf war started. It is the bomb he is still likely to build, despite the war and the most intrusive nuclear inspections in history, unless the United Nations changes its tactics.

"They are pouring concrete as we speak," says a U.N. official at the next desk. Saddam, he says, is rebuilding the bombed nuclear sites in plain view of the U.N. inspectors. "He is even planting trees and re-landscaping," he adds, "to boost employee morale." Another U.N. official has a similar story. During a visit to the Iraqi nuclear weapon testing site at Al Atheer, he says, his Iraqi hosts looked him in the eye and said, “We are waiting for you to leave."

Since the inspections started last spring, the Iraqi disinformation specialists who serve as guides have done their best to outfox the inspectors. In one instance, the Iraqis hid reactor fuel by loading it on the back of a truck and driving it around the reactor site, always staying about 200 yards in front of the inspection team. The fuel contained weapon-grade material.

Perhaps the most notorious confrontation occurred when inspectors followed an intelligence tip to a cache of sensitive documents. In an attempt to elude the Iraqis, each of the 44 team members hid a stack of papers inside his clothing. Rather than strip-search the inspectors before video cameras the Iraqis simply forbade them to leave, leading to a four-day standoff in a Baghdad parking lot under a scorching summer sun. Only alter a unanimous vote of support by the Security Council did Iraq finally relent.

That spirited encounter is now as much a part of history as the brief triumph or the 100-hour war. Under the cease-fire terms, inspectors for a U.N. Special Commission were charged with the "destruction, removal or rendering harmless" of Iraq's nuclear weapon potential. But after months of chasing increasingly fruitless intelligence leads, morale on the Special Commission is scraping rock bottom.

The Iraqis know it, too. "They've started laughing at us," one U.N. official says, adding that the Iraqis have even threatened individual inspectors. "They have basically told our people that they know where we live," he says in exasperation.

The problem is that the inspectors have exhausted their information. The first inspections were fueled by leads from Iraqi defectors and the chance discovery of the sensitive documents in Baghdad. But that luck has run out just as the Iraqis have organized their resistance to the inspections. Recently, in fact, they told the inspectors that "you won't find any more documents in this country."

That remark came after a U.N. team had charged into several suspected reactor sites, following intelligence leads that turned out to be duds. "All we found were empty warehouses, cement factories making real cement and prisons with real prisoners," one inspector says. The inspectors believe they have reached a dead end.

The inspectors' defeat raises a chilling prospect: In the absence of a major new U.N. effort, Saddam Hussein is still likely to get the bomb. Thus, Iraq has become a test case for nuclear proliferation. If war and a full-court press by the United Nations cannot stop an outlaw nation like Iraq from making the bomb, what will it take to stop countries like Iran, North Korea and Libya?

In a sense, what is being played out in Iraq is the first battle of a new cold war, fought with spies, international pressure and export controls. The West may have won the first cold war against the Soviet Union, but it is losing the second to Iraq and other nations that want to get the bomb.

Saddam has hauled himself up the nuclear mountain on a chain of high-tech exports, sold by the very Western countries whose inspectors—now on loan to the United Nations—are trying to find them. Other similarly favored nations could easily follow Saddam's example, given existing export laws. Iran and Libya are now maneuvering into this position.

Iraqi scientists know, for example, how to cast uranium metal into bomb parts in a vacuum furnace. The vacuum prevents molten uranium from burning in air. At Al Atheer, U.N. inspectors found vacuum furnaces made by a German firm, Arthur Pfeiffer Vakuum Technik. The inspectors rejected Iraq's claim that the furnaces were for scientific research.

The inspectors also found a large "isostatic" press, made by a Swedish-Swiss firm, Asea Brown Boveri. This, too, the Iraqis claimed was for research. But the U.N. team thinks the machine was for shaping the high-explosive charges that set off a nuclear chain reaction. These specially shaped charges are wrapped around the bomb core and set off simultaneously, creating a shock wave that travels inward, "imploding" and compressing the core. When the core is compressed to sufficient density, the nuclear chain reaction begins.

How did the Iraqis learn to use such specialized equipment? In large part from the United States Government. In August 1989, the Pentagon and the Department of Energy invited three Iraqis to attend a "detonation conference" in Portland, Ore. Financed by American taxpayers, the meeting brought together experts from around the world to explain to the Iraqis and others how to produce shock waves in any desired configuration. There were even lectures on HMX, the high explosive of choice for nuclear detonation, and on flyer plates, devices that help produce the precise shock waves needed to ignite A-bombs. Both HMX and flyer plates have turned up at Al Atheer, which should surprise no one. The three Iraqis who attended the conference came from the laboratory that eventually provided Al Atheer with its first shaped charges.

To design a successful bomb, the Iraqis also needed computing power to solve the hydrodynamic equations that predict the behavior of shock waves. The inspectors discovered that Iraq was running the equations on a mainframe computer from the Japanese company NEC. Another Japanese firm, Hamamatsu, sold Iraq two "streak cameras," sensitive instruments that can photograph a high-speed shock wave as it implodes. The inspectors confiscated both cameras after determining that they were rapid enough for nuclear weapon work.

Altogether, the Iraqis carried out 20 detonation tests before May 31, 1990 — the date of the last Iraqi progress report on Al Atheer found by the United Nations. The Iraqis had worked their way through five versions of the bomb design, cutting the weapon's total weight from one ton in the first version to about half a ton in the last—light enough to go on a missile.

After May 1990 the Iraqis worked unimpeded at Al Atheer for eight more months. No one knows how much more they achieved. The Iraqis started relocating vital equipment before allied bombing began in January 1991, and as late as last summer tore out concrete floors to prevent inspectors from determining which machines were used there. They even ripped out electrical hookups to hide power usage. Now that Al Atheer is "sanitized," inspectors fear the bomb work has moved elsewhere.

Wherever the work is going on, the Iraqis still have plenty of equipment. During the late 1980's, Baghdad bought machines by the factory load, few of which have been found. The purchases included additional vacuum furnaces, from the German firm Leybold; plasma-coating machines, which could be modified to coat the surfaces of the molds into which molten uranium is poured, from the American company TI Coating; high-speed oscilloscopes, needed to develop firing circuits for nuclear weapons and for nuclear tests, from the American company Tektronix; and two X-ray diffraction systems, capable of analyzing weapon-grade uranium during production, from the German firm Siemens. TI Coating sold directly to an Iraqi factory charged with making A-bomb fuel; Tektronix sold to an Iraqi procurement agent for a string of nuclear and missile sites; Siemens sold to the Iraqi Ministry of Industry and Military Industrialization, which set up Al Atheer.

These purchases followed Iraq's policy of "parallel sourcing." The Iraqis never buy just one machine or build a single plant. If the item is important, they buy or build two. So if one vital machine or plant is bombed or surrendered to inspectors, they almost always have another.

The inspectors found out one other thing about the Iraqi bomb—it is highly unstable. The design calls for cramming so much weapon-grade uranium into the core, they say, that the bomb would inevitably be on the verge of going off — even while sitting on the workbench. "It could go off if a rifle bullet hit it," one inspector says, adding: "I wouldn't want to be around if it fell off the edge of this desk."

Even a "fizzle," when the bomb explodes too soon to get a full chain reaction, would be serious. The minimum blast effect would be equal to filling 20 semitrailers full of TNT, parking them side by side and setting them off simultaneously. The full yield would be like setting off 1,000 semitrailers' of TNT.

With a workable and mostly tested bomb design, Iraq faces only one more barrier: weapon-grade uranium fuel. Iraq started producing this precious substance before the war, but never got close to making enough for a bomb. Whether it finally succeeds will depend on its foreign suppliers.

The key will be the centrifuge. By spinning uranium gas at high speeds, centrifuges separate light, unstable uranium isotope that explodes in an atom bomb from the heavy, stable one that doesn't. A spinning tube called a rotor propels the heavy isotope to the outside wall and leaves the light one at the center. As the gas is run through a series of centrifuges called a cascade, the concentration of the light isotope is gradually raised from less than 1 percent in natural uranium to over 90 percent in uranium of nuclear weapon-grade. This technically demanding process is called enrichment.

Iraq's centrifuges are based on German designs and were built with German help. Iraq somehow got German blueprints in the 1980's. By 1988 it was already running experimental models. When one model developed a hitch in late 1988, Iraq summoned Bruno Stemmler, an ex-employee of M.A.N., the German company that makes centrifuges for the German national enrichment effort. After studying Iraq's illicit blueprints, Stemmler removed the hitch.

Iraq's next goal was mass production. It takes from 1,000 to 2,000 German-style centrifuges to produce a bomb's worth of enriched uranium each year. German firms again obliged. From H & H Metalform — a company subsidized by the German Government — came "flow forming" machines that are specially adapted to produce rotor tubes, the most difficult part of the centrifuge to make. From Leybold's American subsidiary came a giant electron beam welder, equipped with custom-made fixtures for welding the rotors to their necessary end caps. From Dr. Reutlinger & Sohne came machines to balance the rotors vertically and horizontally. From Neue Magdeburger came other specially adapted machine tools. And from Degussa came an oxidation furnace to treat the surfaces of parts so they could withstand corrosive uranium gas.

After surveying this glittering array, the U.N. inspectors concluded that Iraq would be able to produce more than 2,000 centrifuges a year, enough for a full-fledged bomb program. From a recent inspection, we know that Iraq ordered parts for 10,000 centrifuges, although it is not known how many parts were actually delivered, or how many centrifuges Iraq may have made.

The U.N. teams have now destroyed all the centrifuge parts it could find. But the inspectors don't know how many more centrifuge parts there are, because they don't know how many were sold to Iraq by Western companies. They are especially worried about a "missing cascade." They assume that Iraq would not have built a plant to mass-produce centrifuges without first being able to connect them in an experimental cascade. No cascade has been found. As the inspectors warn in their report, Iraq "may still have an undisclosed program."

The inspectors are also worried about a possible cache of weapon-grade uranium. Last July, they found four traces of this material in samples taken from Tuwaitha, Iraq's primary nuclear site. Because of the possibility that the samples were contaminated after they left Iraq, however, the evidence was not considered conclusive. New samples were taken in October, but the test results are still not in. Thus, the U.N. inspectors cannot pursue the lead.

There is also the matter of a hidden reactor. Western intelligence sources believe that the Iraqis have at least started to build one, but the inspectors have not been able to find it. Even a small, 20-to-40-megawatt reactor would be large enough to fuel a few nuclear weapons a year.

And, finally, the inspectors are worried about outside suppliers. They have concluded that to stop Saddam permanently there must be "strict maintenance of export controls by the industrial nations." But nothing in recent history suggests that the industrial nations will exercise such restraint.

In the five years before the Persian Gulf war, for example, the Commerce Department licensed more than $1.5 billion of strategically sensitive American exports to Iraq. Many were for direct delivery to nuclear weapon, chemical weapon and missile sites. Companies like Hewlett-Packard, Honeywell, International Computer Systems, Rockwell and Tektronix sold high-performance electronics either to Saad 16, Iraq's major missile research center; to the Ministry of Industry and Military Industrialization, which set up Al Atheer; to the Iraqi Atomic Energy Commission, responsible for atomic-bomb research; or to Nasr State Enterprise, in charge of Iraq's missile and nuclear procurement. Honeywell even did a feasibility study for a powerful gasoline bomb warhead, intended for an Iraqi-Egyptian missile.

The computer giant Sperry and its successor, Unisys, also benefitted. They got licenses to sell multimillion-dollar computers designed to handle a "personnel data base." The powerful machines—ordered by Iraq's Ministry of the Interior, which houses the secret police—are ideally suited to tracking and suppressing civilians.

The Commerce Department approved all these exports despite strong warnings from the Pentagon, the first coming in November 1986 concerning Saad 16. Commerce nevertheless permitted the sale of hundreds of thousands of dollars' worth of sensitive computers and electronics to Saad 16, all after the warning.

And there was the strange case of the Badr General Establishment, a factory outside Baghdad. In the summer of 1989 it wanted to buy a computer-controlled lathe from Cincinnati Milacron and a high-accuracy measuring system from Brown & Sharpe. Badr said the equipment would make "crankshafts, camshafts, and gears" for automobiles. But the Pentagon was skeptical. Commerce therefore agreed to a "pre-license check," in which an American official would actually visit the site.

After a 30-kilometer trip out from the capital, two embassy officials toured Badr with its production manager, Salam Fadl Hussain. The verdict was unanimous. The American Ambassador, April Glaspie, cabled the good news to Commerce on Sept. 13: “We believe that Badr General Establishment is a reliable recipient of sensitive United States origin technology and technical data." We now know that Badr and another organization were jointly in charge of all the centrifuge production in Iraq.

As bad as the American record is, Germany's is worse. Germany supplied more of Iraq's mass-destruction machinery than all other countries combined. Germany not only sold Iraq most of its centrifuge equipment, it also furnished an entire chemical weapon industry and was Iraq's greatest supplier of missile technology, including a flood of parts that enabled Iraq to extend the range of its Scud missiles. During the Persian Gulf war, enhanced Scuds hit Tel Aviv and a United States Army barracks in Saudi Arabia, killing 28 sleeping soldiers.

To develop an even longer-range missile, Iraq turned to the German armament giant Messerschmitt, now doing business as MBB (Messerschmitt-Bolkow-Blohm). MBB supplied the know-how for a 600-mile nuclear-capable missile called the Condor II that Iraq tried to develop jointly with Egypt and Argentina before the war. The missile's range and configuration are similar to that of the American Pershing, which MBB worked on at the Pentagon. The same MBB employee who worked on the Pershing at the Pentagon also represented MBB in Iraq for the Condor, and thus was in a position tn transfer American missile technology to Baghdad.

Since the inspections began, critics have questioned whether civilian volunteers working under United Nations auspices could eradicate Iraq's weapons of mass destruction. That question has now been answered. Despite great courage and enthusiasm, the inspectors still have not found the hundreds of Scud missiles Iraq is known to be hiding, or the headquarters of the centrifuge program, or exposed the supplier network. Nor have they solved the mystery of the weapon-grade uranium. Obviously, stronger methods are needed.

First, the United Nations has to change tactics. "We have diplomats when we should have detectives," says a knowledgeable United States official. "It's like looking for an escaped murderer. You question everybody who might have a lead and you keep on asking until you get answers."

In other words, shift to police-style investigations. Only the Iraqis know where their nuclear treasure is buried; only they can reveal it. To make headway, the United Nations will have to deploy inspectors by the hundreds, station them in Iraq instead of New York, and use soldiers as well as civilians. The inspectors must be free to interrogate every Iraqi scientist or engineer who might have relevant information and to follow up the leads immediately. And they must have the power to push aside Saddam Hussein's disinformation specialists.

The inspectors also need to know exactly what Iraq has bought. So far, though, not a single country has been willing to tell the inspectors what its companies sold. Only Germany has provided leads, and when it did, the inspectors quickly turned up centrifuge parts. As long as other suppliers sit on their export data, the inspections will be reduced to fishing expeditions, with the Iraqis steering the boat.

The United Nations must also put its own house in order. While the Special Commission has run the missile and chemical inspections with great zeal, the nuclear inspections are assigned to the International Atomic Energy Agency, the equivalent of an oxcart with its brakes on.

In late September, for example, the agency seized more than 60,000 pages of Iraqi documents, many of them describing the supplier network. Five months later practically no translations have been done.

The agency is also timid about destroying illicit equipment. While the Special Commission is destroying every machine it can find that the Iraqis bought, built or used to make chemical weapons or missiles, the Atomic Energy Agency has been willing to destroy only small parts of the machines used to make nuclear weapons. For example, Iraq bought a giant electron beam welder to fabricate centrifuges, but the agency destroyed only the small fixture that holds the centrifuge in place, leaving the giant welder intact. This means that if the Iraqis have extra fixtures — which is likely, given their parallel sourcing plan—they can go back into the bomb business with the same machines.

Assuming the United Nations does manage to eradicate Iraq's nuclear, chemical and missile programs, it still faces the problem of preventing Baghdad from starting over. One solution is to expose Iraq's supplier network, which is still intact. The United Nations has compiled lists of the companies in the network and what they sold, but it has furnished them only to the involved governments. The United States is reported to favor making the lists public, but Germany and France are said to be resisting.

Another way to defeat the network is to toughen export laws. Most of what Saddam bought was licensed. Governments knew he was getting dangerous equipment but hated to see their companies lose a sale. The resulting debacle should have taught the world a lesson, but Western export controls are no stronger now than they were before the gulf war. In fact, with the end of the cold war, the NATO countries and the European Community have been easing export controls.

The outcome in Iraq is now in the hands of President Bush and his gulf war allies. If they are willing to turn the United Nations into a vehicle for curbing the spread of the bomb, the battle in Iraq can still be won. If not, Iraq's bomb makers will pick up where they left off, and the new world order will fail its first important test.