The findings apparently do not undermine the "out of Africa" theory, which holds that there was a relatively modern founding migration of human ancestors into Asia and Europe from Africa.

But they do suggest that there were at least two migrations rather than one — the first about half a million years ago, the other, as in the "out of Africa" theory, beginning some 100,000 years ago.

The study found that the two evolutionary links to Africa were clearly imprinted in the genes of people today. But it also detected another trace of family history in the genes of modern Homo sapiens: on their migrations, the more modern people from Africa did some interbreeding with the less modern people they encountered.

If this is correct, it means that contrary to standard "out of Africa" theory, the newcomers from Africa did not completely replace the local populations — and that the Neanderthals of Europe and western Asia, and similar nonmodern people throughout Asia, escaped complete extinction. That in turn would mean that a few Neanderthal genes must survive in Europeans and those of European ancestry, and that a few genes of some descendants of Homo erectus carry on in Asians.

"It looks very, very convincing," said Dr. John H. Relethford, an anthropologist at the State University of New York at Oneonta, who has advanced his own revised version of the standard "out of Africa" model.

Most previous studies of modern human evolution have been confined to one or two genetic trees. Dr. Templeton included those two with eight other DNA sources. They were analyzed using a computer program developed with the help of Dr. David Posada and Dr. Keith Crandall, both of Brigham Young University.

Of the 10 DNA lines of evidence, Dr. Templeton said, eight bear traces from well before 100,000 years ago — that is, before the beginning of the most recent migration out of Africa. And those eight gene regions yield statistically significant data that are incompatible with a total genetic replacement model for nonmodern humans, the researcher concluded.

If the nonmodern people had been wiped out and replaced, Dr. Templeton explained, the genetic signatures of the older migration and of older recurrent gene flow would not be showing up in people today from various parts of the world.

In the journal article, Dr. Templeton said the results seemed to agree with recent estimates that about 90 percent of the human gene trees appeared to be rooted in Africa.

By e-mail from Israel, where he is a visiting professor, Dr. Templeton explained that the research showed that "human populations in Africa and Eurasia have not been genetically isolated from one another, but rather have been interchanging genes for at least 600,000 years."

The genetic interchange, he continued, "was restricted, primarily by geographical distance, which meant that local populations could and should show genetic differences, as they do today."

"But over a long time," he added, "there was sufficient genetic interchange to insure that all humanity evolved as a single species."

Dr. Templeton's findings fall somewhere between two much-debated theories of modern human origins.

Since the 1980's, molecular biologists have produced strong DNA evidence that people living today stem from a common genetic source in Africa about 150,000 years ago or more. That research underpins the "out of Africa" theory, favored by most anthropologists.

But a few stalwarts, led by Dr. Milford Wolpoff of the University of Michigan, have continued to argue the case for multiregional origins. They contend that humans evolved around the world in different regions about the same time, often living in isolation but not without occasional infusion of outside genes. These early humans were descendants of the protohuman Homo erectus, which first migrated from Africa about 1.7 million years ago.

In the sense that the genetic legacy of humanity is disproportionately from Africa, Dr. Templeton said, the "out of Africa" model still stands, but with modification: the two waves of migration out of Africa did not completely replace the local populations.

Commenting in another article in Nature, Dr. Rebecca L. Cann, a molecular biologist at the University of Hawaii who participated in the early studies of humanity's African genetic traces, saluted Dr. Templeton's research as "strong genetic support for describing the geographical center of our species as African."

But on the issue of interbreeding with local populations, Dr. Cann seemed to avoid a direct comment. She suggested that "perhaps Templeton was overambitious in the scale of his analysis," and concluded that it would take more study "before we can settle on how to interpret the varied signals uncovered by Templeton's analysis on a global scale."

If changes have to be made in the theories, several scientists said, the result could resemble the "mostly out of Africa" model proposed recently by Dr. Relethford, of SUNY Oneonta.

"My reading is that we are mostly out of Africa," Dr. Relethford said this week. "We don't replace, but rather mix to some extent with groups that were already in existence."

Such changes, said Dr. Bernard Wood, an anthropologist at George Washington University, make the "out of Africa" theory "more pragmatic, more understandable."

Dr. Wood noted that paleontological and archaeological findings were already pointing to the likelihood of several "migration pulses" from Africa.

"The notion," he said, "that you had a pulse out of Africa 100,000 years ago and there was not sharing of a single gene between the modern and nonmodern members of the species just doesn't make a lot of sense."