1. Radiological, chemical and histological examinations have been made of the lumbar vertebral bodies in 100 necropsies on patients dying in a general hospital, with a view to determining the range of variation of calcium content and radiographic density in normal and osteoporotic bone.

2. Radiographs were made of sagittal mid-line vertebral body slabs uniformly one centimetre in thickness, and the radiographic density of these specimens was measured in relation to an aluminium step-wedge of one to ten units. Radio-opacity of different vertebrae ranged from four to ten units. The specimen radiographs also clearly revealed the trabecular structure and the lateral profile of the bones.

3. Calcium was chemically estimated and expressed as weight of the element per unit volume of the whole bone mass (that is, of anatomical bone including soft marrow tissue). It ranged from 38 to 102 milligrams per cubic centimetre of bone. In 75 per cent of the cases the range was 50-84 milligrams per cubic centimetre. High calcium values were mostly encountered in young adults, and the calcium per unit volume tended to diminish with age; but a wide range of calcium was still encountered in the older subjects and a better correlation with age was achieved by radiographic density. Both calcium content and radiographic density tended to be higher in the male than in the female bones at all ages.

4. The results of both calcium and radiographic density showed a smooth distribution curve, though skewed through the inclusion in the series of more older people with less mineralised bones; the absence of a double peak in these curves suggests that the examinations were made on a homogeneous population and does not indicate a separate pathological group of osteoporotic subjects.

5. Arbitrary standards must be used to distinguish osteoporotic from normal bones, since neither radiological measurement or chemical assay, nor histological assessment, reveals a point at which the two groups can be separated. In the present series it seemed to us satisfactory to regard as abnormal all bones showing a radiographic density of five or less step-wedge units, and by this standard nineteen of the 100 cases (eight male, eleven female) were deemed to be osteoporotic. Histological examination excluded other forms of bone rarefaction.

6. The regression of calcium on the density measurements proved to be statistically significant and was not affected either by age or by the number of days in bed during the last illness. A small difference between the sexes was apparent, there being slightly less calcium in female than in male bones of equal radiographic density. Provided this is taken into account, the radiographic density scale can be used to predict the calcium content of vertebral bone specimens and should prove a rapid and accurate method in a survey of osteoporosis in post-mortem room material.