That was the normal sequence.

But if the damage was extensive, if it began at the lungs and worked its way . . .

"I'm wondering," Hall said, "if our organism attacks vessel walls. If so, it would initiate clotting. But if clotting were prevented in certain persons, then the organism might eat away and cause hemorrhage in those persons."

"And insanity," Burton said, hunting through his slides. He found three of the brain, and checked them.

No question.

The pathology was striking. Within the internal layer of cerebral vessels were small deposits of green. Burton had no doubt that, under higher magnification, they would turn out to be hexagonal in shape.

Quickly, he checked the other slides, for vessels in lung, liver, and spleen. In several instances he found green spots in the vessel walls, but never in the profusion he found for cerebral vessels.

Obviously the Andromeda Strain showed a predilection for cerebral vasculature. It was impossible to say why, but it was known that the cerebral vessels are peculiar in several respects. For instance, under circumstances in which normal body vessels dilate or contract—such as extreme cold, or exercise—the brain vasculature does not change, but maintains a steady, constant blood supply to the brain.

In exercise, the blood supply to muscle might increase five to twenty times. But the brain always has a steady flow: whether its owner is taking an exam or a nap, chopping wood or watching TV. The brain receives the same amount of blood every minute, hour, day.

The scientists did not know why this should be, or how, precisely, the cerebral vessels regulate themselves. But the phenomenon is known to exist, and cerebral vessels are regarded as a special case among the body's arteries and veins. Clearly, something is different about them.

And now there was an example of an organism that destroyed them preferentially.

But as Burton thought about it, the action of Andromeda did not seem so unusual. For example, syphilis causes an inflammation of the aorta, a very specific, peculiar reaction. Schistosomiasis, a parasitic infection, shows a preference for bladder, intestine, or colonic vessels—depending on the species. So such specificity was not impossible.

"But there's another problem," he said. "In most people, the organism begins clotting at the lungs. We know that. Presumably vessel destruction begins there as well. What is different about—"

He stopped.

He remembered the rats he had anticoagulated. The ones who had died anyway, but had had no autopsies.

"My God," he said.

He drew out one of the rats from cold storage and cut it open. It bled. Quickly he incised the head, exposing the brain. There he found a large hemorrhage over the gray surface of the brain.

"You've got it," Hall said.

"If the animal is normal, it dies from coagulation, beginning at the lungs. But if coagulation is prevented, then the organism erodes through the vessels of the brain, and hemorrhage occurs."

"And insanity."

"Yes." Burton was now very excited. "And coagulation could be prevented by any blood disorder. Or too little vitamin K. Malabsorption syndrome. Poor liver function. Impaired protein synthesis. Any of a dozen things."