1. Sixty-three traction injuries of the brachial plexus in adults are reviewed. Most of the patients were seen at regular intervals for more than three years after injury.

2. The mechanism of injury is described. Forcible separation of the head and shoulder is the essential factor, but the type of lesion is determined by the position of the upper limb at the time of the accident.

3. In traction injuries the main damage is intraneural, and the lesions are of considerable extent. Extraneural scarring is a conspicuous feature of old injuries, but it does not cause any damage to uninjured parts of the plexus.

4. The prognosis of each type of lesion of the plexus is discussed. Satisfactory recovery occurs in most lesions of the upper three roots. Degenerative lesions of the whole plexus never recover completely. Cases with Horner's syndrome always have severe residual paralysis.

5. Conservative treatment is advocated for traction injuries of the plexus and evidence is cited against early or late operations on the plexus. Reconstructive surgical procedures are sometimes indicated.

1. In the first 1600 of the series of peripheral nerve injuries treated during the years 1940-45 at the Peripheral Nerve Injuries Unit, Oxford, 52 per cent. were due to penetrating wounds, and 6 per cent. of these involved the brachial plexus.

2. A specific method of grouping and grading recovery is described: Group I—lesions of the roots and trunk of C.5, 6; Group II—lesions of the posterior cord; Group III—lesions of C.8, T.1, and the medial cord.

3. The recovery of cases in Group I was good, in Group II fair, and in Group III poor.

4. The brachial plexus was explored on twenty-two occasions. Only in four instances was there interruption of continuity—which is in striking contrast with the frequency in more distal degenerative nerve lesions in open wounds.

5. There is no evidence to suggest that neurolysis influenced motor or sensory recovery, but on one occasion it relieved persistent pain in the limb. In general the correlation between operative findings and prognosis was not precise.

6. With one exception, no recovery of consequence took place in the small muscles of the hand after a lesion in continuity, although galvanic stimulation to all paralysed muscles had been given regularly.

7. There is evidence to suggest that innervation of the biceps may not always be limited to C.5 and 6 and that in a proportion of cases C.7 may make a substantial contribution.

8. In five cases (12 per cent.) Horner's syndrome was present; the causation is discussed.

9. In every case of damage to a main vessel there was a marked tendency to joint stiffness, particularly in the hand, but on one occasion only was there evidence of ischaemic changes in the muscles and nerves.

10. The good spontaneous recovery which occurred in Groups I and II; the poor recovery in Group III even in cases where there was an apparently favourable lesion in continuity; the rarity with which division of nerves was found at operation; and the discouraging results of repair in three cases; make it necessary to conclude that routine exploration of open wounds of the brachial plexus is neither profitable nor justifiable.

The use of the upright position of the patient, and a vertical incision for exposure of the brachial plexus, has been attended by no complications or serious hazards. The aid of skilled anaesthetists is acknowledged. The structures are visible and accessible; the operative field is steady; bleeding is controlled easily; and dissection is facilitated. By this technique it has been possible to explore a larger field from above, and division of the clavicle has seldom been necessary. Finally, and of importance, the operative area at shoulder level enables the surgeon to continue tedious dissection for some hours comfortably, and to escape postoperative postural complications in his own back.

Two cases are described in which a traction lesion of the brachial plexus was complicated by sensory loss and anhidrosis in the second, third, and fourth cervical dermatomes. Both patients recovered spontaneously, though in one the recovery of muscle power in the limb was incomplete. It is believed that both were examples of a traction lesion of the cervical plexus. No similar case appears to have been recorded.

After brachial plexus injuries, and other forms of paralysis of the upper limb, even when residual paralysis is very grave indeed, and even when the limb is almost flail—a forearm which can project forwards by the construction of a bone-block behind the elbow, with arthrodesis of the shoulder when necessary; a hand with fingers in the form of a claw and a thumb which opposes them; and a wrist which may perhaps be arthrodesed, but which still better can flex or extend when the forearm is pronated or supinated; together with the wise use of such muscles as are likely to gain function; may enable a patient to do very much more than he can with an artificial limb.

These statistics show that some movement was gained, with good stability and painlessness, in thirty-five of the fifty operations (70 per cent.) and that there was failure, with reankylosis, instability, or persistent pain, in fifteen (30 per cent.).

1. Arthroplasty of the knee joint should be performed only in carefully selected cases. Criteria for the operation are outlined.

2. In our experience, 70 per cent. of properly selected patients secure good or fair results. An additional 12 per cent., whose anatomical or functional results were classified as poor, preferred the movement which had been gained to ankylosis of the joint.

3. The major functional adaptation of the knee joint takes place during the first five years after arthroplasty. Several patients who had a poor range of movement after one or two years developed an excellent range by the end of five years.

4. Instability, when present, usually became apparent within the first five years.

5. Joints which were still stable at the end of five years usually remained so over a long period of time. Four patients have been traced for twenty to twenty-five years, and three have been traced for over twenty-five years.

6. Since the incidence of ankylosis of the knee joint from gonococcal and pyogenic infections has been reduced by the use of antibiotics, fewer patients are suitable subjects for arthroplasty.

1. In tuberculous disease of the hip, premature epiphysial fusion at the knee joint is due to rupture of the epiphysial cartilaginous plates consequent upon resorption of cancellous support and suppression of osteogenesis.

2. In the tibia, premature fusion is usually preceded by near-central "bulging" of the metaphysial and tibial marrow through the epiphysial plate. In the femur, epiphysial changes preceding fusion are of a fragmentary type.

3. Injury, in quiescent disease, plays little or no part in the causation of premature epiphysial fusion.

4. The factors which are responsible for these changes—local toxaemia and prolonged immobilisation—must exist for not less than two years.

5. In cases which are treated conservatively for long periods the incidence of premature fusion, with serious shortening of the limb, is so high that the wisdom of such treatment must be reconsidered.

Retrosternal dislocation of the clavicle is an unusual injury. Serious complications may arise from damage to the trachea, the great vessels of the mediastinum, the oesophagus, and the thoracic duct. Operative reduction and reconstruction of the ligaments is the most reliable treatment.

1. A series of 196 fractures of the patella has been reviewed.

2. The treatment adopted was excision of the whole bone—l0l; excision of part of the bone—33; open reduction and suture—18; suture with later excision—14; no operation—30.

3. The average time of post-operative disability varied from 3·6 to 5·3 months. The time was appreciably less when operation was carried out before the fourteenth day than when it was done later.

4. An attempt was made to follow up, two to five years after injury, those patients in whom the result was not influenced by other major injuries of the limbs or by unexpected complications. Replies to questionnaires were received from 116 patients.

5. Of these, all regained a good range of movement, varying from 90 degrees of flexion to full movement, whether treatment was by excision of part or all the bone, or by open reduction and suture.

6. The late results of excision of the patella, as estimated by the patients themselves two to five years after treatment, showed that there was considerable residual disability.

7. After total excision of the bone only 5 per cent. of patients considered that the knee was normal; 90 per cent. complained of aching; 60 per cent. complained of "giving way." After excision of one fragment, about half the patients regarded the knee as normal and half complained of aching and stiffness.

8. The number of fractures in this series treated by accurate internal fixation was too small to make justifiable comparisons.

9. The indications for non-operative treatment, open reduction and accurate internal fixation, excision of one fragment, and excision of the whole bone are discussed.

10. Excision of part or all the patella is often inevitable, but some claims made in the past for the results of this operation are not substantiated.

1. The arterial supply of the upper end of the femur has been studied in twenty-four children and twenty adults.

2. The arterial system was demonstrated by injection of radio-opaque material, with Spalteholz' method of clarification, and histological section of the neck and ligamentum teres.

3. The upper end of the femur is supplied by the nutrient artery of the shaft, the retinacular vessels of the capsule, and the foveolar artery of the ligamentum teres.

4. The retinacular vessels consist of three separate groups: postero-superior, posteroinferior, and anterior. These vessels are the chief supply to the epiphysis and femoral head at all ages.

5. The foveolar artery constitutes a small and subsidiary blood supply to the femoral epiphysis. In this series, it penetrated the cartilaginous or osseous head in 33 per cent. of young specimens and 70 per cent. of adult specimens. The foveolar vessels increase in size with age.

6. The site of the vascular pathology in various lesions of the femoral head is considered.

1. Muscles acting upon any joint can be divided into two principal groups: muscles of displacement or spurt muscles, and muscles of stabilisation or shunt muscles.

2. Muscles which arise far from the joint are spurt muscles; those which arise near it are shunt muscles.

3. The fibrous tendon sheaths are joint-stabiising mechanisms.

4. The lumbrical and interosseous muscles are muscles of stabilisation of the digits.

5. The arrangement of the musculature is such as to ensure a constant pressure across the joint cavity during rest or uniform movement. The necessary centripetal force during movement is supplied mainly by the shunt muscles.

6. Experiments are described to illustrate these observations.