First, with regard to compression of the cord, the possible sources are three; (i) extra-dural haemorrhage, which may, I think, be dismissed with mention as rarely capable of producing severe symptoms. (ii) The displacement of bone fragments. This is of less importance than in civil practice, because an injury by a bullet of small calibre, capable of seriously displacing fragments, has probably at the same time produced grave changes in the cord. In the presence of severe immediate symptoms we may tentatively assume that a simultaneous destructive lesion has been produced. In such injuries pain, combined with a tendency to improvement in the paralytic symptoms and return of reflexes, is the only point in favour of bone pressure, unless considerable deformity of the spinal column can be detected by palpation or examination with the X-rays. (iii) Pressure from the bullet. This is the most important form of compression, because the mere fact of retention of the bullet is evidence of a low degree of velocity, and therefore opposed to the existence of the most severe form of intramedullary lesion. In a case of apparent transverse lesion with retained bullet, shown to me at No. 3 General Hospital by Mr. J. E. Ker, the pain was very severe, and so greatly aggravated by movement that an anaesthetic had to be administered prior to the renewal of some necessary dressings. The general condition of this patient precluded a projected operation, and after death the bullet was found to be pressing laterally upon a cord not materially altered on macroscopic inspection. In the case of retained bullet recorded (No. 104), the slight degree to which the severed ends of the cord appeared altered has been already remarked upon. Beyond this we are helped by the position of the aperture of entry, and its shape, as evidence of the direction in which the bullet passed, the presence of pain, and positive proof may be obtained by examination with the X-rays. Lastly, we come to the discrimination of total or partial section, destruction by vibratory concussion or contusion, and severe intramedullary haemorrhage. Except in the case of partial section with localised symptoms, which must be rare, I believe this to be impossible from the primary symptoms, although some indication of possible encroachment on the canal may be obtained from careful consideration of the course of the wound, as evidenced by the position and shape of the openings, the position of the patient's body at the time of reception of the injury being taken into consideration. Later we may get some aid from the possible improvement in the symptoms in the case of haemorrhage. In cases with signs of total transverse lesion, however, the discrimination of the conditions is of little practical importance, since either is equally unfavourable and unsuitable for surgical treatment.