“All right,” I said in a mean tone of voice. “You’d better give me the whole story here, all the details of your model, your new improvements that Swann told me about, everything. If you want me to help.”

The two women nodded politely, as if this request were the most ordinary thing in the world. And we got down to it.

So I helped them, yes, I did. And more than ever before, the I who thought and felt was distanced from the I who did the work on this particular example of the BLSS problem — more than ever the work seemed a game, a giant intricate puzzle that we would look at when we finished — we would stand back to look at it, and admire it, and then we would forget it and go home to dinner. In this frame of mind I was especially inventive, and I helped a lot.

It got to the point where I even began to return to the starship in the evenings after dinner, to wander the farm alone and type some figures into the model programs to check the results. Because they had a real problem on their hands — I’d never worked on a harder one. The two ships were Deimos PRs: about forty years old, shaped like decks of cards, just over a kilometer long; powered by cesium reactor-mass, deuterium-fueled, direct-explosion rockets. The crew of forty or forty-five lived in the forward or upper part of the ships, behind the bridges. Below them were the recreational facilities, the various chambers of the farms, and the recycling plants, and below those were the huge masses of the rocket systems, and the shield that protected the crews from them. The ships were biogeocenoses, that is, enclosed ecologic systems, combining biologic and technologic methods to create closures. Total closure was not possible, of course; it approached eighty percent complete for a three-year period, tailing off rapidly after that. So they were good asteroid miners, they really were. But there were loss-points that had never been satisfactorily solved, and although these were the best closed biologic life-support systems ever built, they were no starships.

I walked in circles through the rooms of Hidalgo’s farm, following the course of the various processes as I tried to think my way through the system. Most of the rooms were darkened, but the algae rooms still required sunglasses. Here the whole thing began. Heat and light generated by the nuclear reactions in the rocketry provided energy for the photoautotropic plants, mostly the algae chlorella pyrenoidosa and chlorella vulgaris. These were suspended in large bottles under the lights, and I thought that, despite the nutrient problems, they could be manipulated genetically or environmentally to make the gas exchange as needed. I took off the sunglasses and stumbled around the darkened aqua room until my sight returned. Here the excess algae was brought to feed the bottom of the food chain. Plankton and crustacea ate algae, little fish ate the plankton, big fish ate the little fish. It was the same in the barns farther along; under night lights I could make out the cages and pens for the rabbits, chickens, pigs and goats — and my nose confirmed their presence. These animals ate the plant wastes that humans didn’t use, and provided food themselves. Beyond the animals’ barn was the series of rooms planted with rows of vegetables — the farm proper — and here some lights were still on, providing a pleasant, mild illumination. I sat down against one wall and looked at a long row of cabbages. Beside me on the wall was drawn a simple schematic, left wordless like a religious token — a diagram of the system’s circular processes. Light fed algae. Algae fed plants and fish. Plants fed animals and humans, and created oxygen and water. Animals fed humans, and humans and animals created wastes which sustained microorganisms that mineralized the wastes (to an extent), making it possible to plow them back into the plants’ soil.

The cabbages glowed in the dim light like rows of brains, working on the problem with me. The circle made by the diagram, supplemented by physiochemical operations to aid the gas exchange and the use of wastes, was nearly closed: a neat, reliable, artificial biogeocenosis. But there were two major loss points that had me stumped; and I wasn’t going to see the solution walking around the farm. One was the incomplete use of wastes. Direct use of human waste products as nutrients for plant life is limited by the build-up of chlorine ions not used by plants. Sodium chloride, for instance, is a compound used by human beings as a palatable substance, but it isn’t required in equivalent amounts by the other components of the system. So the use of algae to mineralize wastes on Hidalgo had to be supplemented by physiochemical mineralization — thermal combustion in this case, which resulted in a small but significant amount of useless furnace ash. It would be difficult to find ways to return those poorly soluble metal oxides into the system.

The other major problem was the very minute disappearance of water. Though water could be filtered out of the air, and recaptured in a number of ways, a certain percentage would coat the interior of the ship, bond with various surfaces, pool in cracks and hidden spots on the floor, and even escape the ship if they ever had EVA.

And the more I thought of it, the more little problems appeared to augment these larger ones, and all of the problems impinged on each other, making a large and interconnected web of cause and effect, mostly measurable, but sometimes not… the game. The hardest game. And this time, by these people, played for keeps.

I got up nervously and paced between the long soil strips. They could create water using a fuel cell and electrolysis. With the power plant they had along, that might be all they would need. It would depend on their water recovery, their fuel supply, the amount of time they spent between stars. I turned and headed for the farm computers, intent on trying some figures. And as for those wastes, Marie-Anne had spoken of new mutant bacteria to mineralize them, bacteria that could chew up the metals they would slowly be piling up outside the system…

The whoosh of the vents, the clicking of a counter, the soft snuffling of the animals in their sleep. Maybe they could do it, I thought. A very high degree of closure might be possible. But the question was, once accomplished, would they want to live inside it?

How long could humans live in a spaceship?

How long would they have to?

One morning after a night like that there was a knock at my door, and I opened it. It was Davydov.

“Yes?” I said.

He ducked his head. “I’m sorry about the way I behaved during our talk. It’s been so long since I’ve gotten any criticism of the project, I’d forgotten how to react to it. I guess I lost my temper.” Head raised, a shy little smile — Forgive me? Forgive me for kidnapping you and then yelling at you to boot?

“Umm,” I said cautiously. “I see.”

The smile disappeared, he pulled at his swarthy cheeks with one hand. “Could I perhaps, um, take you on a tour of the starship? Show you what we plan to do?”

I stood thoughtfully for as long as I could, knowing that I would accept the offer, curious to see what they had managed to steal from the Committee. “I suppose,” I said.

I saw from the boat’s dome, during our crossing, that they had finished connecting the two ships, with thin struts that held them side by side, and contained narrow passageways. It was one fat and ungainly-looking starship. Its windows gleamed like the luminous patches of ocean-floor fish. We were still in a tiny cluster of asteroids. The big one, I had learned, was Hilda; and around it were several daughter rocks.

It took Davydov several hours to show me what they had. They had: ore-holds full of minerals, medical supplies, foodstuffs, spices, clothing, equipment for planetfall, color panels and other material for the seasonal changes; a microfiche library of forty million volumes in three hundred languages; an equally vast collection of recorded music, with several each of almost every musical instrument; sports equipment; a lot of movies in English and Russian; a nursery full of toys and games; a room full of computers and computer parts; an observatory with several large telescopes.