It is one of the most evocative traces of humanity's ancestors ever found, a trail of footprints pressed into new fallen volcanic ash some 3.6 million years ago in what is now Laetoli, Tanzania. Discovered in 1978 by a team headed by Mary Leakey, the Laetoli footprints led to the stunning revelation that humans walked upright well before they made stone tools or evolved large brains. They also engendered controversy: scientists have debated everything from how many individuals made the prints to how best to protect them for posterity. Experts have generally come to agree, however, that the tracks probably belong to members of the species Australopithecus afarensis, the hominid most famously represented by the Lucy fossil. Now new research is calling even that conclusion into question.

The case for A. afarensis as the Laetoli trailblazer hinges on the fact that fossils of the species are known from the site and that the only available reconstruction of what this hominid's foot looked like is compatible with the morphology evident in the footprints. But in a presentation given at the American Association of Physical Anthropologists meeting in April, William E. H. Harcourt-Smith of the American Museum of Natural History and Charles E. Hilton of Western Michigan University took issue with the latter assertion.

The prints show that whoever made them had a humanlike foot arch, and the reconstructed A. afarensis foot exhibits just such an arch. So far, so good. The problem, Harcourt-Smith and Hilton say, is that the reconstruction is actually based on a patchwork of bones from 3.2-million-year-old afarensis and 1.8-million-year-old Homo habilis. And one of the bones used to determine whether the foot was in fact arched--the so-called navicular--is from H. habilis, not A. afarensis.

To get a toehold on the Laetoli problem, the researchers first compared the gaits of modern humans walking on sand with two sets of the fossil tracks. This analysis confirmed that the ancient footprints were left by individuals who had a striding bipedal gait very much like that of people today. The team then scrutinized naviculars of A. afarensis, H. habilis, chimpanzees and gorillas. The dimensions of the H. habilis navicular fell within the modern human range. In contrast, the A. afarensis bone resembled that of the flat-footed apes, making it improbable that its foot had an arch like our own. As such, the researchers report, A. afarensis almost certainly did not walk like us or, by extension, like the hominids at Laetoli.

But according to bipedalism expert C. Owen Lovejoy of Kent State University, other features of the australopithecine foot, such as a big toe that lines up with, rather than opposes, the other toes, indicate that it did have an arch. Even if it did not, Lovejoy contends, that would not mean A. afarensis was incapable of humanlike walking. "Lots of modern humans are flat-footed," he observes. "They are more prone to injury, because they lack the energy-absorptive capacities of the arch, but they walk in a perfectly normal way."

For their part, Harcourt-Smith and Hilton note that a new reconstruction of the A. afarensis foot built exclusively from A. afarensis remains is needed to confirm these preliminary findings. As for identifying the real culprit, if A. afarensis did not make the prints, that would put the poorly known A. anamensis in the running. But just as likely, speculates Harcourt-Smith, an as yet undiscovered species left the prints. That is to say, consider the world's oldest whodunit an unsolved mystery.