Brawn of humanity.

Estimates of human ancestors' body size based on their fossil remains have varied greatly from one study to the next. It's tough to rebuild, from remnants of their bony architecture, creatures that no one has seen in the flesh.

A new examination of fossils from various sources indicates that, on average, the body mass of Homo species gradually increased between 1.8 million and 36,000 years ago and exceeded that of modem humans by nearly 13 percent. In addition, large boosts in brain size relative to body size occurred in the Homo lineage sometime between 600,000 and 150,000 years ago, assert anatomist Christopher B. Ruff of Johns Hopkins Medical Institutions in Baltimore and his colleagues.

Ruff's group applied two different methods of estimating body mass to 163 fossil individuals found in Africa, Asia, and Europe. One approach focused on the breadth of the knob that attaches the upper leg to the pelvis. The second method calculated body size from the pelvis' maximum width. In both cases, the measures bear a consistent relation to body size in modern human groups.

The two techniques yield comparable estimates of body mass for members of the fossil Homo sample, the researchers report in the May 8 Nature. As in modern populations, human ancestors inhabiting higher latitudes were larger than those living closer to the equator, they note.

Relative brain size in Homo remained stable from 1.8 million to 600,000 years ago, according to the scientists. Pronounced brain growth ensued sometime within the next several hundred thousand years. Although European Neandertals in the period from 75,000 to 36,000 years ago were generally 30 percent larger than modern humans, their relative brain size was nearly equal to ours, Ruff's group contends.

Modern humans have experienced parallel declines in absolute brain and body size over the past 35,000 years, the investigators maintain.

The trend toward increasing bulk in much of the Homo lineage prior to modern humans "is probably quite robust," writes anthropologist John Kappelman of the University of Texas at Austin in an accompanying comment. Still, he remarks, the new study has limitations. For instance, its fossil sample is confined largely to European and African specimens that date to no more than 200,000 years old. Moreover, Kappelman says, world-class athletes may provide a better model than folks of more modest build for estimating ancient Homo body sizes from incomplete fossil evidence.