It's a safe bet that no one will ever know, but here's where I decided to call on Nancy, who I thought could at least help me identify the weapon that had been used to make the cuts. Nancy had studied the macabre practice of human butchering and the evidence this practice left on bones-just the kind of science that I needed to wrap up this case.

When Nancy had a chance to examine the bones, she confirmed that the preliminary cuts on Bill's legs had been made with a thick, smooth-bladed knife, while the final amputations had been performed with hacking blows from an axe-like tool. She explained that a saw or a knife often leaves its “signature” on the bone, so that a hacksaw, for instance, makes fine irregular lines across the cut end of a bone, whereas a large table saw cuts cleanly in a single direction until the bone is severed. A chainsaw rips and chews through the bone in an instant, leaving gouges and chips in its wake, while a serrated knife leaves a pattern of dips and points-not to be confused with the straight, smooth cut mark often left by a butcher knife or a meat cleaver. The work done with cut marks by Nancy and my fellow forensic anthropologists-Steve Symes of Pennsylvania and the late William Maples of Florida-has helped to put numerous suspects behind bars.

I've had occasion to use cut-mark evidence in several other Kentucky cases, in sometimes surprising ways. One of the things that haunts investigators is knowing that a person can die violently-stabbed, shot, poisoned-without a single mark being left on the bone. And when the flesh has decomposed or been burned away, the bones are all you've got left.

Luckily, bones enable you to roughly determine the time a wound was inflicted, and fairly easily, too, because the nature of bone changes so radically after the body dies. When a person is alive or very recently dead, his or her bones resemble green wood. If you stick a knife into what we call a “green bone,” you can pry up a little sliver, because the bone-living tissue-is still pliable. If you try to make the same cut days or weeks after death, the bone is more like firewood-dead and dried-out wood-and it's not going to have that flexibility. That's why the cut marks made at or around the time of death look completely different from those made after death-if you know what to look for. So when Nancy and I reviewed the evidence in the case of Bill Conley, we concluded that the bone had been sliced “perimortem”-either at the time of death, immediately before, or fairly soon thereafter. Although we couldn't tell exactly why the amputation had happened, at least Nancy had identified the butchering tools.

From a death investigator's point of view, there are two types of fires: the kind that kill people, and the kind that somebody sets to disguise a homicide. Kentucky has far more than its share of the latter, and nobody really knows why. Is it that investigators in other states just aren't as suspicious about fire-related deaths as we are? Or does the criminal element in Kentucky really not know that even the most all-consuming fire inevitably leaves behind some human bones?

Of course, I'm glad they don't know. I'm kind of reluctant to tell them. But for the record, here it is: If you ever plan to incinerate a person, don't count on the body being completely destroyed. Trying to burn a human body-which after all is about 80 percent water-is like trying to burn a huge, sopping sponge. The fluid-filled organs, muscles, and bones can often withstand the fiercest of flames.

Ironically, one of my first major cases of homicide disguised by fire also happened in Pulaski County, where it was initially discovered by my old friend Sheriff Sam Catron. By the time this case broke in 1995, Sam, like so many law enforcement officers in Kentucky, had learned to keep my personal phone numbers in his pocket. My colleagues across the state know I'm available to them at any time of the day or night, so I wasn't surprised to get Sam's call at five o'clock one April Sunday afternoon.

“So here's the story,” Sam said wearily after we'd exchanged the usual pleasantries. “A small wood-frame farmhouse in the northeastern corner of the county burned to the ground earlier this afternoon. The fire department found two charred bodies in the living room. We found one more in one of the bedrooms.”

Sam and I both knew that this case wasn't necessarily a homicide. In Kentucky, it's not that unusual for remote dwellings to burn to the ground with sleeping or incapacitated occupants inside. Lots of mountain folks rely on wood-burning stoves and kerosene heaters for heat, and they sometimes use coal-oil lamps or even candles for light. The rural volunteer fire departments do their best, but sometimes they're not even aware of the fire until it's too late to help. In this case, a distant neighbor just happened to see the blaze and call it in. But the ramshackle old house was pretty much rubble by the time the firefighters got there.

Generally, the coroner, local law enforcement officers, and an arson specialist examine a fire scene. If the fire appears to be truly accidental, the bodies are simply recovered and brought to the M.E.'s office for autopsy and positive ID.

So what had made Sam and the coroner suspicious in this case? For one thing, the fire had occurred in the middle of a mild day in April. Not much chance that anyone was using a heater. Then there was the time of day. How likely was it that three able-bodied people were sleeping so soundly during daylight hours that they couldn't make their way out of this small one-story house, especially since there were plenty of doors and windows? The third clue, and the one most significant to trained fire-death investigators, was the fact that the bodies themselves were in abnormal positions.

Of course, probably no charred fire victim can be said to have a “normal” position, but there are certain things you look for when investigating a fire death. If a person dies from smoke inhalation-the usual cause of death in a fire-carbon monoxide builds up in the blood, causing a rapid loss of consciousness. Even after the person has blacked out, though, his or her body continues to react. The windpipe, or trachea, sucks in soot and smoke, and the organs and muscles turn a bright cherry red as carbon monoxide replaces oxygen in the blood. Last-minute chemical reactions in the muscles of the victim cause him or her to contort into what we call the “pugilistic” position-arms bent at the elbows; wrists and forearms drawn in toward the shoulders; hands balled into fists, as if the person were engaged in a boxing match. The legs, too, often flex slightly at the hips and the knees, so that the victim looks to be sitting in some imaginary chair.

However, none of the bodies in the Pulaski County farmhouse were in that position, Sam told me, meaning that there had been no physiological muscle reactions to the fire. And since Sam could actually see one victim's internal organs through rents in the abdominal wall, he'd noticed that there was no sign of the cherry-red discoloration that would have been there if the victim had died while inhaling smoke.

When I arrived at the scene, Sam walked me through the remnants of the little house and pointed out the other reasons he was suspicious. Something about this case had gotten to him: There was a catch in his normally soft voice, lending an air of uncharacteristic sadness to a man I had come to know as cheerful and completely professional even at a crime scene. Then, as we tiptoed through the rubble, Sam suddenly bent down beside a bright-yellow marker flag, and I saw what had affected this career lawman so profoundly. This victim was tiny-by my estimation, a child no older than two or three.