1. The long-term results in a consecutive series of 323 healed subcapital fractures of the femur show that, with few exceptions, the capital fragment maintains its integrity when the fragments are aligned within the narrow limits of good reduction, but undergoes superior segmental collapse when reduction is poor.

2. The effect of malreduction on the congruity of the hip joint is examined, and a remodelling response to malalignment of the aspherical femoral head in the imperfectly round acetabulum is proposed as an alternative interpretation of the radiological changes now considered to be the result of capital ischaemia.

1. Practical experience has shown that subcapital fractures of the femur unite freely if reduction is stable and fixation is secure.

2. Stable reduction is obtained when the muscular and gravitational forces tending to redisplace the fracture are opposed by equal and opposite counterforces, and inherent stability is believed to depend upon the integrity of the flared cortical buttress at the postero-inferior junction of the femoral neck and head.

3. In the stable subcapital fracture a state of equilibrium is reached when the forward and upward thrust of the fixation appliance in the femoral head is opposed by the counterthrust of the closely applied and cleanly broken fragments at the postero-inferior aspect of the fracture. When the postero-inferior cortical buttress is comminuted, inherent stability is lost, lateral rotation deformity recurs and the fixation device is avulsed from the cancellous bone of the head.

4. Stability may be restored by reduction in the "valgus" position, by various forms of osteotomy, by refashioning the fracture fragments or by a postero-inferiorly positioned bone graft. Theoretically, stability may also be obtained by a double lever system of fixation in which an obliquely placed fixation device or bone graft is combined with a horizontally disposed wire, pin, nail or screw crossing it anteriorly. Multilever fixation by three or more threaded wires or pins inserted at different angles and lying in contact at their point of crossing may likewise provide stability.

5. Fixation by two crossed screws has been chosen for clinical trial in 100 displaced subcapital fractures. Imperfect positioning of the screws in seven patients has been followed by early breakdown of reduction and non-union, but satisfactory positioning has been associated with radiological union in fifty patients who have been observed for twelve months or more.

6. Ultimate breakdown in some of these fractures is certain to follow avascular necrosis, and this complication has already been seen in a few patients treated by cross screw fixation more than two years ago. It is also expected that non-union will occur in some of those patients still under observation for less than a year. Even so, these preliminary findings indicate a percentage of union far greater than that obtained by previous methods of treatment, and, although statistically inadequate, they are presented in support of the belief that it should no longer be considered impossible to achieve the same percentage of union in subcapital fractures of the femur as we are accustomed to expect in the treatment of fractures elsewhere. It is not implied, however, that this ideal will be reached merely by the adoption of some form of double or multilever fixation, and much will continue to depend upon the quality of the radiographic services, the precision of reduction and the perfection of operative technique.

7. Every advance in our understanding of what is meant by "perfection of operative technique" lends increasing support to the ultimate truth of Watson-Jones's (1941) dictum: "A perfect result may be expected from a technically perfect operation; an imperfect result is due to imperfect technique." But the simple and foolproof method of fixation which will end the search for technical perfection in the treatment of the displaced subcapital fracture has yet to be evolved, and many questions remain to be answered about this injury. Nevertheless, it is clear that the surgeon should now be prepared to attribute early mechanical failure in the treatment of femoral neck fractures to his own shortcomings, and the temptation to blame capital ischaemia for every disaster should be resisted.

The successful management of femoral neck fractures is obviously based upon many factors. The forces acting upon the proximal end of the femur are believed to be mainly compressive in nature, and the low-angle nail by stabilising the fully reduced fracture in the line of these forces is held to allow weight bearing to take place. Low-angle nailing is believed to offer many advantages over conventional methods of treatment but only in the presence of stability. Stable reduction is the essential preliminary to any form of treatment, and low-angle fixation with early weight bearing in the absence of stability is regarded as futile.

It is suggested that those subcapital separations which follow trivial injury may originate as stress fractures accompanying the process of bone remodelling in the aged, and that many of these fractures may remain unrecognised and heal spontaneously. With rare exceptions, subcapital fractures are regarded as being of the same essential pattern, and their varying radiological appearance is considered to be due to the different degrees of displacement to which they have been subjected. A new classification based on this premise has been suggested.

In a series of eighty subcapital fractures the incidence of avascular necrosis was not adversely affected by early weight bearing, but reduction in the extreme valgus position was invariably followed by this disaster. This is probably also true of any malposition in extreme rotation which must stretch and obliterate the vessels in the ligamentum teres.

A rough alignment index of reduction was found to provide an almost infallible guide to the prognosis both in regard to union and to avascular change. It may therefore be possible to base prognosis on the quality of reduction before the fixation appliance has been inserted. The unsatisfactory results in those cases apparently destined to non-union or avascular necrosis may then be avoided by alternative means of treatment at an early stage. Whether this will prove to be true must depend upon a much longer experience of low-angle fixation, and, in common with almost every communication on this subject, premature publication must largely offset the value of the present findings.

Many analyses of the geometric arrangement of trabeculae in the proximal end of the femur have accepted and perpetuated the theories of Ward (1838), Culmann (1866) and Meyer (1867), and have contributed to the belief that the structure of the femoral neck embodies mechanical principles which are foreign to bony formations elsewhere. This isolated departure from the normal pattern of skeletal behaviour is considered to be most unlikely, and an attempt has been made to show that the structure of the femoral head and neck departs but little from the normal anatomy of the long bone.

From a developmental point of view, the proximal end of the human femur is believed, in its simplest interpretation, to represent an upward continuation of the original shaft which has undergone rotation and expansion. The cancellous arrangements of the internal weight-bearing system are likewise believed to represent the expanded and rotated lamellae of the neck as they are presented radiologically or on coronal section.

The forces acting upon the proximal end of the femur are considered to be mainly compressive in nature, and both crane and street-lamp bracket theories have therefore been rejected. The spiral conformity of the proximal end of the femur has been related to the spiral disposition of the soft-tissue structures that surround the hip, and in the interpretation of hip joint mechanics the principle of the screw has been preferred to the principle ofthe lever.

1. Tennis elbow is largely an affliction of middle age, and it is believed that degenerative changes in the orbicular ligament may underlie its pathology.

2. It has been shown that most patients may be relieved by either extra-articular or intra-articular injection of hydrocortisone, and operative intervention is required only in a minority of cases.

3. Contraction of the extensor carpi radialis brevis is considered to be the principal pain-producing factor, and Z-lengthening of the tendon of this muscle has been found to relieve the symptoms when conservative measures have failed.

4. The late results in fifty patients have shown that this operation causes diminution neither of the power of wrist dorsiflexion nor in the efficiency of the grip. The operation may therefore be undertaken with every prospect of relieving the discomfort of tennis elbow without inviting alternative disability.