Ultrasound scans were made of the hips of 209 neonates born consecutively over a two-week period. Of the 418 scans, 62 images were selected at random and 25 of these were duplicated to give a total of 87 scans. These static images were then presented to five experienced observers who each made nine different assessments and measurements. Interobserver and intraboserver agreement was calculated and expressed as kappa values. Our results showed poor reliability on both counts.

The results are reported of an audit of allografts collected in the first 18 months of the Leicester Bone Bank. We retrieved 161 femoral heads at primary arthroplasty of which 103 were implanted into 59 patients. There were deep infections in two recipients and wound infections in five. In two of these cases, culture of the femoral head at implantation was positive but the organisms grown were not those which caused the clinical infection. We retrieved 22 large allografts from six cadavers. Four of these were contaminated at retrieval and required irradiation for sterilisation. There has been one clinical infection in the nine large allografts implanted so far. We recommend that all bone banks undertake prospective audit to ensure that high standards are maintained and wastage minimised.

An alternative to Syme's amputation for congenital absence of the fibula is described in this paper. Nine children have had the ankle reconstructed using the Gruca technique with a very satisfactory result in eight. This procedure is not suitable for every patient and in most unilateral cases the operation can only be regarded as an interim procedure because of progressive leg-length inequality. The decision to remove the foot may be delayed and it allows childhood to be spent without resort to prostheses. However, the procedure can be considered as the definitive operation in cases of bilateral deformity.

Seven patients with Ewing's sarcoma of the pelvis were treated by chemotherapy followed by wide resection of the primary tumour. Although good function after operation is possible, survival in this series reflects the poor prognosis associated with the disease; two patients died, two are alive with local recurrence and metastases and three patients are alive with no evidence of disease.

Intra-osseous phlebography and the measurement of intramedullary pressure (IMP) have been used clinically and in experimental animals as qualitative methods of measuring blood flow in the bone. The normal phlebographic appearances in long bones are not clearly understood and the correlation between these appearances and the IMP is not known. The distal femora of 10 anaesthetised rabbits were cannulated percutaneously. The IMP was measured and phlebography performed by injecting a radio-opaque dye (Conray 280).The mean resting IMP was 33 millimetres of mercury with a range of 7 to 81 millimetres of mercury. The rate of elimination of dye from the marrow varied from less than 1 minute to 40 minutes. There was no correlation between the rate of elimination of dye and the IMP. Variation in the medullary phlebographic appearance and in the routes of drainage were noted. We concluded that the wide range of resting values for both techniques suggest that neither is a true measure of blood flow in the bone and that the results of research or clinical investigation using these techniques should be viewed with caution.