Lee went back to Newport steaming — and worried. If Hughes wouldn’t take their bid seriously, who the hell would? And not having a partner could leave a big hole right in the middle of the bid.

But the more he thought about it, the more he began to see that he might be able to turn this, too, from a weakness to a strength.

“Look at it this way,” he said to Art Cane. “Fuck ’em. Fuck Hughes. And the rest. We’ll go it alone. We’ll go to NASA as a coach, not a team. That’s what they need. Once you have Columbia as the coach, you can put anybody you like on the goddamn team. Let the customer make the choice when they’re ready; we don’t need to lock them in now.”

Art Cane shook his head. “You’re a crazy man, JK. Get out of my office.”

When the canteen had been converted, Lee set himself up at a desk on a stage. He appointed team leaders, his senior and smarter people like Bob Rowen and Julie Lye. But all the time he himself was going to hover over the whole thing like a hawk, looking out for trouble.

Rapidly, the concept of the Columbia MEM began to emerge.

Lee wanted something that would look, to the NASA evaluation panel, easy to build. So the MEM started out pretty much as Lee had outlined it that night he’d driven out to the Mojave with Ralph Gershon. You had your basic Apollo Command Module shape, a squat cone thirty feet tall, with a heat shield all of thirty feet across at the base. And his teams of engineers focused on that concept. Inside the conical, heat-resistant shell, the MEM would have two stages, like Apollo’s Lunar Module: a descent stage for landing on Mars, which would later serve as a platform for an upper ascent stage, which would return to orbit.

Lee laid down rules about how he wanted as little innovation as possible. “You can start being creative when we’ve won the goddamn bid, not before.” For instance, he didn’t want to see any changes to the overall Command Module shape. The angle of that cone had been determined years back by wind-tunnel tests at NASA’s own Ames Laboratory, and all the experience of Apollo had proved out that analysis since then, and he wasn’t about to let anyone in his organization challenge it.

In Lee’s first sketch, the MEM would land on five fold-down landing legs. Five was chosen in case one of the legs broke off on landing; the MEM could stay stable even if that happened. The descent-stage engine, the rocket which would carry the MEM down the last few miles to the surface of Mars, stuck out of the base of the MEM, with its fuel tanks clustered close by.