1. Experimental avascular necrosis of the femoral head was produced in rabbits and the histological and radiographic changes were compared.

2. Avascular bone which was not re-ossified or altered in any way showed no change in density to x-rays.

3. Areas of avascular bone which were being repaired by the formation of appositional new bone showed an absolute increase in density in the radiographs. This is apparently due to simple increase in bulk of bone to be penetrated by the x-ray beam. On this basis, increasing density in radiographs of bone suspected of being dead is a sign of increasing re-ossification rather than of increasing necrosis.

4. Re-ossification of dead bone occurred rapidly in the absence of simultaneous resorption of necrotic trabeculae. It is thus suggested that the term "creeping substitution" is misleading and does not reflect accurately the histological findings, at least as they occur in rabbits.

1. The methods by which epiphyses receive their blood supply was studied by means of India ink injections in monkeys. Two types were identified depending upon whether the epiphysis was entirely or partly covered by articular cartilage. In the former, nutrient vessels enter the epiphysis by traversing the perichondrium at the periphery of the plate. In the latter they enter the epiphysis by penetrating the cortex at the side of the epiphysis at a point remote from the epiphysial plate.

2. The histological changes after separation of the second type of epiphysis were also studied. After temporary interference with endochondral ossification marked by increased thickness of the epiphysial plate, healing occurred so rapidly that within three weeks it was difficult to determine that the epiphysis had been separated at all.

3. It is concluded that when nutrient vessels enter an epiphysis at a point remote from the epiphysial plate, that epiphysis can be separated without serious disturbance to its blood supply and accordingly without interference with its capacity for growth. As it has been established that an epiphysis which is entirely covered by articular cartilage cannot be separated without destruction of its blood supply and subsequent avascular necrosis (Harris and Hobson 1956), it is concluded that the prognosis of an epiphysial separation is dependent upon the degree of damage to its blood supply rather than the mechanical disturbance of the epiphysial plate.

An experimental method is described which permits observations on the early stages of repair after acute displacement of the upper femoral epiphysis. Because the epiphysis is intra-articular, displacement brings about avascular necrosis which is slowly repaired by ingrowth of callus and blood vessels from the stump of the neck. As the bulk of the epiphysial plate remainsattached to the epiphysis, it acts as a barrier to successful revascularisation. Deliberate removal of the epiphysial cartilage allows earlier revascularisation. It is suggested that in clinical cases reduction be done through the epiphysial plate rather than through the neck, and that it be accompanied by curettage of the remaining part of the epiphysial plate from the under surface of the head.

1. An apparatus was designed to determine the shearing strength of the upper tibial epiphysis in the rat. Observations were made with this instrumenton normal animals, on animals receiving growth-hormone, and on animals receiving oestrogen.

2. When the epiphysis separates from the diaphysis, the plane of cleavage is constant, passing through the third layer of the epiphysial plate.

3. Growth-hormone decreases and sex-hormone increases the shearing strength of the epiphysial plate. These changes are due to alterations produced by these two hormones in the thickness of the third layer of the epiphysial plate.

4. It is suggested that these findings may be of significance in providing an anatomical basis for slipping of the upper femoral epiphysis in man, especially when it is associated with the adiposo-genital syndrome or with rapid adolescent growth.