Brisbin acquired more of these feral pariahs, which he'd taken to calling Carolina dogs, from shelters and from the wild in South Carolina and Georgia.  A number of them came from land surrounding large federal reservations, like Fort Gordon near Augusta, where much of the terrain remains undeveloped and undisturbed.  Brisbin kept his Carolinas in a complex of kennels in an 18-acre enclosure where they could roam in the woodlots and fields.  They started producing puppies--and surprises.
   Regardless of their origin, the Carolinas did things never before observed in domestic dogs.  They had peculiar breeding cycles, starting with a rapid-fire, thrice-annual estrus in young females (perhaps a way to ensure quick breeding before diseases like heartworm took their toll) that later settled into seasonal reproductive cycles peaking in spring and late summer-the period, Brisbin notes, when small mammals are most abundant. Some pregnant females dug elaborate underground dens in which to give birth, unlike most domestic dogs, porch or into a handy pile of brush. When she was in estrus or after her puppies were born, a captive female would carefully cover her excrement by scraping sand over it with her nose.
    To Brisbin's bafflement, the animals also dug what he calls "snout pits, "hundreds of small, conical depressions in the dirt that exactly fit the dogs' muzzles.  Most snout pits are dug by females, between September and January.  The dogs seem to be eating something, "but when I pull them out and look, there's never anything there," Brisbin says.    They are particular about where they dig, and Brisbin can only conclude they're eating the soil itself, perhaps for its minerals.
    With time, Bris became convinced he had stumbled onto something unique.  The size, shape, color and behavior of the Carolina dogs, so similar to the the traits of other primitive canids, suggested they might be a relic of the first dogs to enter the region.
   He compared their skulls with those of Indian dogs from 2,ooo-year-old archaeological digs at Savannah River; they were similar, but there was too much individual variation among the fossils to be certain. But the fact that Carolina dogs are most often found in wild, swampy, sparsely settled regions, instead of more heavily populated areas where stray dogs are most common, is a strong indication to Brisbin that these are more than just mongrels. 
   Others agree; Bris has convinced both the American Rare Breeds Association and the United Kennel Club to recognize the Carolina dog. As with any registered breed, there is now a Carolina dog studbook to document and control breeding, and Brisbin's animals have even started winning "best in show" at multiple-breed dog shows. They are, of course, individuals. There are Lucy and Cici, both captured as wild pups, who vanish down their den holes before a stranger even steps from Brisbin's car. Dibble, the dominant female of the pack, is a bit standoffish, but Bo Pup, an adolescent, is all over me in seconds with openmouthed greetings. Surrey is a medium-sized female, daughter of Horace and Marion; Morgan is a solid, friendly chap whose drop ears belie his Carolina dog genes. If Marion is emblematic of a pariah dog's ancestral shyness, Taz is the polar opposite: a Carolina dog deeply, passionately, enthusiastically in love with people. He's Brisbin's home dog, a white-and-tan whose parents were both taken from the wild. Bris is bringing him up through the ranks of obedience training, aiming for the highest level. On the lead, Taz reacts instantly to hand and voice signals, but when the leash comes off as we walk in the woods, he vaults up steep banks with ease, and scrambles out on fallen tree trunks that span 20-foot-deep gullies.
   On an oppressive June evening, with the temperature still above l00 degrees Fahrenheit and the sun a fierce red ball in the western sky, I walk with Bris and a dozen or so Carolinas across Banbury Cross Farm, near Aiken. The dogs lope easily ahead of us, the low sun backlighting them, creating little moving nimbi of gold. Billy Morgan Benton, a big, outgoing man with dark hair slicked back wet under his cap, whistles to his pack, he and farm owner Jane Gunnell breed the dogs here, working with Brisbin to establish the Carolina dog as a domestic pet. They are frolicsome, friendly animals, Gunnell says, but out in the wooded nooks and weedy fields of the farm, their primitive traits become obvious.
   The dogs are ginger, a red-brown that fades to pale buff on the flanks and belly, the same color as fallen pine needles and dead grass. They fan out through a low scrubby field, moving into the damp breeze, zigzagging and coursing with their noses low and their curved tails at half-staff.
   Suddenly one dog makes a sidestep, its supple neck arching; the tail snapping high, the longer, whiter hairs along its underside flaring, reminding me of a deer. The effect on the rest of the pack is electric. Within seconds, all the dogs converge on the spot, tails moving like semaphore flags. One plunges its head into the grass with the speed of a heron's stab, but it misses. Something small skitters through the weeds, and another dog leaps, coming down with mouth and front feet together, a predatory exclamation mark-Bam! There is a tiny squeak, and the mouse vanishes in one gulp. "I would suggest that you're watching a reenactment of a dog pack out hunting 8,ooo years ago," Brisbin says. While he doesn't claim that Carolina dogs are direct, genetically pure descendants of the original dogs that crossed the land bridge, he believes that they re-create their look and behavior.  "The Carolina dog is a hypothesis, he says. "A hypothesis that there still exists in certain parts of the United States, most likely in relative uninhabited broad expanses of natural habitat within the Southeast, remnant groups of dogs whose morphological, behavioral, ecological and genetic traits my approximate those of the first dogs to enter North America.