Dan shakes his head. “Won’t work. The missiles are too hard. We’ve got to burn them out, just like I used to say we’d have to, back before we got this development contract. The mirrors are up there and they won’t be getting any bigger, the power pulse is already incredible when you think that over a hundred and fifty laser stations will have to be supplied all at once, and we can’t change the wavelength of the lasers without replacing the entire systems. And that’s the whole ball game.”

“But that means that dwell time is longer than boost time!”

“That’s right. Each laser can bring down about eight-tenths of a missile. And there’s a hundred and fifty laser stations, and about ten thousand missiles.”

McPherson feels himself gaping. He takes the pen from Houston, starts writing on the tablecloth himself. He surveys the figures. Takes another drink.

“So,” he says, “how did we get this development contract, then?”

Dan shakes his head. Now he’s looking out the window at the sea.

Slowly he says, “We got the contract for Ball Lightning by proving we could destroy a stationary hardened target in ground tests, with the sudden pulse shock wave. They gave us the contract on that basis, and we were put in competition with Boeing who got the same contract, and after three years we have to show we can do it in boost phase, in real-time tests. It’s getting close to time for the head-to-head tests. The winner gets a twenty-billion-dollar project, just for starters, and the loser is out a few hundred million in proposal and development costs. Maybe it’ll get a follower’s subcontract with the winner, but that won’t amount to much.”

McPherson nods impatiently. “But if we could do it on the ground?”

Dan polishes off another glass in one swallow. “You want another pitcher?”

“No.”

He pours foam and ice into his glass. “The problem,” he says carefully, “is that the test wasn’t real. It was a strapped chicken.”

“What?” McPherson sits up so fast his knee knocks the table and almost tips his glass over. “What’s this?”

But it’s clear what Dan means. The test results didn’t mean what LSR said they did.

“Why?”

Dan shrugs. “We were out of time. And we thought we had the problem licked. We thought we could send a beam so bright that it would create a shock wave in even the hardened skins, the calculations made it look like all we needed was a little more power and the brightness would be there. So we simulated what would happen when we did solve the problems, and figured we could validate the tests retroactively, after we had the contract. But we’ve never been able to.” He stares at the table, unable to meet McPherson’s gaze.

“For God’s sake,” Dennis says. He can’t get over it.