1. Evidence is presented that the basic lesion in scoliosis is relative lengthening of the anterior components of the spine compared with the posterior elements.

2. The logical treatment is to reduce this relative lengthening either by lengthening the posterior elements or shortening the anterior elements. This may be achieved by anterior lumbar wedge osteotomy or by epiphysiodesis; and correction of lumbar lordosis can improve a thoracic scoliosis.

Severe kypho-scoliosis, lateral curvature and lordo-scoliosis are ultimately caused by disturbance of vertebral growth. The results of treatment by destroying the growth potential opposite the area of growth inhibition have been encouraging. When the operation has been adequate further deterioration has been prevented; in younger children there has been improvement with further growth. It is important that the growth arrest should be at the right site and that it should be sufficiently extensive. Accurate pre-operative diagnosis of the type and extent of the curve is important.

1. Evidence is presented that certain types of cervical spine injury are due mainly to lateral flexion forces.

2. These injuries are often complicated by a brachial plexus lesion as well as a lesion of the spinal cord.

3. It is not always easy to detect the brachial plexus injury when the patient is first seen.

4. In the cases reviewed there has been little or no recovery of cord function, and the existence of a brachial plexus injury has, of course, made rehabilitation much more difficult.

5. The practical importance of recognising the mechanism of this type of injury is that treatment which will cause further separation of the vertebrae is inadvisable.

Radiographic examination of a giraffe-necked woman shows that there is no true elongation of the cervical spine. The apparent lengthening of the neck is due to incorporation of part of the thoracic spine and thorax in the neck.

1. Compression forces are mainly absorbed by the vertebral body. The nucleus pulposus, being liquid, is incompressible. The tense annulus bulges very little. On compression the vertebral end-plate bulges and blood is forced out of the cancellous bone of the vertebral body into the perivertebral sinuses. This appears to be the normal energy-dissipating mechanism on compression.

2. The normal disc is very resistant to compression. The nucleus pulposus does not alter in shape or position on compression or flexion. It plays no active part in producing a disc prolapse. On compression the vertebral body always breaks before the normal disc gives way. The vertebral end-plate bulges and then breaks, leading to a vertical fracture. If the nucleus pulposus has lost its turgor there is abnormal mobility between the vertebral bodies. On very gentle compression or flexion movement the annulus protrudes on the concave aspect–not on the convex side as has been supposed.

3. Disc prolapse consists primarily of annulus; it occurs only if the nucleus pulposus has lost its turgor. It then occurs very easily as the annulus now bulges like a flat tyre.

4. I have never succeeded in producing rupture of normal spinal ligaments by hyperextension or hyperflexion. Before rupture occurs the bone sustains a compression fracture. On the other hand horizontal shear, and particularly rotation forces, can easily cause ligamentous rupture and dislocation.

5. A combination of rotation and compression can produce almost every variety of spinal injury. In the cervical region subluxation with spontaneous reduction can be easily produced by rotation. If disc turgor is impaired this may occur with an intact anterior longitudinal ligament and explains those cases of tetraplegia without radiological changes or a torn anterior longitudinal ligament. The anterior longitudinal ligament can easily be ruptured by a rotation force and in my experience the so-called hyperextension and hyperflexion injuries are really rotation injuries.

6. Hyperflexion of the cervical spine or upper thoracic spine is an anatomical impossibility. In all spinal dislocations a body fracture may or may not occur with the dislocation, depending upon the degree of associated compression. In general, rotation forces produce dislocations, whereas compression forces produce fractures.

Three cases of ulnar dimelia, one in an adult, are described and reference is made to earlier reported cases. The etiology of the condition is discussed.

1. Some of the factors responsible for vertebral growth have been discussed.

2. In kyphosis and scoliosis it is important to prevent progressive epiphysial damage.

3. In selected cases of progressive scoliosis, epiphysiodesis on the convex side will correct unequal growth.

4. The technique of spinal epiphysiodesis is described and the results that may be expected are discussed.

The principle of Occam's razor proves nothing. Nevertheless, it is possible to explain all the phenomena of severe scoliosis on the basis of a primary rotation deformity alone. The typical rotation type of scoliotic deformity can be reproduced artificially by fitting vertebrae together in an abnormal rotatory relationship without any element of lateral flexion. From this, certain mechanical factors inevitably come into play which must tend to increase the deformity. Above all, the forces responsible for progressive scoliosis are dynamic and active, not just passive. The spine readily compensates for a passive, non-progressive deformity such as a simple wedge vertebra. It is my belief that rotation is usually the dominant factor and that correction and control of severe scoliosis can only be achieved by concentrating on the rotation deformity. I am well aware that this is an old idea but its essential truth has been insufficiently appreciated in recent years and we have not faced its full implications. Failure to correct rotation invites recurrence. Conversely, even a slight reduction in rotation usually produces a marked cosmetic improvement, often out of all proportion to the radiographic appearances.