Here’s the plot of an episode of a fifties TV show called Science Fiction Theatre:

A baby woolly mammoth has been found preserved in the ice of the Arctic Circle. A group of scientists, including a sorrowful-looking female zoologist in a Peter Pan collar, melt the ice and stare in wonder at the prehistoric creature lying motionless on the floor. Could it possibly still be alive, in a state of suspended animation? The scientists administer a “galvanic shock,” and the baby mammoth miraculously scrambles to his feet. But something is wrong: He misses his mother. He begins to pine away until the zoologist, whose own child has died, rushes to comfort him.

But then she is injured in a car wreck, and by the time she recovers and gets back to the baby mammoth, it’s too late. The weakened, heartbroken creature greets her with a forlorn bleat from his trunk and falls over dead.

“Funny, isn’t it?” the zoologist says as she dabs at her eyes with a handkerchief. “Crying over an animal that should have died half a million years ago?”

When I tracked down this show recently and watched it for the first time since boyhood, it looked itself like an artifact of the Ice Age. The images on the DVD transfer were faded and skittery. The mammoth was played by a baby elephant wearing plastic-looking tusk extenders and scraggly tufts of hair glued to its head. But I had no trouble recalling the story’s haunting effect on me. Deep into adulthood the memory lingered: a lonesome lost creature imprisoned in time, a frozen heart made to beat again, a block of ice like a window through which the light of prehistory still dimly shone.

Back in the fifties in Abilene, as I sat in my cowboy pajamas watching Science Fiction Theatre, it did not occur to me to find this scenario far-fetched. I had a calm certainty that in my lifetime a mammoth would indeed be brought back to life. As it turns out, I might have been right. A number of frozen mammoths have been extracted from permafrost over the past century, and though they are very much dead and cannot be reconstituted by a galvanic shock, their nuclei could conceivably be inserted into the egg cells of an elephant, which would act as a surrogate mother. A group of Japanese scientists recently predicted that by 2016 the world may witness the birth of a baby mammoth.

If it happens, it will be a reunion of sorts, since humans and mammoths go back a long way. In Texas, the people we know today as Paleo-Indians were hunting mammoths a mere 10,000 or 11,000 years ago. By that time, the glaciers that had once covered much of North America were long gone. And they had never extended this far south, so there is no chance of finding a frozen mammoth in Texas, much less one in suspended animation. But mammoth bones are everywhere: emerging from eroded cutbanks, plowed up in farmers’ fields, unearthed in construction sites.

These bones are from Mammuthus columbi, the Columbian mammoth, not the woolly mammoth that was featured in Science Fiction Theatre. Columbian mammoths ranged farther south than their cold-weather cousins, throughout the Pleistocene grasslands and woodlands of North America, all the way down to what is now Mexico. They were larger than the woolly mammoths, reaching up to fourteen feet high, with domed heads and huge sweeping tusks.

Sixty or seventy thousand years ago, a herd of these creatures was foraging in a floodplain a few miles west of what is now downtown Waco. The landscape, in the waning millennia of the Ice Age, would have been prairies and lightly forested savannas populated with all sorts of vanished megafauna: camels, saber-toothed cats, ground-dwelling sloths that could rear up to twenty feet high, and massive armadillo-like glyptodonts that shuffled along like living boulders.

There were a few juveniles in the mammoth herd, but most were adult females. A flash flood, roaring out of the ancestral Bosque River, caught them at the bottom of a steep-sided tributary. Perhaps a few members of the herd were able to scramble up the banks of the tributary and get away, but the rest were struck by a lethal blast of moving water. The positions in which the mammoths’ bodies were found suggest that the adult females—in a protective gesture familiar to us from the behavior of modern elephants—tried to form a shield around the juveniles. But even full-grown mammoths could not keep their footing against the sudden velocity of the flood. Old and young drowned together as the water swept them away. Their bodies were then buried by subsequent deposits of soil and sediment and remained hidden from sight until 1978, when two Waco men, Paul Barron and Eddie Bufkin, came across what they thought was a mammoth bone exposed in the eroded wall of a ravine.

They knew they had found something interesting, but they didn’t know how interesting until archeologists from Baylor University began to dig at the site and uncovered the remains of five mammoths. Later excavations revealed as many as nineteen more, as well as the bones of an extinct camel and a single tooth from a young saber-toothed cat. Taken all together, it was a mother lode—“the largest single-herd, non-human-related [Columbian] mammoth death site in the world,” as John D. Bongino, a Baylor graduate student, described it in his 2007 master’s thesis.

In 2009 the Waco Mammoth Site opened to the public, adding to the reputation for eclectic tourism that Waco already enjoyed with the Texas Ranger Hall of Fame and the Dr Pepper Museum. The site is located about four miles west of Interstate 35, where the Brazos River makes a northeasterly bend at its confluence with the Bosque. There is a parking lot for fifty or so cars and a small welcome center and gift shop, from which visitors walk down a path toward the site of the tributary. “Follow me to the Mammoths,” the signs read. High up on one of the light poles along the way, a ribbon marks the height of a Columbian mammoth. It’s a simple but startling reminder. The day I visited, I stared up at the ribbon, trying to imagine the colossal beast whose shoulder muscles would have rippled six or seven feet above my head.

Beyond the light pole is a steel bridge spanning a narrow ravine, the spot where the original bones were found. Most of the remains were removed years ago, encased in plaster jackets and stored in Baylor’s Mayborn Museum. Of the roughly 24 mammoths found at the Waco Mammoth Site, only 6 are still there, and they are on spectacular display in the elegant building on the far side of the ravine.

The building is known as the dig shelter. I found it to be as airy and hushed inside as an art gallery; indeed, the first thing I saw as I entered was a pony-tailed artist named Lee Jamison, standing on a scaffold and surveying his work on an almost completed mural of a mammoth herd fleeing from a wall of water. Visitors to the dig shelter can view the mural and exhibit from a catwalk that is suspended from the ceiling. The gentle smell of overturned earth rises from about six feet below, where in situ bones of the remaining mammoths lie in a matrix of the alluvial soil that was deposited over many thousands of years by the slackwater floods of the Bosque River. The three deepest-lying specimens are part of the nursery herd, but the most complete and spectacular remains on display were found in shallower soil, which means that these animals were killed not in the flood that engulfed the nursery herd but in a separate event, 10,000 or 20,000 years later, in more or less the same place, under probably more or less the same conditions.

“Mammoth Q” reads the label next to an impressive pile of sprawled and spavined bones near the entrance. Mammoth Q, a victim of the more recent flood, is the site’s star attraction, the only bull found during the three excavations. One of his ribs had been broken, not in the flood but long before; there’s a bulge where it had healed over. The beast’s massive pelvic bone is in one compressed piece, and most of the material is flattened and scattered, the great toe bones lying there like the playing tokens from some inscrutable ancient board game, the skull a caved-in shambles. In spite of the wreckage, though, it’s clear enough what this creature had been. All you need is one glance at the serpentine tusks growing out of the smashed skull, their tips curving emphatically toward each other. They’re immediately distinguishable from the shorter tusks of a modern-day elephant or from the long, sabre-like tusks of the mammoth’s slope-headed contemporary relation, the mastodon.

Scattered nearby are the bones of Mammoth R, a juvenile that was killed along with the bull. The younger animal was found in an interesting position, more or less athwart the bull’s tusks, and this gave rise to speculation that the bull, in a noble final gesture, had tried to heave the younger animal up and out of danger as the waters rushed in. It’s a touching thought, the idea of this prehistoric behemoth selflessly concerned with the safety of his young as he himself fought to keep from drowning. But more likely the receding floodwaters had simply trapped the body of the drowned juvenile against the tusks of its elder.

Nevertheless, at the Waco Mammoth Site the drama of that long-ago day is still eerily tangible. For me, there was something particularly poignant about Mammoth W, the female adult killed along with the bull and the juvenile in the later flood, whose bones lie forty or fifty feet away from theirs. I stood on the catwalk staring down at Mammoth W for a long time, at a heap of ribs and vertebrae and disarticulated leg and shoulder bones still half-buried in the soil. The lower jaw had broken off and sat upright, so that I looked down upon the single enormous molar on each side of the mouth. The teeth were the size and shape of shoes, and the enamel striations on their surfaces created a pattern that made me think of astronaut footprints on the surface of the moon. Emerging from the ground next to the jawbone was the skull. It lay on its side, its occipital cavity still covered in dirt, the near tusk broken off, the other sloping down and disappearing into the earth.

The overall effect, though, was one of living cohesion. The bones were close enough to their original positions that you saw the creature whole, and once you saw her as whole you couldn’t help but see her as real, something—someone—who actually existed, whose death agony had been caught in a geological freeze-frame almost as vivid as a photograph.

Texas is an especially good place for such Ice Age frissons, particularly if you’re on the scout for Mammuthus columbi. In far north San Antonio, for instance, hidden among the new subdivisions and unfinished strip malls claiming the open land between Loop 1604 and Bulverde, there is a remnant strip of juniper and oak guarded by a locked gate. If you happen to be a homeowner in the Villas at Canyon Springs, standing at the window of your master suite in one of the houses that edge up to this oversized vacant lot, you have a backyard view of one of the continent’s great prehistoric treasure troves. If you had been here 20,000 years ago, you might have noticed a saber-toothed cat grunting and heaving as it dragged a baby mammoth down a sloping cave opening to its den.

The cat’s den is still there, and for almost a hundred years scientists have been mining it for the bones of Pleistocene predators and their prey. The site is called Friesenhahn Cave.

“It’s kind of like the La Brea Tar Pits of San Antonio,” Laurence Meissner, a 65-year-old biology professor at Concordia University, in Austin, told me on the day I joined his zoology class on a field trip. Meissner has been working at the site for thirteen years. Though he asked me to emphasize that Friesenhahn Cave is not a public site, it’s clear enough that he regards it as a public treasure, a place whose secrets have only begun to be revealed.

The cave is small, with a single oval-shaped room sixty feet long, though Meissner has lately been doing electrical conductivity tests to determine if there are other chambers. The only entrance is a thirty-foot vertical shaft, covered with a steel grate. Before Meissner opened the grate to lead his students into the cave, he pointed out a shallow depression a few yards away to the north. This had been the original cave opening, he said, the sloping entryway down which the cats and other predators had hauled their kills. Ten to fifteen thousand years ago, it filled up with rubble, sealing the cave. Millennia later, the ceiling fell in, creating the other opening. But since this newer entrance is a sheer thirty-foot drop rather than a natural ramp, the cave was apparently never again used as a den for big predators.

The students put on hard hats and descended the ladder one at a time. When they reached the bottom, Meissner pointed out the spot where an entire skeleton of Homotherium serum (now on display at the University of Texas) was found in the early fifties. Homotherium, a species of saber-toothed cat, was tough and bandy, smaller than the iconic Smilodon, whose curving canines were often seven inches long and whose remains have also been found in Friesenhahn. But the abundance of Homotherium and immature mammoth bones suggests that this less imposing cat was particularly adapted to preying on young mammoths, no small accomplishment if we assume that mammoths, like modern-day elephants, were formidable protectors of their offspring.

Meissner ushered the students just beyond the natural spotlight created by the shaft, where an impacted mass of rocks and boulders indicated the location of the original entranceway. He shone a light along the wall, revealing what looked like a seam of friable rock but was in reality the protruding tusk of a young mammoth.

“And I’m pretty confident that’s an adult mammoth tibia bone,” he said, shining the light on a triangular-shaped object sticking out of the cave wall. Since it’s unlikely that any animal could have dragged an adult mammoth down here, he speculated that it had been scavenged postmortem by a tiger or bear—not just any bear, but the extinct short-faced bear, which was almost twice the size of a modern grizzly.

“Ooh, look at this!” he called out to the students when he crossed the entrance chamber to the other side. Recent rains had washed away some of the soil at the bottom of the cave, revealing even more hidden material. “Look at this big bone! It’s a leg bone!”

No sooner had he found the new bone than one of the students pointed out yet another lying nearby. “And look at that!” he almost yelped. He looked up at the students gathered around him, their plastic hard hats gleaming in the light from the entrance shaft, as he directed their attention to the bone that was just now emerging from the mud into human sight. “Christmas is coming!”

“Mammoth bones show up all over the place,” Ernest Lundelius, professor emeritus of paleontology at the University of Texas told me a few days later as he led me into the basement of UT’s Vertebrate Paleontology Laboratory. “A tooth here, a tusk there, once in a while you get a bit more.”

He flipped a light switch, illuminating an immense room filled with rows and rows of floor-to-ceiling metal shelves.

“This is where we keep the big stuff.”

It took me a moment to process my amazement. It was as if I had taken a wrong turn in Costco and wandered into the prehistoric bone section. The warehouse-size room was filled with tusks, femurs, and pelvic cradles from mammoths; skulls and arm bones from giant ground sloths, ancient rhinoceroses, and short-faced bears; and shells the size of industrial woks that had belonged to long-extinct species of turtles.

Lundelius, a vigorous, folksy man in his eighties, had personally extracted some of the material on these shelves from the ground. He pointed to a giant irregular lump of plaster that contained a mammoth skull he had found in Bee County. “God, that block was heavy,” he recalled. “It took six of us to get that up on a little cart and drag it out of the gulley.”

Lundelius indulged me while I browsed through his shelves a while longer, then we went back up to the first floor, where there were banks of tall specimen cabinets. He opened one that read “Friesenhahn Cave” and pulled out trays that held the jawbones and teeth of juvenile and baby mammoths.

“It’s a neat system,” he said as he traced with his fingers a series of ridges and valleys that made up the chewing surface of a paperweight-size mammoth tooth. “There are three different substances here—enamel, cement, and dentene—that have three different hardnesses. It’s a self-sharpening system. It’s really nifty.”

An older mammoth would have had only four such teeth in its mouth at a time. Over the creature’s lifespan, which could reach eighty years, they would be worn out and replaced six times. Mammoths lived on grasses and woody vegetation. They ripped the grass up in tufts or stripped leaves off trees with their trunks, which at the tip likely had two gripping “fingers” delicate enough to pick flowers. A mammoth might have weighed ten tons, but it moved in a surprisingly light-footed way, almost on tiptoe, as the spongy tissue of its footpads expanded and contracted with each step.

The mammoth’s elaborately corkscrewing tusks grew throughout its life, between one and six inches a year. They were used for protection from predators and in dominance battles with other mammoths, as well as for stripping the bark off trees and plowing up plants. These mighty tusks were rooted in the countervailing mass of the mammoth’s skull. In its skeletal nakedness, its inscrutable blankness, the skull is almost supernaturally riveting, as I was reminded when Lundelius and I paused at a specimen on display in the Vertebrate Paleontology Lab’s entrance hall. Viewed from the front, this skull was a sheer wall of bone, with an odd oblong hollow near the crest. This had been the mammoth’s nasal cavity, but its position at the front of the skull gave it the look of a disturbing all-seeing eye.

“The occipital orbits are here on the side,” Lundelius said, “but the early guys in the Mediterranean thought this nasal opening was the eye hole, and it has been suggested that this is where the whole cyclops idea came from.”

Closer to home, there was a similar misconception. The Comanche came to the Southern Plains long after the mammoths had disappeared, and when they encountered such colossal skulls, there were no living creatures to assign them to. They believed they were the remains of the mythical cannibal owl, which grabbed children in the night and bore them away on silent wing beats.

We know from cave paintings, ivory carvings, and the remains of huts that were supported by bones and tusks that mammoths and humans co-existed in Europe as far back as 35,000 years ago. In North America, the evidence is not as abundant, but it’s still compelling, and the best place to find it is at the Blackwater Draw site, a mildly picturesque declivity in a stretch of windblown farmland just across the Texas state line between Portales and Clovis, New Mexico. There is a little museum in Portales that showcases an impressive collection of the mammoth bones, Paleo-Indian tools, and spear points found there. But to get the full impact you need to drive another five miles, keeping an eye out for the easy-to-miss sign next to a lonely ranch gate. From there a dirt road leads to a no-frills visitors center.

Unless, of course, you regard mammoth bones as frills. There are plenty of them on display at the visitors center, including a towering femur and a pair of tusks long enough to rival the world’s record holder, a sixteen-foot tusk found near Post, Texas, that’s now in the collection of New York’s American Museum of Natural History.

But it is not the bones themselves but what has been found with them that makes Blackwater Draw such an important site. It was here, eighty years ago, that archeologists first discovered the distinctive fluted projectile points that now bear the name Clovis, which established human occupation of this area back at least twelve thousand years, during the twilight time of the mammoths.

The Clovis people didn’t just share the land with mammoths; they hunted them. Over the decades, the bones of 28 separate mammoths have been found in Blackwater Draw. George Crawford, the site archeologist, told me as we headed down into the draw that 8 or 9 of them had been killed by humans.

How do we know this? “Because,” Crawford said, “there were points found in and around the carcasses. Everybody feels pretty comfortable when you’ve got a lot of spear points like that. It’s pretty clear that these mammoths were killed by people.”

Crawford had to speak above a ferocious wind that came howling down off the Llano Estacado. Back in the thirties, during the time of the Dust Bowl, part of the site had been nicknamed Elephant Tusk Lake because of the mammoth remains exposed by that same ceaseless wind.

It is still a parched-looking place, though there is some greenery in the form of cottonwoods at the bottom of the shallow draw, and archeologists have planted native grasses and shrubs on the slopes to prevent erosion. But Blackwater Draw had been a small, bucolic lake once, thanks to an artesian spring that continued flowing until historic times, and the water had lured wildlife and people to its banks.

Not much remains to indicate the presence of that Ice Age watering hole to a casual visitor. Pleistocene-era dunes still stand on what was once the southern shore, but the draw was dried out by dropping water tables, scoured bare by wind, and in the sixties robbed of its prehistoric contours and no doubt many of its fossil treasures by a gravel mining operation. The miners, however, had helped uncover what the Portales museum boasts was “the largest known mammoth kill site in North America.”

Five mammoths died here, killed—it seems likely—by the Clovis points that were strewn through the bone heaps. One mammoth, known as Big Momma, was found with four of those points, all of them lying in and around her rib cage.

How do you kill a mammoth if you’re a prehistoric hunter? Various modern-day experiments have shown that Clovis technology was adequate to the task. Spears using replica Clovis points, for example, have penetrated the hide of living elephants from 65 feet away. But if mammoths traveled in nursery herds like elephants—as the Waco Mammoth Site seems to confirm—then they would also likely have been guarded by ill-tempered matriarchs who would have made any approach a lethal risk. In order to strengthen the odds in their favor, hunters may have dug pit traps or taken advantage of places like Blackwater Draw, where the mammoths, as they cooled themselves in the water of a now-vanished lake, would have been crucially less maneuverable.

“The mammoths that were killed here,” Crawford said, “look like they were killed facing out toward the water. So the people here may have had a system—waiting for the mammoths to wade out into the water before they attacked.”

Blackwater Draw is one of several drainages that cut through this part of the High Plains and provided a route for those ancient peoples as they ventured into the Brazos River valley to hunt or trade. It is still possible today to follow the draw all the way to Lubbock, a hundred miles away, where it joins with Yellowhouse Draw at the northwest side of town.

Highway 84, southeast out of Clovis, follows roughly the same route. I took the highway to Lubbock, and when it hit Loop 289 I found myself at another remarkable archeological site, the Lubbock Lake Landmark. Lubbock Lake is hard to miss, mostly because of a full-size statue of a mammoth and her calf that overlooks a small museum and a walking trail leading past active archeological digs.

Lubbock Lake is a national historical landmark, a dry river valley like Blackwater Draw in which the exposed strata contain artifacts that neatly trace the human occupation of this part of the country from the earliest Paleo-Indians to nineteenth-century buffalo hunters. And, like Blackwater Draw, it may well be a kill site. The three mammoth carcasses that have been found in the deep Clovis sands here show evidence of human butchering when they were still freshly dead.

Eileen Johnson, the director of the Landmark, who has worked at the site since 1972, met me in her office. With her curly dark hair, oversized glasses, and pleasant academic mien, she did not seem like a woman whose research interests include cutting up big game with stone tools. (“I’ve only butchered one elephant,” she said. “I’d like to do more.”) She said that the mammoths found at Lubbock Lake had definitely been butchered, but it was still an open question whether they had been hunted or merely scavenged.

“If an animal has been scavenged after it has stiffened,” she explained, “the cut marks on the bone are different—they’re wider, they’re deeper. You may also see far more marks, because it’s hard work. The carcass when it’s fresh is easier to butcher.”

Clovis hunters didn’t just use mammoths for meat, she told me. They regarded a pile of mammoth bones the same way they regarded an outcropping of flint, as a quarry to be exploited. They would take a leg bone, say, break it over a rock anvil, then select pieces to carry off to their camps to fashion into bone foreshafts for spears or into cutting implements that could be as sharp as stone knives.

Lubbock Lake and Blackwater Draw are two of the three possible mammoth kill sites on the Southern Plains. The third is in a wheat field a few miles outside Miami (pronounced “Miama”), a little railroad town tucked into the Panhandle Plains northeast of Amarillo.

“Hold still a minute,” Larry Kaul told me the next day, as I struggled in the grip of a barbed wire fence that I had tried to climb, following his nimble lead. He carefully released the barbs, freeing me to fall into the dry chaff on the far side of the fence. I did my best to stanch the bleeding in my palm as we walked out into the wheat field. Kaul, springy and seventy-ish, owns a pest control business in Miami. Among his many jobs in the past was trapping coyotes for the Russian fur trade. (The Russians liked coyote fur, he said, because it doesn’t collect frost.)

Kaul is also an amateur archeologist who has been looking for Indian artifacts since he was a teenager. He was present in the nineties when archeologists returned to this field to further investigate the original discoveries that had been made sixty years earlier.

“This low spot here is the center of it,” he said, coming to a stop about a hundred yards before the farmland dropped down into a landscape of spidery breaks. This was where, in the early thirties, a farmer plowing deep furrows to keep the soil from blowing away in the Dust Bowl struck something solid. The bones he dragged up turned out to be those of at least a half dozen mammoths that had died in what had once been a muddy sink. The evidence that humans had killed them came in the form of Clovis-era stone tools, including a broken spear point lying less than three inches away from a mammoth vertebra.

“The bison wallowed in these little lakes and the mammoth did too,” Kaul said. “What happened here was the mammoth were pretty well half-mired in the mud and those Paleo people killed them with spears. That would be my opinion of it, anyway.

“This is a hide-working tool right here,” he said, handing me a sliver of bone he had just picked off the ground. Then he found what he said was a piece of enamel off a tooth, then a piece of worked Alibates flint.

We drove in Kaul’s pickup back to Miami and parked in front of the Roberts County Museum, where some of the bones found at the site are displayed in a kind of tableaux that includes a stuffed sandhill crane and tufts of native grass. A backdrop painting depicts two Clovis hunters with spears in hand observing a mammoth as it approaches them across the infinitude of the broken plains.

Emma Bowers, who runs the museum, directed my attention to the mammoth vertebra that had been found near the spear point. She said the point is a replica. The original had been stolen. “We had the spear point in this case and all of a sudden it came up missing.”

But she was proud of the display, secreted away in a back room of the surprisingly expansive museum, in a belowground extension that she described as a tornado-proof “fraidy hole.”

“Now look at that mammoth in that mural,” she instructed. “Watch his eye. When you move, that eye follows you.”

The mammoth’s eye did seem freakily all-observant, in a probably unintended trompe l’oeil way. It gave the impression that this great beast still regarded itself as the presiding entity of the landscape, even though mammoths had disappeared from North America about 10,000 years ago, trailing along to extinction as the earth warmed. No one knows what the tipping point was. Mammoths may have died out from attrition as their browsable paradise of open grasslands and spacious woodlands gave way to dense forests. Or they may have been reduced to smaller and smaller pockets of congenial habitat, then hunted to extinction by a surging population of Paleo-Indians.

Mammoths are gone, but the human imagination is a long way from relinquishing them. As I stared at the mural in the Roberts County Museum, I remembered another depiction of a mammoth eye I had seen, in a photograph of a startling millennia-old painting from Rouffignac Cave, in France. It was an image that was as haunting to me as an adult as that episode of Science Fiction Theatre had been to me as a child. In the painting, the mammoth is shown in profile, his right eye fixed on the viewer with a searching, intoxicating scrutiny. There is nothing prehistoric about it. It feels timeless and immediate, as if this long-extinct mammoth is looking at you, assessing you as you stand there in your own pocket of time, holding you in an unbreakable gaze as deep and ageless as a trance.