1. Thirteen cases of traumatic separation of the upper femoral epiphysis have been studied. Four were previously reported and nine new cases are analysed.

2. This is a rare injury occurring in young children and is due to severe violence.

3. Separation occurred at the epiphysial plate, and severe posterior displacement was frequent. The fracture line does not penetrate the epiphysis.

4. Premature fusion, avascular necrosis or non-union (individually or together) occurred in eleven of the thirteen patients.

5. Separation of the upper femoral epiphysis is a serious injury which is likely to lead to permanent deformity.

1 . Thirty-four cases of Perthes' disease followed into adult life were reviewed twenty-five to forty years (average thirty years) after diagnosis.

2. Rather more than one-third of the patients developed hips which were good, an equal number were considered fair and about one-quarter were poor.

3. Four out of five patients were fully active and free from pain but only two out of five had hips which were radiologically good.

4. Clinical and/or radiological deterioration had seldom occurred in the last twelve years.

5. A good result in childhood is likely to be maintained with no pain and good function up to the age of forty years and perhaps longer.

1 . Two cases of osteochondritis dissecans after Legg-Calvé-Perthes' disease observed for thirty years are described.

2. Osteochondritis dissecans of the hip can remain in an apparently unchanged state for many years and in these two patients is associated with excellent function, not requiring surgery.

A study of a collected series of femoral neck fractures in seventy-one children observed for one to nineteen years shows:

1. This injury is rare but occurs in children of all ages from three to sixteen years old.

2. The fractures may be classified as transepiphysial, transcervical (the commonest), basal and pertrochanteric. Displacement was frequent.

3. The fracture usually followed severe violence, especially falls from a height or motor accidents.

4. Complications were frequent and included avascular necrosis, delayed union (seventeen cases), non-union (seven cases) and disturbances of growth at both the upper and lower ends of the femur.

5. Avascular necrosis occurred in thirty patients (42 per cent). Three patterns of necrosis are described : diffuse, localised and confined to the femoral neck. The radiographic appearances of avascular necrosis after this fracture are different from those of pseudocoxalgia (Legg-Calvé-Perthes' disease).

6. Non-union did not occur after adequate primary internal fixation or after primary subtrochanteric osteotomy.

7. The management of an undisplaced fracture presented no great problem and the results were good. A plaster spica is recommended for treatment of this fracture. Exceptionally, avascular necrosis developed.

8. Treatment of the displaced fractures (forty-nine cases) was less satisfactory. A good result was obtained from primary treatment in only fifteen patients.

9. The value of different methods of primary treatment is discussed, including manipulative reduction and immobilisation in a plaster spica, manipulative reduction and internal fixation, and primary subtrochanteric osteotomy. Manipulative reduction and immobilisation in a plaster spica is not recommended.

10. Salvage operations were required in nineteen patients. Late subtrochanteric osteotomy is of value in the management of some of these problems.

1. The results of compression arthrodesis of the ankle performed on fifty-five patients (four bilateral) are presented.

2. Fifty per cent of these ankles were immobilised for a total period of no more than eight weeks.

3. Bony fusion occurred in 91 per cent.

4. The subjective result was good or excellent in 88 per cent.

5. The theoretical disadvantages of the transverse incision are not serious in practice.

6. The best position for arthrodesis of the ankle is at, or close to, the right angle.

A study of limb shortening after poliomyelitis in 225 children in whom paralysis was confined to one leg shows:

1. The paralysed leg became shorter than its fellow in 219 patients (97 per cent).

2. The discrepancy in leg length only once exceeded three and a half inches.

3. Both the tibia and the femur were shorter than their fellows in 171 out of 184 studied (93 per cent). In only one patient was the femur alone shortened.

4. Three patterns of progress of shortening are described. No evidence was found that reduction of shortening ever occurs.

5. It is impossible accurately to predict shortening. In general, the more severe the paralysis the greater the shortening, but there are notable exceptions.

6. No relationship could be found between the amount of shortening and the incidence of paralysis of any individual muscle-group.

7. There was no significant difference in leg shortening in adult life between those who had developed the disease in the first two years of life and those who had developed it later.

8. A cold blue limb is not more likely to undergo severe shortening.

9. When the paralysis was confined below the knee the greatest shortening seen was one and three-quarter inches. When muscles both above and below the knee were involved severe paralysis may produce shortening up to three and a half inches.

10. Lengthening of a paralysed leg can occur during the first two years after the onset of the disease, but this is always a temporary phase.

11. The cause of leg shortening is unknown. In only two patients in this series was there evidence of premature epiphysial fusion.

A review of fifty cases of idiopathic pseudocoxalgia (Legg-Calvé-Perthes disease) followed into adult life for periods of eleven to thirty years (average seventeen years) after diagnosis shows:

1. In the whole series rather more than one-third of the patients developed hips which were normal or nearly normal.

2. An equal number had hips which could only have been considered "fair."

3. About one quarter had hips which gave pain and which showed marked loss of movement and gross degenerative changes radiologically.

Judged from the point of view of symptoms, the results were better than the foregoing would suggest. Three-quarters of the patients were fully active and free from pain but only two-fifths had hips which were radiologically good. It is possible that such apparently good results are unlikely to be permanent, and I hope, therefore, to continue this follow-up for another ten or fifteen years. It can, however, be concluded at this stage that an immediate good result is likely to be maintained at least until the age of twenty-five years, even though half of such patients will have radiologically abnormal hips.

There is a characteristic pattern of deformation of the femoral head and neck in the adult resulting from this condition in childhood, based on the degree of flattening of the head and shortening of the neck.

The end-results are better in adequately treated cases.

Two cases are reported in which osteochondritis dissecans developed as a late complication.