And he drew a deep breath and spewed it out hastily in one long succession of syllables. Then, in the charged air that lingered, he concluded almost in tears, “. . . more or less. I don’t remember the exact order of what I called him, and maybe I left out something or put in something, but that was about it.”

Only the slightest flush betrayed any feeling on the part of the robopsychologist. She said, “I am aware of the meaning of most of the terms used. The others, I suppose, are equally derogatory.”

“I’m afraid so,” agreed the tormented Black.

“And in among it, you told him to lose himself.”

“I meant it only figuratively.”

“I realize that. No disciplinary action is intended, I am sure.” And at her glance, the general, who, five seconds earlier, had seemed not sure at all, nodded angrily.

“You may leave, Mr. Black. Thank you for your cooperation.”

It took five hours for Susan Calvin to interview the sixty-three robots. It was five hours of multi-repetition; of replacement after replacement of identical robot; of Questions A, B, C, D; and Answers A, B, C, D; of a carefully bland expression, a carefully neutral tone, a carefully friendly atmosphere; and a hidden wire recorder.

The psychologist felt drained of vitality when she was finished.

Bogert was waiting for her and looked expectant as she dropped the recording spool with a clang upon the plastic of the desk.

She shook her head, “All sixty-three seemed the same to me. I couldn’t tell—”

He said, “You couldn’t expect to tell by ear, Susan. Suppose we analyze the recordings.”

Ordinarily, the mathematical interpretation of verbal reactions of robots is one of the more intricate branches of robotic analysis. It requires a staff of trained technicians and the help of complicated computing machines. Bogert knew that. Bogert stated as much, in an extreme of unshown annoyance after having listened to each set of replies, made lists of word deviations, and graphs of the intervals of responses.

“There are no anomalies present, Susan. The variations in wording and the time reactions are within the limits of ordinary frequency groupings. We need finer methods. They must have computers here. No.” He frowned and nibbled delicately at a thumbnail. “We can’t use computers. Too much danger of leakage. Or maybe if we—”

Dr. Calvin stopped him with an impatient gesture, “Please, Peter. This isn’t one of your petty laboratory problems. If we can’t determine the modified Nestor by some gross difference that we can see with the naked eye, one that there is no mistake about, we’re out of luck. The danger of being wrong, and of letting him escape is otherwise too great. It’s not enough to point out a minute irregularity in a graph. I tell you, if that’s all I’ve got to go on, I’d destroy them all just to be certain. Have you spoken to the other modified Nestors?”

“Yes, I have,” snapped back Bogert, “and there’s nothing wrong with them. They’re above normal in friendliness if anything. They answered my questions, displayed pride in their knowledge—except the two new ones that haven’t had time to learn their etheric physics. They laughed rather good-naturedly at my ignorance in some of the specializations here.” He shrugged, “I suppose that forms some of the basis for resentment toward them on the part of the technicians here. The robots are perhaps too willing to impress you with their greater knowledge.”