Ever since the 'hobbits' (Homo floresiensis) were discovered in 2003, on the Indonesian island of Flores, there has been an on-going debate about their exact relationship with our own species. One interpretation of the fossils sees them as members of our own species, with the most complete individual (LB1) having suffered from microcephaly (ie an abnormally small head). Another group of scientists, including those who first discovered and described the remains, regards them as belonging to a dwarfed species, most probably descending from Homo erectus.
Now a new paper, describing the morphology of the wrist bones from the best-preserved floresiensis specimen, appears to support the latter view (Tocheri et al., 2007). These bones, called carpals, were found close to the head of LB1 and at the same depth in the cave.
In modern humans one of these bones, the trapezoid, is shaped like a boot, but it is more wedge-like in non-human primates. This difference in shape influences the shape of the other carpal bones, and carpals in modern humans are described as 'derived' (changed) compared to the presumed ancestral state in non-human primates such as chimpanzees. What's more, these 'derived' features of carpal bones are also found in Neanderthals, and in remains attributed to H. antecessor and dated to around 800,000 years ago. This suggests that the 'modern' carpal bones evolved at least 800,000 ya.
In contrast, the 3 carpal bones from Flores don't show any modern characteristics. Instead, they resemble carpals from living great apes – and from hominin remains that include the wrist bones and date before 1.7mya (including Australopithecus). This strongly suggests that floresiensis is derived from a lineage that predates the origin of modern humans. An analysis of features of skeletal development, and also of skeletal abnormalities in modern humans, confirmed that these similarities are a reflection of evolutionary relationships and not the outcome of disease in a modern human individual.
This confirms an earlier study of the brain morphology of floresiensis (there's an article here and an enlarged image here), which found it to be more similar to an erectus brain than to one from either a normal or a microcephalic modern human. However, Tocheri and his coworkers note that their analysis can't add anything to the discussion about whether floresiensis is a descendant of erectus – because there are no known erectus carpal bones to compare with the Flores remains.
Reference: M.W. Tocheri et al.(2007) The primitive wrist of Homo floresiensis and its implications for hominin evolution. Science 317: 1743-1745