Analyses of measurements taken from adult femurs of a British population

The femur remains an important bone in palaeoanthropology and this paper reports various analyses of measurements taken from long‐cassette radiographs of adult femurs of a reference population from the south of England (n=81, males 33, females 48). Discriminant analysis confirmed that the male femur is usually larger than the female femur, and that a femur may be sexed with some confidence. Discriminant scores have been derived to permit estimates to be made, from the whole or a part of a femur, of the probability of the male gender. Outcomes from a principal component analysis suggest: that the female proximal femur is a morphological unit; that neck length and shaft length are dissociated and there is no common factor representing linearity; and that the male bicondylar width is anomalous, a finding that is confirmed by other analyses. An estimate of the robusticity of the complete female, but not male, bone may sometimes be made from a proximal fragment. The robusticity of the complete bone may be made confidently from the shaft for both genders. The two width variables relating to a distal fragment cannot be used to predict the robusticity of the complete bone for either gender. A method of defining the position of the waist in terms of displacement in deciles down the shaft is described. The position of the waist seems to have a bimodal distribution. The positive association between age and distal displacement may be due to remoulding of the shaft with age or to a cohort (generational) effect. There is a weak correlation between a small neck–shaft angle and the distal displacement of the waist, two archaic phenotypic traits. Ten statements in the literature relating to the femur have been examined and tested; six are confirmed and four are unconfirmed, and it is suggested that there is a need for further studies relating to the morphology of the femur.