DUKE HAD once given me a very interesting compliment.

It was after a burn, after the debriefing, after the usual bull and beer session; he and I had retired to the office to "hoist a jar" privately.

Duke didn't usually say much after a burn; he just sat and sipped. This time, however, he looked like he had something on his mind, so I nursed my drink and waited.

He had turned his chair to face the window and put his feet up on the little filing cabinet. He was holding his glass against his forehead, as if he had a headache and was enjoying the coolness of the ice.

"You know," he said, "you really impressed me this afternoon."

"Uh-thanks. What'd I do?"

"Amy Burrell."

"Oh," I said. "Yeah." I'd been wondering if he was going to say anything about that.

"You did right," Duke said. He lowered the glass from his forehead and glanced over at me.

I shrugged. "If you say so."

"I do say so," he said. "You didn't have a choice. You've known it for months that she's your weak link. I've seen it in your planning. And you knew it this afternoon. You did what you had to do."

"But I still feel bad about decking her."

"If you hadn't, it would be worse next time. Or it would be someone else. Think you can knock down Jose Moreno?"

"No way."

"Well, you'll probably never have to," Duke said. "Not now. Not after today."

"I hope so," I said. I shook my head. "But I keep seeing the look on her face-"

"You mean the tears? That's just the racket she runs on men. That crap doesn't work on officers."

"No, I mean when I jerked her back to her feet and shoved her at the dome. If she'd been carrying a weapon instead of a camera, I'd be dead now."

"That's precisely why she's carrying a camera instead of a gun. Because she can't be trusted with one." He sipped at his drink thoughtfully, then added, "Let me tell you something about integrity, Jim. It's like a balloon. It doesn't matter how good the rubber is; the air still goes out the hole."

"Uh ... sure," I said. I still wasn't sure where he was going with this.

"Integrity means airtight. No leaks. No holes in the balloon. A hundred percent."

"So, what you're saying is-?"

"What you did was appropriate. You closed up a hole. It was a good lesson for all of them. You showed them that there's no alternative to doing the job. Your team will be a lot tighter the next time out. You'll see the difference."

"Thanks," I said, and I meant it. "But the truth is, I did it without thinking. I just got pissed off at her continual whining."

Duke raised his glass in my direction. "Absolutely. And you administered the appropriate response. I congratulate you. I salute you." And he drank to my health.

I remembered that now. I wondered what kind of a salute Duke would give me if I punched out a general.

Well ...

At least, I could think about it.

I strode down to the front of the room and said, "Hi."

Fletcher looked up at me with a weary smile. "Hi, yourself." I plunged right in. "I have a question for you."

"The answer is probably `I don't know.' What's the question?"

"Well, your demonstration here was very impressive-despite General Whatsisname's reaction-"

"General Poole."

"That was General Poole? I didn't know they were so hard up for generals."

Fletcher allowed herself a hint of a smile. "What's your question, James?"

"Well, I was remembering something you said before, about the gastropedes' fur. You said it wasn't fur."

"Right. It's nerve endings."

"Well-that's my question. When two worms go into communion, isn't it possible that they're experiencing direct nerve-to-nerve contact?"

She nodded. "They very definitely are."

"Well-couldn't that be your mechanism? Maybe they're passing nerve impulses directly from one to the other."

She raised an eyebrow at me. "You think so?"

"You don't think much of the idea, do you?"

"As a matter of fact," Fletcher admitted, "I like the idea very much. It would explain a lot of things."

"But-?" I prompted.

"But-" she agreed, "it was one of the first things we tested for when we started putting Lucky and Tiny together. And it was one of the first hypotheses we had to discard. We kept finding arguments against it. Too many arguments."

"Really?"

"Really." She glanced at her watch. "All right, I'll have to give you the brief version. Here's what we know. Most of the Chtorran nerve-strands are sensory receptors of one type or another. We've identified at least seventeen distinct types of nerve-strands-different functions, different shapes of cross section, different colors, and so on. Each of those types are further divisible into categories of shade, length, and specialization of function. So far, we've identified over five hundred different sub-categories of nerve-strand. We presume that there is considerable overlap of function among the strand types, but we don't have the people available to do the necessary research.

"We do know that most of the strands are sensory receptors of one type or another-but maybe one strand in a thousand is a `tickler nerve.' It's a little transmitter; it can trigger any nerve it touches. That accounts for the tingly feeling of the fur. So, yes-it does look like a very good mechanism for communication. Pat yourself on the back for recognizing the possibility. Now here's the bad news. It can't possibly work. Do you want a minute to figure it out yourself?"

I thought about it. "It's a connection problem?"

"Not quite. The worms have no problem connecting. When they're in communion, they're connecting at least twenty percent of their surface area. But you're on the right track. It's a networking problem."

"Huh?"

"When you plug one computer into another, how many lines are you connecting?"

"Just one-oh, I see what you mean. There are one thousand and twenty-four individual channels in a standard lux-cable."

"Right. Now suppose you were working with wires instead of light and you had to connect each wire by hand-and suppose also that you didn't know which one went where. What are the chances of you plugging each of those lines into the right socket?"

"None, and less than that," I said. "There're billions of wrong combinations, and only one right one."

"That particular problem," she said, "would take longer than the life of this universe to solve. Now, raise it to the power of itself, and you have the odds against two worms forming a direct nerve-to-nerve contact for communication. Don't take my word for it," she added. "Run a simulation on the nearest terminal."

"No, it's all right. I'll take your word for it. But couldn't the worms have some kind of internal decoding?"

"We thought of that too," Fletcher said. "We had two fellows from the Minsky Foundation looking into that very problem. They said it was possible only if the creature was almost entirely brain and very little else. So far, we haven't found the evidence of that. Have you had the opportunity to see any of the photo-isotomographs?"

"I've seen the demonstrations, but I haven't had the opportunity to poke around on my own." A photo-isotomograph was a three-dimensional map. Easy to make. You thin-slice a frozen worm, taking a picture of the cross section after each slice. You store all the pictures in a computer-the computer holds the data as a three-dimensional array that can then be explored as a visual display. You can examine any part of the worm's body, inside or out, from any angle. With a joystick you can move around through the entire body, tracing the paths of blood vessels, nerves and other structures. So far, most of what we'd seen still fell into the category of "other structures." There were organs inside the worms with no apparent function. Were they evolutionary leftovers, the equivalent of the human appendix-or were they something else, on biological standby and still waiting to be activated?