"Oh, hell," said Siegel, waving one hairy paw. "I didn't want to live forever, anyway. Count me in."

"Yeah, me too," grumbled Reilly. Marano and Lopez followed, as did the others. They grunted their assents with their usual sullen good nature and started gathering their gear and ecosuits.

I put Lopez in charge of outfitting, Reilly to double-check her, Locke and Valada to ready the rollagons, Braverman to take care of logistics, Bendat in charge of weaponry, and Nawrocki to take care of supplies. I picked Willig for mission specialist, which meant she got to handle the datawork. She gave me a dirty look, but she was already booting up the checklist.

In the meantime, I had my own work to do. I snagged a stool, hooked it around beneath me, plopped my butt onto it, and rolled up to a terminal, already shouting commands at it. In less than fifteen minutes, I'd filed a mission plan, the same one as before, plugged in a standard support program (coupled with a set of safeguard macros I'd written myself), waited for LI (lethetic intelligence) analysis from Green Mountain, noted the projected risk margins with a skeptical snort, and signed off on the orders anyway.

Ninety minutes later, we were in the air-and ninety minutes after that, we were on the ground again in northeastern Mexico. The rollagons trundled down the ramps, the VTOLs lifted off with a whisper, and we were once more crunching through the red crust of the waxy Chtorran infestation. The ocher sun was still high overhead, and the day was overcast with brick-red dust. There wasn't enough wind to clear it away.

We had at least six hours of daylight, all day tomorrow, and most of the day after that. If I couldn't find my answer in three days-or at least some kind of clue-I probably wasn't going to find it at all. At least, not this trip. I climbed up into the forward turret and started studying the scenery.

The crumpled hills were spotted and streaked with pustules of rancid vegetation; an appalling sight. It made me think of virulent sores spreading across the feverish body of a dying plague victim. Some of these plants lived only to die; they mulched the ground for the next generation to come.

Even through the filters, we could smell it-the sweaty, overripe stench of cancerous growth and fruity decay. The sickly-sweet odor had a druglike quality; nothing seemed real anymore in this nightmare landscape.

Siegel's voice came quietly through the all-talk channel. "Hey, ah… what exactly are we looking for, Cap'n?"

I didn't answer immediately. The same question had been echoing through my own mind. Finally, I said, "Y'know, there used to be a theory that the worms were only the shock troops to soften us up. Slum clearance. When the worms got the human population under control, then we'd see the next step of the infestation. The theory was that whatever was going to come after the worms would be higher up on the food chain… "

"You mean we're looking for something mean enough to eat a full-grown worm? Uh-oh…"

"That's one theory. But a lot of other people think that if the purpose of the infestation is to establish a stable ecology, then it's got to have its own checks and balances. Therefore, the infestation has a built-in controlling mechanism for each and every species that we see, some kind of biological governor-so when the worms get to be too widespread, something else wakes up or kicks in. Maybe it's some kind of phenomenon like seventeen-year locusts; only when the conditions are right does it start munching worms. I'm just thinking; if we can find out what it is, whatever it is, it might be useful."

"Right. That's what I thought you said. We're looking for something mean enough to eat a full-grown worm."

After a while, the view from the turret became oppressive. I couldn't stand looking at the dreadful red hills anymore; I dropped down into the command bay, wiping my forehead. I realized I was sweating. Drops were rolling down the back of my neck. "Is there something wrong with the air-conditioning in here?" I demanded.

Willig shook her head. "It has that effect on everybody, remember?"

I didn't answer. She was right. I sank down into my command seat—the information cockpit-and began reviewing the satellite pictures again-for the umpteenth time. The thing about satellite surveillance was that there were so many different ways to analyze scans, so many various filters and enhancements, so many possible patterns, that it was as much an art as a craft. We didn't have the trained personnel that we needed, and as good as the lethetic intelligence engines were, they still lacked the ability to make intuitive leaps. LIs could give you statistical probabilities; they couldn't give you hunches-although the last I'd heard, they were working on adding that function too.

The resolution on this latest set of pictures was good. We could have been looking down from the top of a ten-story building. I had downloaded six months of aerial surveillance into the vehicle's memory; that should have been more than enough. I dialed up last week's pictures and watched from above as five rollagons approached the dead worm, scanned it, and then moved on.

Unfortunately, a backward scan from that moment revealed little else of use. A ragged streak of clouds, the southernmost tail of a Gulf storm that had never gotten big enough to be called a hurricane, slid across the coast of Mexico and obscured the view of the target area. Before the clouds came by, there was no dead worm on the ground. Afterward, there was.

The creature's death had most likely occurred sometime around dawn; the mission log showed that the interior temperature of the carcass, at the time we had scanned it, had still been several degrees warmer than the noonday air. Whatever had killed the worm could not have been more than six hours away.

Back to the satellite record

LI image restoration, including infra-red scanning and ultrawide spectrum enhancements, suggested that an event of some kind had occurred on the site just before dawn, confirming my hypothesis. There was a cluster of infra-red activity, plus that peculiar flurry of spiky electromagnetic radiation that worms sometimes gave off when they went into multiple communion; it sounded like a burst of whistling static on an AM radio.

So… the dead worm-John Doe-obviously had had a meeting with several other worms. Or had it? The evidence only demonstrated the possibility of communion. It didn't prove it.

Assume that there were other worms in the neighborhood. Where-were they now? Were they in danger too? And what was their relationship to John Doe? Great questions-all I needed was the trenchcoat, the hat, and a half-cigarette dangling from my lower lip.

Another question-how many worms had participated in the encounter?

The satellite scan wasn't precise enough. Less than six. More than two. Three large worms? Five small ones?

I hunched over the keyboard, mumbling and typing in commands. I watched the displays change on the screens. If there were worms, there had to be worm nests within-oh, figure ten klicks to start with. Look for circular structures or patterns. Look for anything that could be a mandala seed… Look for worms. Scan outward and backward for large movements

Bingo!

The record showed three worms. Not too large… following a fourth worm. Toward the direction of the event under the clouds that left one of them dead.

Hmm, so where did these worms come from? Are they all from the same nest? I typed another command. Backtrack the worms. This time the wait was longer. The worms had come from the northwest, but their origin was unclear. Okay. Try a different way. Move the center of the search, enlarge the radius, scan the surrounding terrain again. Look for a nest.

Tick, tick, tick. The LI engine considered probabilities. Sorry. No nests in the surrounding terrain; now checking for circular anomalies…