Then there were the inconsistencies that Steinfield had talked about. . . . An absurd idea came from nowhere, which Hunt rejected impatiently. But as the night wore on, it kept coming back again with growing insistence. Over breakfast he decided that he had to know the story that lay below those billions of tons of rubble. There had to be some way of extracting enough information to reconstruct the characteristics of the surface just before the bombardment commenced. He put the question to Steinfield later on that morning, back in the lab.

Steinfield shook his head firmly. "We tried for over a year to make a picture like that. We had twelve programmers working on it. They got nowhere. It's too much of a mess down there—all ploughed up. All you get is garbage."

"How about a partial picture?" Hunt persisted. "Is there any way that a contour map could be calculated, showing just the distribution of radiation sources immediately prior to the bombardment?"

"We tried that, too. You do get a degree of statistical clustering, yes. But there's no way we could tell where each individual sample was when it got irradiated. They would have been thrown miles by the impacts; a lot of them would have been bounced all over the place by repeat impacts. Nobody ever built a computer that could unscramble all that entropy. You're up against the second law of thermodynamics; if you ever built one, it wouldn't be a computer at all—it would be a refrigerator."

"What about a chemical approach? What techniques are available that might reveal where the prebombardment craters were? Could their 'ghosts' still be detected a thousand feet below the surface?"

"No way!"

"There has to be some way of reconstructing what the surface used to look like."

"Did you ever try reconstructing a cow from a truckload of hamburger?"

They talked about it for another two days and into the nights at Steinfield's home and Hunt's hotel. Hunt told Steinfield why he needed the information. Steinfield told Hunt he was crazy. Then one morning, back at the laboratory, Hunt exclaimed, "The Nearside exceptions!"

"Huh?"

"The Nearside craters that date from the time of the storm. Some of them could be right from the beginning of it."

"So?"

"They didn't get buried like the first craters on Farside. They're intact."

"Sure—but they won't tell us anything new. They're from recent impacts, same as everything that's on the surface of Farside."

"But you said some of them showed radiation anomalies. That's just what I want to know more about."

"But nobody ever found any suggestions of what you're talking about."

"Maybe they weren't looking for the right things. They never had any reason to."