"I don't mind putting in an extra shift-"

"Thanks, but it isn't necessary. We've done a good day's work. Call it quits and we'll hit it again tomorrow. We can ready the equipment for the session tomorrow evening then. Good day, Tickler."

Tickler peered back as if he were trying to read a message that was written on Spence's face in a foreign language. "Are you sure there's nothing else?"

Spence shook his head and smiled as broadly as he could. "You sure are a workhorse. No, I can't think of a thing that can't be done tomorrow. You're free. I'll see you tomorrow."

Tickler did not reply; he only dipped his head in a smarmy little bow and then hurried away like a rat heading back to his burrow after a night in the pantry. Spence watched him go and then went to the portal himself. He cleared the access code on the doorplate and reset it with a new code so that he would not be disturbed.

"Now to business!" he muttered to himself as he sank back into his cav chair behind the console. Throughout the day as he worked, the thought kept nagging him that he should check out the riddle of the identical scan more thoroughly. Actually, the urge was not a new one-it had nagged him before, but he simply had not had time to do anything about it until now.

He fell to with a will. He retraced in his mind the steps he had followed to discover the similarity of the two scans in the first place. As to what the significance of the supposed similarity could mean, he was still at a loss for an answer. But deep inside he believed it to be important in some way. What he proposed to accomplish next was to establish that it had been no glitch, no momentary foul-up in the electronic circuitry or in the program which had fed him spurious information.

Spence picked up at the point where he had made his strange discovery three days before.

"MIRA, Spence Reston here. Ready for command."

"Ready, Dr. Reston," said MIRA's feminine voice.

"Compare all PSG Seven Series LTST entries. Display entries with similarities with less than one percent variability."

He sat back to wait, tapping his fingers on the table before him while MIRA worked. MIRA-the initials stood for Multiple Integrated Rational something or other which he could not at the moment remember-was the largest of a breed of biotic computers whose circuitry was in part derived from organic moleculesprotein grids which had been integrated with electronics. She was faster, smarter, and more creative in a dozen ways than any computer before her.

Within seconds the wafer screen spelled out the message, which to Spence's grim satisfaction matched the previous one: PSG Seven Series scans 3/20 and 5/15.

There it was again. The chance that it was a computer error ceased to be a possibility. Glitches did not repeat themselves. The chance that it was a kink in his program was also remote. The command was well within the program's range of flexibility.

Retracing his steps completely, he opened the yellow log book and matched the two disputed scans. They were, as he had previously discovered, quite different.

Next he pushed the inquiry a step further and went to the cabinet, getting out the tray of spools for the week of 5/15 and the tray for the week of 3/20. He set the trays down on his nearby desk and fished out the spools in question. He snapped the seal on each of them and rolled out a portion of the scan. The four red wavy lines undulated evenly across his desk. He matched up two intervals and placed one tape over the other and held them to the light.

The two scans, viewed one through the other as they overlapped one another, were clearly different. He could see peaks in one where there were valleys in the other. Laid one on top of the other all similarities between them ceased to exist. He checked the interval again and even tried to force the comparison by matching peaks and valleys, but could not. The scans were simply quite different one from the other. MIRA had apparently goofed after all.

But there was still one more wrinkle to check: the bubble memory. As an added backup to the overall design of the project, Spence had recorded each scan on a bubble plate. This was the source of the numbers entered in the log book. The rising and falling motion of the scanner's red ink lines was recorded within the thin sealed cartridge whose magnetic bubbles were interpreted by the computer as a continuous series of numbers. For every place the needle rested on the paper tape, there was a corresponding number. By reading the numerical values the computer could reconstruct the wavy lines on the paper tape.

He opened the bubble file and pulled the cartridge for the two sessions. He popped one cartridge into each of the slots in the memory reader of the console and gave the display command.

Instantly the numbers on the plates began filling the screen. He quickly scanned the columns and his breath caught in his throat; the two scans were exactly alike!

He dropped into his swivel chair and propped his feet up on the edge of the table. He stared at the rows of identical numbers on the screen and then closed his eyes, retreating into thought.

Here at last was the corroborating evidence he had been seeking-only instead of helping to solve the mystery, it deepened it. He began to think through the steps of his experiment and how it was recorded in all its various stages to determine how a situation such as the one glimmering at him from the wafer screen could ever have happened.

Given the fact that it was impossible for any two scans to be perfectly alike-even the same man on the same night could not produce two identical scans-he was forced to reckon the evidence an error, either human or electronic.

Now, with the evidence of the bubble memory, the likelihood of an electronic error diminished to the point of infinite improbability. The cloud of doubt in which he had so far carried his investigation began to condense into suspicion: someone had been tampering with his records.

The longer he thought about it, the more suspicious he became until the unproven hypothesis hardened into certainty. Someone had been tampering with his materials. Assuming that much, the next question was why? Why would anyone want to sabotage his experiment?

No, that was the wrong approach. Not sabotage-alter. That seemed closer to the mark. Why would anyone want to alter the evidence? And why these particular scans, in this particular way?

To puzzle this latest wrinkle in this confusing development he got up from his chair and shoved it across the room. He began pacing with his arms folded across his chest and his head bent down as if he expected the answer to form itself upon the floor.

The answer, when it came, hit him like a closed fist between the eyes; it nearly knocked him down.

The simplicity of it staggered him-it was so obvious. The scans had not been altered; they had been duplicated. The scan of 5/15 was a copy of scan 3/20. That was why they were identi cal. What about the other pieces of the puzzle? The tape, the log book, the main computer memory? Those simply had been man ufactured to fill in the gap. too Spence's mind raced ahead like lightning along a oncetraveled path.

The morning of 5/15 had been the morning after his first blackout when he awakened in the sleep chamber. That much he remembered clearly. He remembered Tickler remarking that the scan had gone well that night. He also remembered that he had not actually seen the scan at that time; it was not until after breakfast that he examined it. Plenty of time for someone to manufacture the missing pieces and place them in position.

Was the scan of 3/20 somehow significant? Probably not. It had just been selected at random from among the first of the experiment's records. It was used to fill in the gaps in the bubble memory and the data base memory.