And if there had been unreceived MCN messages, then the computers were not functioning properly. Captain Morris watched as the computer ran its usual internal check program, which read out as all circuits functioning.

Unsatisfied, he punched in the CHECKLIM program, a more rigorous testing of the circuit banks. It required 0.03 seconds for the machine to come back with its answer: a row of five green lights blinked on the console. He walked over to the teleprinter and watched as it typed:

MACHINE FUNCTION ON ALL CIRCUITS

WITHIN RATIONAL INDICES

He looked and nodded, satisfied. He could not have known, as he stood before the teleprinter, that there was indeed a fault, but that it was purely mechanical, not electronic, and hence could not be tested on the check programs. The fault lay within the teleprinter box itself. There, a sliver of paper from the edge of the roll had peeled away and, curling upward, had lodged between the bell and striker, preventing the bell from ringing. It was for this reason that no MCN transmissions had been recorded.

Neither machine nor man was able to catch the error.

18. The Noon Conference

ACCORDING TO PROTOCOL, the team met every twelve hours for a brief conference, at which results were summarized and new directions planned. In order to save time the conferences were held in a small room off the cafeteria; they could eat and talk at the same time.

Hall was the last to arrive. He slipped into a chair behind his lunch—two glasses of liquid and three pills of different colors—just as Stone said, "We'll hear from Burton first."

Burton shuffled to his feet and in a slow, hesitant voice outlined his experiments and his results. He noted first that he had determined the size of the lethal agent to be one micron.

Stone and Leavitt looked at each other. The green flecks they had seen were much larger than that; clearly, infection could be spread by a mere fraction of the green fleck.

Burton next explained his experiments concerning airborne transmission, and coagulation beginning at the lungs. He finished with his attempts at anticoagulation therapy.

"What about the autopsies?" Stone said. "What did they show?"

"Nothing we don't already know. The blood is clotted throughout. No other demonstrable abnormalities at the light microscope level."

"And clotting is initiated at the lungs?"

"Yes. Presumably the organisms cross over to the bloodstream there—or they may release a toxic substance, which crosses over. We may have an answer when the stained sections are finished. In particular, we will be looking for damage to blood vessels, since this releases tissue thromboplastin, and stimulates clotting at the site of the damage."

Stone nodded and turned to Hall, who told of the tests carried out on his two patients. He explained that the infant was normal to all tests and that Jackson had a bleeding ulcer, for which he was receiving transfusions.

"He's revived," Hall said. "I talked with him briefly."

Everyone sat up.