The crime scene was bare of clues. Even DNA fingerprinting scored no hits in the databases. The police had no leads. As a last resort, they sent a DNA sample to a private company -- which soon supplied the surname of the man who had left it.

Science fiction as yet, but the company, Oxford Ancestors, exists and has applied for a patent on the idea of deriving surnames from DNA. Its founder, Dr. Bryan Sykes, is a geneticist at the University of Oxford in England who has made an intriguing discovery about English surnames: many of them once had a single bearer.

Surnames came into use in England between 1250 and 1450, probably in connection with the development of inheritable property rights and trades.
"The surname acted like documentation a time when they didn't have forms," said Dr. George Redmonds, a historian of place names and surnames. It is not known how many people took the same surname, but those based on common trades, like Smith, may have had many original owners.

Dr. Sykes, who analyzes DNA to track ancient population changes, wondered if he could trace the origins of his own surname. The word comes from a type of moorland stream used to mark land boundaries. Genealogists expected that as a landscape feature it would have been adopted by many people in search of surnames. Nearly 10,000 Sykeses are registered as voters in the United Kingdom, many near the town of Huddersfield in Yorkshire.

Dr. Sykes sent out letters to a random sample of his male namesakes, asking them to send him cells brushed from the inside cheek on a cotton swab. His interest lay in the subjects' Y chromosomes, which of course are bequeathed from father to son in the same pattern as surnames, except in the case of what geneticists delicately refer to as a "non-paternity event."

Unlike the other chromosomes, the Y is transmitted unchanged and would remain identical from the Adams of the human race to all of their sons, except for the rare mutations or accidental changes that accumulate over the centuries. Because each lineage of Y chromosomes carries its own signature set of DNA changes, these mutations provide a perfect system for tracking male genealogies, and that is what Dr. Sykes looked for in the cotton swabs mailed back to him.

Only one DNA signature was common among his namesakes' Y chromosomes, Dr. Sykes reports in the current issue of the American Journal of Human Genetics. This means there was only one original Mr. Sykes, at least as reflected in today's population. The first Sykeses on record lived in the 13th century in Flockton, Slaithwaite and Saddleworth, three villages close to Huddersfield.

Sykeses who do not carry the genetic signature have presumably had a non-paternity event somewhere in their ancestry. In fact, 50 percent of the modern-day Sykeses did not have the signature.

Despite appearances, this is a remarkable testimony to the fidelity of the Mrs. Sykeses of past centuries, because it amounts to a non-paternity rate of 1.3 percent per generation. The non-paternity rate in the present English population is conjectured to be between 2 percent and 5 percent.

From court records, it seems Dr. Sykes's ancestors, at least in the 14th century, were "quite a rough lot -- always being fined for cutting down trees and stealing sheep."

"Nonetheless," Dr. Sykes said, "their wives were faithful through all this."

Asked if he was a true Sykes or an out-of-wedlock Sykes, he replied, "I'm proud to say I have the aboriginal chromosome."

Sykeses who live in the United States would not necessarily bear the original chromosome in the same proportion as their English cousins because the emigrants might not have been a representative sample, Dr. Sykes said.

He has analyzed three other surnames in the same way and found that, as with the Sykeses, all can be traced to a single bearer of the family name.

Even the ubiquitous Smiths and Clarks might trace their surnames to just a handful of very prolific early ancestors, not the hundreds that might be presumed, Dr. Sykes said.

This suggests it would be feasible to construct a library of DNA signatures linked with English family names. The library could be used forensically to provide a surname that matched a DNA sample, and for genealogy.

Dr. Redmonds, the historian, said a library of DNA matched to surnames would be useful in connecting branches of a family that lacked historical records to document their kinship. The library might be of particular interest, he said, to Americans of English ancestry seeking to Identify their origins; their Y chromosomes might link them directly to the villages where the first bearers of their surnames once lived.

"Bryan had no interest in genealogy whatsoever before this started," Dr. Redmonds said of his geneticist friend. "When I was able to take him to the precise place in Yorkshire where his ancestors came from, he was hooked."