Even in these globetrotting times, getting from Shanghai to the Gobi Desert is seen as an arduous haul. Now imagine getting there from Rome, on foot, more than two thousand years ago. As unlikely as it seems, such a migration is the most plausible explanation yet for the fair hair and eyes of Gansu's Zhelaizhai villagers. We didn't say airtight, mind you, but plausible.

True or not, it makes a heck of a story, Chinese villagers genetically linked to Roman soldiers. So does the body of evidence. You be the judge.

Ancient Roam

In 53 BCE, Roman Emperor Marcus Licinius Crassus learned that it doesn't pay to mess with Iran. Back then, Iranians were called Parthians, and literally handed Crassus his head at the Battle of Carrhae. Fish-scale formations of Roman stalwarts were reduced to pin-cushions by free-wheeling Parthian cavalry, under an unending storm of arrows. Twenty thousand of Crassus' men perished with their emperor. But some ten thousand were captured alive.

Of that group, many legions were redeployed as Parthian soldiers. When Rome and Partha reconciled some thirty years later, a prisoner exchange followed. However, the exchange failed to account for all captured legions. The missing Romans wandered off the charts of Western history.

They remained in obscurity until 1955, when Oxford's Sino-expert Homer Dubs presented his research to London's China Society. Citing ancient Chinese historical texts, he claimed that some of the lost soldiers had wound up as mercenaries for the Kangju, Hun antecedents. The Kangju spun them off to anti-Han confederates the Xiongnu, whose leader Zhizhi Chanyu suffered a spectacular loss to Chinese forces in 36 BCE.

Sorry for the twists and turns; here's the upshot: Dubs found a passage in the Han Shu, a contemporaneous history, describing the odd fish-scale formation Zhizhi's troops used in the battle. The account also remarked on the odd double palisade of wooden stakes ringing Zhizhi's fortress walls, a structure thought to have been employed exclusively by Romans. Dubs concluded that the vanquished prisoners had been led inland and left as frontier guards by the Chinese at the site of today's Zhelaizhai. The appearance of a new county called Liqian in China's 5 CE government register lends credence to the hypothesis - Liqian was the ancient name for the Roman Empire.

More Clues

Scholar Guan Yuquan devoted the last twenty years of his life to writing a huge but unpublished book corroborating Dub's theory and adding further evidence, none indisputable. However, he pointed to official documents from 9 CE, recording that then-emperor Wang Jang renamed Liqian, calling it Jielu, "Prisoners taken in storming a city". In 1989, Australian writer David Harris became intrigued with Dub's theory and ventured to Gansu in search of Liqian. Aided by Guan, he arrived at Zhelaizhai village as the location. His book "Black Horse Odyssey" details the official scrapes and other challenges he encountered on the quest.

Other clues unearthed in the area - a Roman-style pot and bowl, as well as a helmet with the characters for zhao an, "one of the surrendered" - have bolstered the hopes of nearby Yongchang city, which is invested in a Roman heritage. Its museum exhibits Han-era skeletons far larger than average for Chinese at the time, unearthed in 2003 and discovered facing westward, a burial practice otherwise unheard of in Han custom.

Naturally, DNA testing would go far in settling the matter. DNA studies of blood samples from area residents were launched in 2005 and 2007, but as of yet have proven inconclusive.

Other Cauc - Asians

Considering the historically strategic location of Zhelaizhai and Yongchang, there could be any number of ruddy, big-nosed ancestors lurking in the locals' genes to explain their western features. Perhaps Liqian was originally a Greek settlement. Remnants of Alexander the Great's empire still had a significant presence in Central Asia when the Parthians rose to power. Varying historical accounts place the last Greek king of Kabul between 40 BCE and 50 CE, proving the viability of Greek culture and military power at the time. Maybe it was Greek mercenaries who contracted out to the Xiongnu, to be defeated with Zhizhi and left to dry in remote Gansu. After all, Liqian sounds quite a bit like Yaliqian, the traditional Chinese name for Alexandria.

Also not to be discounted are the 4,000 year old mummies of the nearby Tarim Basin in Xinjiang. Indisputably Caucasian, they are believed to be an off-shoot of the multifarious Celts. Celtic heritage would go a long way in accounting for not only the locals' long straight noses and light eyes, but also their fascination with bulls, a sacred animal in the Celtic world.

Let's not forget that both Yongchang and Zhelaizhai lie hard by Silk Trade routes from Central Asia into China. There are any number of scenarios by which traders or other travelers of western origin may have settled into a stationary rut in the region.

Render Unto Caesar

Whatever the reason for the western appearances, they make Yongchang special. Cities trade on their distinctiveness, and in hardscrabble Gansu, lack of hard scientific proof is hardly reason to ignore a boon. It started in 1993, when local party chief Jia Xiaotian realized the touristic potential in marketing Yongchang as the modern successor of a long-lost Roman outpost. At the top of the city's main street he erected a statue of three people: Han, Hui, and Roman, with a plaque commemorating Roman contributions to Yongchang.

Local industry has done its best to turn the legend into coin of the realm: a TV series, exhibits at the museum, a new hotel, even a KTV house named ‘Caesar'. Tourism hasn't exactly boomed, but officials have commissioned a Roman pavilion to get the odd visitor into the spirit of things.

It is unlikely that Zhelaizhai, the humble settlement of some seventy families most physically suggestive of Roman influence, will see much of any potential tourist windfall. But one villager has changed his name from ‘Cai' to ‘Luoma' to acknowledge his connection. Other villagers are pleased to lay claim to such a romantic heritage. If scientists determine they're wrong, will it matter? History is the version of past events that people have decided to agree upon.