When seen soon after the accident, the existence of the fracture of the humerus is liable to be overlooked, the condition being mistaken for dislocation alone, or for a fracture through the neck of the scapula. On careful examination under an anæsthetic, however, it is observed that not only is the head of the humerus absent from the glenoid cavity, but that it does not move with the rest of the bone, abnormal mobility and crepitus are recognised at the seat of fracture, and the upper arm is shortened. The extravasation in the axilla is usually greater than that accompanying a simple dislocation, and the pain and shock are more severe. A fracture through the neck of the scapula alone is readily recognised by the ease with which the deformity is reduced, and the way in which it at once recurs when the support is withdrawn. In many cases it is only by the aid of a radiogram that an accurate diagnosis can be made (Fig. 25). [Illustration: FIG. 25.--Dislocation of Shoulder with Fracture of Neck of Humerus. (Sir Robert Jones' case. Radiogram by Dr. D. Morgan.)] _Treatment._--Unless the dislocation is reduced at once, the movements of the arm are certain to be seriously restricted, and painful pressure effects from excess of callus are liable to ensue. An attempt should first be made, under anæsthesia, to replace the head in its socket, by making extension on the arm in the hyper-abducted (vertical) position, and manipulating the upper fragment from the axilla. On no account should the lower fragment be employed as a lever in attempting reduction. When reduction by manipulation fails, recourse should be had to an open operation. The upper fragment should be exposed by an incision over its lateral aspect, and made to return to the socket by using Arbuthnot Lane's levers or M'Burney's hook, or a long steel pin may be inserted into the fragment to give the necessary leverage. Reduction having been accomplished, the fracture is adjusted in the usual way, advantage being taken of the open wound, if necessary, to fix the fragments together by plates. The best position in which to fix the limb is that of abduction at a right angle. Massage and movement should be commenced early to prevent stiffness of the joint. When it is found impossible to reduce the dislocation, it is usually advisable to remove the upper fragment. The method of allowing the fracture to unite without reducing the dislocation, and then attempting reduction, usually results in re-breaking the bone, or else in failure to replace the head in the socket, and has nothing to recommend it.