At the end of ten hours of work, he sat back to consider what he had learned. He drew up a short list:

1. The lethal agent is approximately 1 micron in size. Therefore it is not a gas or molecule, or even a large protein or virus. It is the size of a cell, and may actually be a cell of some sort.

2. The lethal agent is transmitted by air. Dead organisms are not infectious.

3. The lethal agent is inspired by the victim, entering the lungs. There it presumably crosses over into the bloodstream and starts coagulation.

4. The lethal agent causes death through coagulation. This occurs within seconds, and coincides with total coagulation of the entire body vascular system.

5. Anticoagulant drugs do not prevent this process.

6. No other pathologic abnormalities are known to occur in the dying animal.

Burton looked at his list and shook his head. Anticoagulants might not work, but the fact was that something stopped the process. There was a way that it could be done. He knew that.

Because two people had survived.

17. Recovery

AT 1147 HOURS, Mark Hall was bent over the computer, staring at the console that showed the laboratory results from Peter Jackson and the infant. The computer was giving results as they were finished by the automated laboratory equipment; by now, nearly all results were in.

The infant, Hall observed, was normal. The computer did not mince words:

SUBJECT CODED—INFANT—SHOWS ALL LABORATORY VALUES

WITHIN NORMAL LIMITS

However, Peter Jackson was another problem entirely. His results were abnormal in several respects.

SUBJECT CODED JACKSON, PETER

LABORATORY VALUES NOT WITHIN NORMAL LIMITS

FOLLOW

BLOOD SMEAR SHOWS MANY IMMATURE ERYTHROCYTE FORMS

Some of the results were easy to understand, others were not. The hematocrit, for example, was rising because Jackson was receiving transfusions of whole blood and packed red cells. The BUN, or blood urea nitrogen, was a test of kidney function and was mildly elevated, probably because of decreased blood flow.

Other analyses were consistent with blood loss. The reticulocyte count was up from 1 to 6 per cent. Jackson had been anemic for some time. He showed immature red-cell forms, which meant that his body was struggling to replace lost blood, and so had to put young, immature red cells into circulation.

The prothrombin time indicated that while Jackson was bleeding from somewhere in his gastrointestinal tract, he had no primary bleeding problem: his blood clotted normally.

The sedimentation rate and SGOT were indices of tissue destruction. Somewhere in Jackson's body, tissues were dying off.

But the pH of the blood was a bit of a puzzle. At 7.31, it was too acid, though not strikingly so. Hall was at a loss to explain this. So was the computer.