I saw no instance of wound of the _neural arch_ from a direct shot in the back in any of our men, neither was I ever able to detect an injury to the articular processes as a localised lesion. Injuries to the _centra_ were very frequent, but differed extraordinarily in their importance. Perforation by bullets travelling at a relatively low grade of velocity, but still one sufficient to allow them to pass through the body, produced in many instances no symptoms whatever when the track did not lie in immediate contiguity to the spinal canal or perforate it. In all the wounds which I had the opportunity of examining post mortem, the fracture was of the nature of a pure perforation of the cancellous tissue of the centrum, with no comminution beyond slight splintering of the compact tissue at the aperture of exit. In one instance the bullet passed in a coronal direction so close to the back of the centrum as to leave a septum of only the thickness of stout paper between the track and the spinal canal. In this case signs of total transverse lesion were present. I never happened to meet with a case in which the canal was encroached upon from the front by displaced bone. In some cases at the end of six weeks there was difficulty in determining the position of the openings, and section of the bone was necessary in order to assure oneself as to the direction of the track. In some instances the centra were pierced in the coronal direction with varying degrees of obliquity; in others the direction was more sagittal; in two of the latter the bullet was retained in the spinal canal. The tracks were sometimes confined to one vertebra, but often implicated two. In others the bullet passed longitudinally through the thorax, grooving or perforating one or more centra. The accompanying evidences of nerve injury varied from nil to those of pressure or irritation of the nerve roots, transient signs of spinal concussion, signs of contusion and haemorrhage, or to evidence of total transverse lesion. Instances of all these conditions will be quoted under the heading of injuries to the cord or nerves. _Signs of injury to the vertebrae._--Separation of the spinous processes was often indicated by slight deformity, either evident or palpable, local pain, tenderness, mobility, and crepitus. In some cases these local signs were reinforced by evidence of cord injury. Fractures involving the laminae differed merely in the degree to which the above signs were developed. Fractures of the transverse processes were generally only to be assumed from the position and direction of the wounds, the assumption being sometimes strengthened in probability by evidence of injury to the cord and nerves.