"X" marks were clearly visible on the right side.

Benson leaned back in the chair. He did not open his eyes again. One of the technicians began to fix the monitor leads to his body, strapping them on with little dabs of electrolyte paste. They were attached quickly; soon his body was connected to a tangle of multicolored wires running off to the equipment.

Ellis looked at the TV monitor screens. The EEG was now tracing sixteen jagged lines; heartbeat was recorded; respirations were gently rising and falling; temperature was steady. The technicians began to punch pre-op parameters into the computer. Normal lab values had already been fed in. During the operation, the computer would monitor all vital signs at five-second intervals, and would signal if anything went wrong.

"Let's have music, please," Ellis said, and one of the nurses slipped a tape cartridge into the portable cassette recorder in a corner of the room. A Bach concerto began to play softly. Ellis always operated to Bach; he said he hoped that the precision, if not the genius, might be contagious.

They were approaching the start of the operation. The digital wall clock said 6:29:14 a.m. Next to it, an elapsed-time digital clock still read 0:00:00.

With the help of a scrub nurse, Ross put on her sterile gown and gloves. The gloves were always difficult for her. She didn't scrub in frequently, and when she plunged her fingers into the gloves she caught her hand, missing one of the finger slots and putting two fingers in another. It was impossible to read the scrub nurse's reaction; only her eyes were visible above the mask. But Ross was glad that Ellis and the other surgeons were turned away attending to the patient.

She stepped to the back of the room, being careful not to trip over the thick black power cables that snaked across the floor in all directions. Ross did not participate in the initial stages of the operation. She waited until the stereotaxic mechanism was in place and the coordinates were determined. She had time to stand to one side and pluck at her glove until all the fingers were in the right slots. There was no real purpose for her to attend the operation at all, but McPherson was insistent that one member of the non-surgical staff scrub in each day that they operated. He felt it kept the Unit more cohesive. At least that was what he said.

She watched Ellis and his assistants across the room draping Benson; then she looked over to the draping as seen on the closed-circuit monitor. The entire operation would be recorded on video tape for later review.

"I think we can start now," Ellis said easily. "Go ahead with the needle."

The anaesthetist, working behind the chair, placed the needle between the second and third lumbar spaces of Benson's spine. Benson moved once and made a slight sound, and then the anaesthetist said, "I'm through the dura. How much do you want?"

The computer console flashed "OPERATION BEGUN." The computer automatically started the elapsed-time clock, which ticked off the seconds.

"Give me thirty cc's to begin," Ellis said. "Let's have X-ray, please."

The X-ray machines were swung into position at the front and side of the patient's head. Film plates were set on, locking in with a click. Ellis stepped on the floor button, and the TV screens glowed suddenly, showing black-and-white images of the skull. He watched in two views as air slowly filled the ventricles, outlining the horns in black.

The programmer sat at the computer console, his hands fluttering over the buttons. On his TV display screen, the words "PNEUMOGRAPH INITIATED" appeared.

"All right, let's fix his hat," Ellis said. The boxlike tubular stereotactic frame was placed over the patient's head. Burr-hole locations were fixed and checked. When Ellis was satisfied, he injected local anaesthetic into the scalp points. Then he cut the skin and reflected it back, exposing the white surface of the skull.

"Drill, please."

With the 2-mm drill, he made the first of the two holes on the right side of the skull. He placed the stereotactic frame - the "hat" - over the head, and screwed it down securely. Ross looked over at the computer display. Values for heart rate and blood pressure flashed on the screen and faded; everything was normal. Soon the computer, like the surgeons, would begin to deal with more complex matters.

"Let's have a position check," Ellis said, stepping away from the patient, frowning critically at Benson's shaved head and the metal frame screwed on top of it. The X-ray technician came forward and snapped the pictures.

In the old days, Ross remembered, they actually took X-ray plates and determined position by visual inspection of the plates. It was a slow process. Using a compass, protractor, and ruler, they drew lines across the X-ray, measured them, rechecked them. Now the data were fed directly to the computer, which did the analysis more rapidly and more accurately.

All the team turned to look at the computer print-out screen. The X-ray views appeared briefly, and were replaced by schematic drawings. The maxfield location of the stereotactic apparatus was calculated; the actual location was then merged with it. A set of coordinates flashed up, followed by the notation "PLACEMENT CORRECT."

Ellis nodded. "Thank you for your consultation," he said humorlessly, and went over to the tray which held the electrodes.

The team was now using Briggs stainless-steel

Teflon-coated electrode arrays. In the past, they had tried almost everything else: gold, platinum alloy, and even flexible steel strands in the days when the electrodes were placed by inspection. The old inspection operations were bloody, messy affairs. It was necessary to remove a large portion of the skull and expose the surface of the brain. The surgeon found his landmark points on the surface itself, and then placed his electrodes in the substance of the brain. If he had to place them in deep structures, he would occasionally cut through the brain to the ventricles with a knife, and then place them. There were serious complications; the operations were lengthy; the patients never did very well.

Now the computer had changed all that. The computer allowed you to fix a point precisely in three-dimensional space. Initially, along with other researchers in the field, the NPS group had tried to relate deep brain points to skull architecture. They measured their landmark points from the orbit of the eye, from the meatus of the ear, from the sagittal suture. That, of course, didn't work - people's brains did not fit inside their skulls with any consistency. The only way to determine deep brain points was in relation to other brain points - and the logical landmarks were the ventricles, the fluid-filled spaces within the brain. According to the new system, everything was determined in relation to the ventricles.

With the help of the computer, it was no longer necessary to expose the brain surface. Instead, a few small holes were drilled in the skull and the electrodes inserted, while the computer watched by X-ray to make sure they were being placed correctly.

Ellis picked up the first electrode array. From where Ross stood, it looked like a single slender wire. Actually, it was a bundle of twenty wires, with staggered contact points. Each wire was coated with Teflon except for the last millimeter, which was exposed. Each wire was a different length, so that under a magnifying glass, the staggered electrode tips looked like a miniature staircase.

Ellis checked the array under a large glass. He called for more light and turned the array, peering at all contact points. Then he had a scrub nurse plug it into a testing unit and test every contact. This had been done dozens of times before, but Ellis always checked again before insertion. And he always had four arrays sterilized, though he would need only two. Ellis was careful.

At length he was satisfied. "Are we ready to wire?" he asked the team. They nodded. He stepped up to the patient and said, "Let's go through the dura."

Up to this point in the operation, they had drilled through the skull, but had left intact the membrane of dura mater which covered the brain and enclosed the spinal fluid. Ellis's assistant used a probe to puncture the dura.

"I have fluid," he said, and a thin trickle of clear liquid slid down the side of the shaved skull from the hole. A nurse sponged it away.

Ross always found it a source of wonder the way the brain was protected. Other vital body organs were well-protected, of course: the lungs and heart inside the bony cage of the ribs, the liver and spleen under the edge of the ribs, the kidneys packed in fat and secure against thick muscles of the lower back. Good protection, but nothing compared to the central nervous system, which was encased entirely in thick bone. Yet even this was not enough; inside the bone there were sac-like membranes which held cerebrospinal fluid. The fluid was under pressure, so that the brain sat in the middle of a pressurized liquid system that afforded its superb protection.

McPherson had compared it to a fetus in a water-filled womb. "The baby comes out of the womb," McPherson said, "but the brain never comes out of its own special womb."

"We will place now," Ellis said.

Ross moved forward, joining the surgical team gathered around the head. She watched as Ellis slid the tip of the electrode array into the burr hole and then pressed slightly, entering the substance of the brain. The technician punched buttons on the computer console. The display screen read:

The patient did not move, made no sound. The brain could not feel pain; it lacked pain sensors. It was one of the freaks of evolution that the organ which sensed pain throughout the body could feel nothing itself.

Ross looked away from Ellis toward the X-ray screens.

There, in harsh black and white, she saw the crisply outlined white electrode array begin its slow, steady movement into the brain. She looked from the anterior view to the lateral, and then to the computer-generated images.

The computer was interpreting the X-ray images by drawing a simplified brain, with the temporal-lobe target area in red and a flickering blue track showing the line the electrode must traverse from entry point to the target area. So far, Ellis was following the track perfectly.

"Very pretty," Ross said.

The computer flashed up triple coordinates in rapid succession as the electrodes went deeper.