Donnelly tried to get confirmation of the problem from EECOM. “You see a Bus undervolt, EECOM?”

“…Negative, Flight.”

But EECOM had hesitated.

He knows more than he’s telling me. He’s still trying to figure it through. What the hell was happening? The mission seemed to be falling apart before his eyes.

Donnelly pressed EECOM again; he needed more data. “The crew is still reporting the undervolt, EECOM.”

“Okay, Flight. I have some instrumentation problems. Let me add them up.”

Instrumentation problems. EECOM sees the undervolt, all right. But he doesn’t believe what the instruments are telling him. He’s looking at a lot of ratty data, he thinks some kind of major telemetry failure is under way. He wants to be sure before he reports it to me.

Donnelly said, “I assume you’ve called in your backup EECOM to see if we can get more intelligence applied here.”

“We have him here.”

“Roger.”

Then INCO, the instrumentation and communications controller, called in. “Flight, INCO. The high-gain antenna has switched to high beam.”

What in hell did that mean? “INCO, can you confirm the time when that change occurred?” If he could, it might be a clue in pinning down what was happening…

Before INCO could reply there was another call. “Flight, Guidance. We have attitude changes.”

“What do you mean, attitude changes?”

“The RCS valves appear to be closed. They should be open.”

Reaction control problems. Antenna problems. Problems with the oxygen tanks, and the fuel cells.

He’d never seen a systems signature like this before, not in any of the sims he’d gone through. But then, even after twelve years of flights, Apollo-Saturn was still an experimental system. You’d test-fly an airplane far more times than any spacecraft had ever flown, before declaring it operational.

So what was hitting him? It could be instrumentation problems, flaky readouts, as EECOM seemed to suspect. Or it could be that the Service Module had blown out, knocking the whole stack sideways. Or something else might have blown, and damaged the Service Module.

INCO’s timing came in. His antenna problems dated from a few seconds after they’d lit the NERVA.

For the first time in several seconds Donnelly glanced at the trajectory plot board. The spacecraft was diverging, markedly, from the path it should have followed, had the NERVA been burning smoothly.

The S-NB looked to have shut down.

“Guidance, you want to confirm that deviation?”

“Rog, Flight.” Guidance was the ground navigator. Guidance must be looking at multiple problems, too, as the spacecraft drifted from its trajectory and tumbled away from its intended attitude.

“Booster, you have anything to report?”

Mike Conlig did not reply. Donnelly could see how he was hunched over his console. “Booster?”

York said, “The crew is reporting a smell of ozone, inside their helmets.”

“Flight, this is Surgeon. I have a contrary indication.” The flight doctor on this flight was a crop-headed Oklahoman sitting in the row in front of Donnelly, with the systems guys, at the left-hand end next to Natalie York. He was wearing a button badge which read FUCK IRAN. His voice was taut, urgent.

Donnelly switched him onto a closed loop. “Go, Surgeon.”

“Flight, I’m monitoring a surge of radiation flux through the spacecraft cabin. And some changes in the crew’s vital readings.”

Donnelly was thinking through York’s brief report. They can smell ozone. Oxygen, ionized by radiation. Radiation from the NERVA. Jesus Christ almighty.

It was real, then. Not just flaky instrumentation. And the Russians orbited a goddamn Vietnamese in Salyut this year. The press will crucify us.

Because of the two simultaneous missions in progress Bert Seger had been away from the office for three days, and he was taking a chance to work through his mail. He’d only been at it for a few minutes when he got a call on the squawk box, the line that linked up the senior staff in Building 2.

There had been some kind of problem with the Apollo-N flight, and Seger had better get on over to the MOCR.

Angrily, Seger folded up his mail. With the NERVA, it was one damn thing after another.

The voltage needle on Bus A sank past the bottom of its scale. More warning lights came on.

Dana checked the Service Module’s fuel cell 1, which was supposed to feed Bus A. It was dead. His gloved fingers clumsy with the switches, Dana began to reconnect the Command Module’s systems from Bus A onto Bus B.

Another red light came on. Bus B was losing voltage as well. He checked fuel cell 3, the feed for Bus B; it was dead, too.

Jesus. We’ve lost the Service Module. It’s Apollo 13 over again.

He made his report, trying to keep his voice level. Mary would be listening, probably the kids. “Okay, Houston, I tried to reset, and fuel cells 1 and 3 are both showing gray flags. I’ve gotten zip on the flows.”

“Acknowledged, Apollo-N. EECOM has copied.”

Earth, beautiful, unperturbed, drifted past the windows.

The spacecraft and booster had been set rotating by that mysterious bang. Dana knew the ship’s attitude control systems should have been trying to steady their slow tumble, but there was no sign of any correction.

“Chuck, I think the Service Module’s RCS must be out.”

“Rog,” Jones said. “Houston, we don’t have reaction control, either from the Service Module or from S-NB.”

If the Service Module had blown, it was the end of the mission. But still, the crew ought to be able to get home, from this low Earth orbit.

As the spacecraft rolled, a cloud of ice crystals, sparkling, dispersing, drifted past the window to his right. It seemed to be venting from somewhere in the stack. It was quite beautiful, coalescing above the shining face of Earth.

More alarms lit up, as the problems multiplied and spread.

Donnelly had Surgeon feeding radiation dosimeter readings into his ear on the closed loop.

EECOM said, “Flight, I want to throw a battery on Bus A and Bus B until we psyche out the anomalies. We’re confirming undervolts.”

Donnelly tried to shut out Surgeon’s voice so he could figure out EECOM’s suggestion.

EECOM wanted to run the Command Module off battery power. It was a reasonable short-term suggestion. But, looking ahead, the Command Module’s batteries would have to be conserved to allow the crew to reenter the Earth’s atmosphere. “What about limiting that to a single Bus, rather than both?”

“Hold on that, Flight.” EECOM would be conferring with his team of experts in the back rooms.

It was obvious from multiple indications, not least the crew’s report, that the NERVA had indeed shut itself down after only a few seconds of the planned burn. “Booster, you have anything you want to say to me?”

Conlig didn’t reply. The guy seemed to have frozen out.

“The crew’s health is going to be severely impacted,” Surgeon said on the closed loop. “Though they probably don’t know it yet. In fact, Flight, you can’t expect them to function normally for much beyond a few more minutes.”

Guidance came on line. “The bird’s attitude is still changing. They’ve got to stop it. We’re heading for gimbal lock.”

“I hear you, Guidance.”

Gimbal lock meant the spacecraft was tumbling beyond the tolerance of the inertial guidance system. The platform could be reset by eye again. But if Donnelly was forced to go for an emergency reentry, he needed alignment control immediately.

Somehow, though, he felt that alignment loss, even a gimbal lock, was the least of the spacecraft’s problems just then.

“Houston, Apollo-N.” It was Jim Dana; to Natalie York, Jim’s voice sounded thin, frail, but controlled. “We’re seeing some kind of gas, venting from the stack.”

York’s skin prickled with a sudden chill.

“Rog, Apollo-N,” she said. “Can you tell if it’s coming from the S-NB tank, or the Service Module?”