Apparently, as well as working on laser systems, the Livermore people have been working on defences as well. They've designed a highly reflective, collapsible mirror specifically for the shuttle. It's been rocked in a warehouse for some time. The shuttle swings this thing overboard with the manipulating boom and positions it to reflect any laser blast as they close in. Just how they immobilize the satellite to get it in the cargo bay and bring it home isn't clear to me.'

'Isn't it too big?' Isaacs wanted to know.

'In a sense, but the Soviets know how big the shuttle bay is. The satellite is basically the upper end of one of their big booster rockets.'

Isaacs nodded.

'Apparently, they added some external gew-gaws specifically designed to make the whole thing too large to fit in the cargo bay. The idea is that the crew should take a torch to it with a space walk, cut it up into manageable-size pieces. In principle it'll fit.'

'Great,' exclaimed Isaacs with irony. 'And when do they advise trying to attempt this insanity?'

'The next shuttle launch is in the middle of April, two weeks from now. That's what they're pushing for. The idea being, of course, to strike while the iron is lukewarm. They'd like to launch yesterday, but the shuttle isn't so flexible.'

'Madness! And they think the Soviets won't then blow away one of our communication link satellites, Comsat or some such thing?'

'The argument is that Cosmos 2112 is the only laser they have flying.'

'But we didn't know that until two days ago!'

'Tell that to mah buddy, the President.'

'How's he leaning?'

'I didn't get any feeling for that, third hand, but the brass is pushing hard. They've pumped a lot of dollars sideways into NASA for the shuttle. They want to play with their toy.'

'But they must have war-gamed this kind of thing.'

'I suppose it can be contained in some scenarios.'

'Yeah, in one per cent of them. Voice, we've got to convince our side about this meteorite, too. That seems to be the only sure way to show that the Soviets had some justification and that we don't need to retaliate.'

'You'll have to start in-house. Drefke will relay any report you write, but you know how his antennae are tuned to the White House. He's apt to take his cues from the President. And McMasters clearly won't be much help.'

'That's a fact,' Isaacs agreed. That was quite a show he put on the other day.'

'It was clearly his only tack. He had to really push the Russians as bad guys to keep Drefke from thinking too deeply about why FireEye was shifted in the first place. Now he's painted himself into a corner. He'll have trouble turning around and saying, well, maybe they're not so nasty after all, a little hasty with their death ray, but really not bad chaps.

'The other factor is,' Martinelli continued, 'that this meteorite idea and follow-up has to come from your group and his negative instincts won't allow him to embrace it with a lot of enthusiasm.' The two men sat in silence for a moment, then Martinelli rose.

'I'll go see Art; we'll try to get some dope on the channels this report went through.' Martinelli waved the document as a farewell gesture and paused.

'There's a bright side to all this, you know. If this trick with the shuttle backfires badly enough, we won't have to worry about getting our taxes done on time.'

'Thanks a lot, Vince.' Isaacs grinned at the black humour. 'Silver linings like that I can do without.'

Isaacs watched his friend shut the door. He began an outline of the questions to be addressed concerning the possible impact of a meteorite on the Novorossiisk. He would turn it over to his technical staff to flesh it out.

The preliminary report was already late the next day, a rush job to which some thirty people had contributed in an intense surge of effort. It looked pretty good, plausible enough for a first pass. There were some troubling points. A meteorite would progressively disintegrate as it passed through metal walls. To go all the way through the carrier, a meteorite would have to drill larger holes than had been reported in the upper decks, and the holes should get smaller in the lower decks. It was not clear-from the stolen Soviet report that that pattern was reproduced.

Isaacs downplayed such doubts in working over the final draft. He wanted to make as much impact as possible to forestall a decision to go after Cosmos 2112 with the shuttle. He relied on the state of emergency to go out of channels and took the report directly to Drefke. The Director was clearly impressed with the idea. Isaacs knew he would then show it to McMasters, but by then the original impact would have had its maximum effect. He would get the most positive response possible when Drefke in turn reported to the National Security Council and the President.

Korolev stirred at his desk, reached up and punched off the button on the neck of the gooseneck lamp, leaving the room to share the deepening light of dusk. He rose and moved to the window. From this upper floor of the Academy of Sciences building he could see a stretch of lights now winking on over Moscow. For years, no, decades now, he had stood at this window watching those lights at odd hours of the night as he contemplated some problem. How many there had been. Practical earth-shattering problems imposed by the voracious military: explosions, implosions, shock waves, the bomb. Later, intense radiation, hyper velocities, directed energy weapons. What did the Americans call them? Buck Rogers stuff. Lovely, basic problems. Microscopic, the innards of particles, and the innards of those in turn, and then of those. Cosmological problems, the wondrous workings of Einstein's mind on vast scales.

Tonight, a small but troubling problem. Some American was quick and thoughtful. He could see the mental play behind the words. Yes, the suggestion of a meteorite was bold, for all its obviousness. It was one of the first which had occurred to him as well. The author of this report had pushed it for all its worth, but he also knew the limitations. Korolev could read between the lines and see where the American had suppressed his reservations. What the American did not know were the results of the follow-up report which had come directly to him. The punctures were all wrong for a meteorite with enough impact to penetrate the carrier decks. There was no downward flaring, the holes looked drilled, not punched. They had done a metallurgical test: there was no meteorite material. The Americans had not yet stolen that report. It was no meteorite.

Although there were features that did not fit, a lack of heat searing, for instance, Korolev had been compelled to state that a beam weapon seemed the most plausible explanation. His superiors had demanded some hypothesis and he could think of no other. He had not anticipated that they would mistrust their intelligence so badly as to suspect that the Americans had leap-frogged them and orbited such a weapon.

What troubled him, beyond the still unexplained nature of the Novorossiisk event, was the sincerity in this report. He was convinced that the author would eventually come o the conclusion that a meteorite could not be involved, but this report was not a sham. The author pushed the meteor idea too strongly because he wanted it to be true. The whole tone told Korolev that the report was based on the secure knowledge of the author that the Americans were not involved with the Novorossiisk. That was the trouble. His government knew he had already considered and rejected the meteor hypothesis. They would reject the suggestion by the Americans. Could he convince them of the Americans' uninvolvement with the Novorossiisk based not on the contents, but on his sense of the motivation of the report on his desk? There would be much resistance. They were convinced the Americans were involved, somehow, and now there was the irrevocable act of the destruction of the American spy satellite. Korolev continued to stare out over the streets until the dusk faded to deepest black.

The first half of April slipped away as Isaacs spent two hard weeks probing the meteor theory. He called in projectile experts from around the country, and his top people visited various test sites. The harder they worked, the less likely the idea seemed. Boswank had traced the Novorossiisk report to one of the most respected members of the Soviet Academy of Sciences, Academician Viktor Korolev. That seemed a positive note: his reputation as a profound and unprejudiced thinker was well-established. Then at the end of the first week came the curt Soviet reply to Isaacs's report. A meteor had been previously considered and rejected. With Korolev's reputation behind that statement, and the increasingly negative results of his own team's study, Isaacs knew he was losing any power to influence events. To make matters worse, the Soviet reply was defensive and belligerent. It yielded no hint that they conceded the innocence of the Americans in the Novorossiisk affair, certainly no confession that they might have mistakenly overreacted m the destruction of the FireEye. The only good news was that with the act of retribution the Yellow Alert had been cancelled. The Backfire aircraft were returned to normal routine; the missiles recapped snugly; the troops redeployed.

Isaacs walked slowly down the hall from Drefke's office and punched the elevator button. The Director had just returned from the meeting of the National Security Council. Isaacs had read the result on his face. The shuttle was going up tomorrow. The crew was to disable the Cosmos 2112 and bring it back in the cargo bay. Or fry trying.

He got off at his floor and continued his thoughtful pace. He opened the outer door to Kathleen's office and was surprised to see Pat Danielson sitting there with an expectant smile and a pile of computer output and charts on her lap. The smile faded when she saw the-heavy cloud on Isaacs's face.

'Is this a bad time?'

She detected Isaacs's quick visible effort to compose himself. His voice had a forced heartiness.

'Not at all.' He smiled ruefully. 'No worse than any other time. You have something important?'

She glanced down at the bundle of paper clutched possessively on her lap, and her voice carried an overtone of excitement.

'I think you're going to find your curiosity about this seismic signal justified.'

Isaacs had to think for a second to recall what she was fouling about. He was too preoccupied with the historical clash scheduled to take place over their heads tomorrow to give much attention to the task he had assigned her, but the little wheels had to be greased, just like the big ones. A little investment of his time would keep Danielson performing efficiently.

He crossed to the door to his office and held it open in welcome as she rose and hustled through. She deposited the material on his desk and took the chair across from him.

'It wasn't just a transient then?' he asked.

'On the contrary, the more we learn about it, the longer we can trace it back through the earlier data — several months' worth now.' She pointed to the stack of paper. 'Here's the latest output, hot off the printer.'

He gestured outward with both hands, palms up, encompassing the output and the young woman.

'Shoot,' he said, striving to concentrate on what she had to say.

'With a longer time base, more information becomes available. At first all one could tell was that the signal repeated itself. We had only a crude idea of the period and no notion of the location. We've worked very hard to obtain a better estimate of the period. The figure of an hour was an alias. The true period is somewhat less than ninety minutes. This update shows that we're beginning to get a handle on the location. Would you care to guess?'

Danielson did not usually play such little games, but came straight out with the facts. She thinks there's something special here, thought Isaacs. Aloud he said, 'Undoubtedly, it's coming directly from the situation room in the Kremlin.'

'Wrong, of course,' smiled the young woman. She turned serious. 'But you've hit on an important point. The first algorithms used in the signal analysis were based on the assumption of a static source, that the signal was coming from a single location. That assumption proved to be self-inconsistent and we abandoned it. When we allowed for the possibility that the source moved, things began to fall in place.

'I won't show you all the data, but look at these two clear stretches when the background noise was low.' Danielson unrolled a strip chart on Isaacs's desk. 'See here, the signal comes from the vicinity of Egypt. Here, this is a week later, it comes from the mid-Pacific basin. That proves it moves. A more careful analysis hints, but doesn't yet prove, that the period is not due to a change in power at the source, but is due to a source of roughly constant power moving from one side of the earth to the other.'

'A reflected wave of some kind,' put in Isaacs.

'Perhaps,' replied Danielson, 'but not like any the seismologists have ever seen before. Any strong earthquake will set up reverberations which travel diagonally through the earth, but those die out quickly. Something continues to drive this wave — that's the mystery.'

'So the actual energizing source might still be located in one place and the apparent movement is just due to the random bouncing of the subsequent wave.'

'Possible,' allowed Danielson, 'and more comfortable, but the data still seem to suggest that the source is moving.'

'How much energy is involved?' queried Isaacs.

'Well, of course, the power we detect depends on both the power at the source and the distance to our detectors. If we assume the source is, on the average, at the distance of one earth radius, about four thousand miles, then the seismic energy flux at the detector corresponds to a source power of about one thousand megawatts — big for a power station, but pretty small potatoes compared with all the seismic energy in the earth at a given time. Which is why the signal is hard to detect and analyse.

'Since we don't really know the nature of the source, it's difficult to associate an energy with it; that is, it could sit in one place and emit bursts of energy that reverberate, or it could represent a continuous supply of energy, as we believe. A ballpark estimate is the total energy liberated in one characteristic period, ninety minutes. In one period that would be about one per cent of the energy of a one kiloton nuclear event.'

'That's a maximum estimate, isn't it?' asked Isaacs.

'Yes, sir,' replied Danielson, 'within a factor of a few, given that the source is confined to the earth.'

'One hundredth of a kiloton,' mused Isaacs. 'That's too small to be a nuclear device, and if the source is closer, the energy estimate only goes down. Still, if that amount of energy is being liberated artificially on the surface, we should be able to see other signs of it in the optical or infrared — somewhere.

'The most reasonable assumption,' Isaacs continued, 'is that this is some natural seismic event which happens to have a period of about an hour and a half, regular fault slippage of some kind.'

Danielson raised a finger and opened her mouth to interject, but Isaacs interrupted her, 'Unless, of course, you can prove the source is actually moving about.

'Obviously, I'm unconvinced this signal is anything but some sort of natural phenomenon,' Isaacs said, 'but I am convinced we need to nail it down. Suppose you're right and it's not related to natural fault slippage somewhere, do you have any guess as to what it might be?'

'No. If the source is moving around in the earth as I think the data suggests, it's a total paradox. Fault slippage at different points on the earth shouldn't be correlated.'

Isaacs leaned back in his chair, toying with a pencil. 'A period of ninety minutes still sounds suspiciously like some artificial phenomenon — keyed to somebody's time clock. If your positions are right, Egypt and whatnot, it's not a local man-made thing, but I'd like to make sure that is ironclad.'

Isaacs sat up at the desk and gestured to Danielson with the pencil. 'You had better make this a matter of some priority until it's resolved. We need to know the period, if it really is one, more accurately. If the period is not precisely defined, that's good evidence of a natural phenomenon. If the period turns out to be exactly ninety minutes, it will be a man-made event despite present evidence to the contrary.

'We need to know the location, whether or not it is moving around. When you have a location, we can look for some other evidence of its existence and nature. If it's seismic in nature, there should be some correlation with fault location and activity. Any other suggestions?'

Danielson paused a moment in concentration before she spoke. 'No sense speculating without more data. It will probably be useful to get records from civilian seismic stations, universities here and abroad. We can look for correlations among events that would pass unnoticed in any single record. That should help with both the period and the location.'

'That's fine,' said Isaacs with a note of finality. 'Let me know how this develops.'

'Right,' said Danielson, rising to leave, collecting the bundle from his desk. 'We'll continue to monitor our own AFTAC data, and that may begin to pin things down. But it will take a month or so to acquire and analyse the civilian records.'

'Okay, keep in touch.'

'Yes, sir.'

Isaacs watched the door close behind her. He stared at it, unseeing, as her problem diffused from his mind and his consciousness flowed out along tangled diplomatic channels. From his office to Drefke's to the White House. To Moscow. Academician Korolev. Why did he rule out the meteorite? What had happened to the Novorossiisk? What would happen to the shuttle?