A technique is described by which lithium fluoride powder is introduced into the marrow cavities in specimens of human trabecular bone to determine the excess photoelectron dose to marrow, when bone is irradiated by X rays of energies between 20 keV and 140 keV. Three specimens of trabecular bone, containing respectively 10, 15 and 25% bone by volume, were investigated and the results compared with those derived on the basis of earlier calculations for mono-energetic electrons by Whitwell. Reasonable agreement was found between the experimental and theoretical results, although there was some indication that scatter influenced the practical measurements at the higher photon energies. Theoretical calculations are then used to derive photoelectron dose enhancements for complete bones from the measured results on the bone specimens, and mean enhancements of the marrow dose for the whole human skeleton are calculated for subjects aged 44, 9 and 1.7 years.

Using techniques of bone scanning and ashing, the areas of the endosteal surfaces in cortical and trabecular bone have been determined for the proximal, mid and distal thirds of each of the six long bones of an adult human subject. The relative frequency of occurrence of bone sarcomas, scored as to site, has been analysed in relation to these measured areas. Data on tumour occurrence have been drawn from three sources: radium-case data from Rowland and Keane (33 cases), naturally-occurring cases from series by Sissons (139 cases) and by Dahlin (473 cases). A strong correlation is demonstrated between tumour frequency and trabecular area, but correlation with cortical area is poor. By comparing the tumour frequency in the mid thirds of the bones with the total recorded it has been possible to show that the probability of tumour occurrence per unit area of cortical bone, relative to that of trabecular bone, is 0.16 +/- 0.06. Analysis of the available dose data for the radium cases shows that in this instance dose has not contributed to the observed correlations. The results lend support to the thesis that tumour occurrence depends on surface area, i.e. on the number of cells at risk.

Techniques are described that have been developed to contaminate post-mortem specimens of human trabecular bone uniformly and artificially with three different radionuclides, 14C, 45Ca and 32P, so as to produce either a uniform volume deposition of radionuclide through the bone matrix or an 'infinitely' thin layer of radionuclide on the trabecular surfaces. Uniform volume depositions of 45C and 32P were produced by neutron activation and the 14C volume distribution was successfully achieved by rehydrating the bone matrix with 14C-labelled urea solution. Surface depositions of isotope were attempted by introducing a radionuclide into solution. Surface depositions of isotope were attempted by introducing a radionuclide into solution which was in equilibrium with trabecular bone and depositions approximating to an 'infinitely' thin layer of isotope were achieved for 32P and 14C using adenosine triphosphate-labelled molecules. Ionic calcium penetrates the bone matrix very rapidly and the best 45Ca surface deposition that could be produced penetrated the trabecular surface to an average depth of about 10 micrometer. The uniformity of deposition was analysed by autoradiography and the thickness of the 'surface' layer of radionuclide was estimated using microdensitometer scans across selected strands of the autoradiograph and comparing the film density profiles thus produced with calculations.

An X-ray technique has been developed to enable reproducible and quantitative data to be obtained from radiographs taken by a diagnostic X-ray machine. Careful attention has been paid to the alignment of the specimen and the diagnostic tube and to operating conditions. The choice of X-ray spectrum is described together with the experimental geometry used to minimize the scattered radiation incident on the film. The resultant scattered radiation intensity achieved is about 4% of the unattenuated direct radiation. Careful calibration procedures using objects the same size as the unknown specimen are necessary to avoid the need for complicated corrections for the resulting variation in scatter across the film. The LiF powder packing density is determined as 1-32 g cm-2 with a standard error of 0-52%. Values of projected bone thickness as a function of position in the bone are given. The average percentage of bone in the trabecular specimen is determined as 12-7 +/- 0-5, which is in close agreement with measurements of the same parameter determined from an optical scanning technique. The relationship between average packing density or rho y and the percentage of bone at different positions in the specimen packing density increase in percentage of bone, as would be expected.