Fracture of the outer or of the inner third rarely leads to deformity, and treatment can therefore be advanced more rapidly. It is often possible to allow full "underhand" use from the outset. It is always necessary to pay special attention to the structures just above the bone; as, failing this, it is not uncommon for a piece of platysma to be caught between the fragments. This is a fertile source of subsequent pain and disability, as some fibres of the superficial cervical plexus are almost certainly involved. It can be loosened only by very slow stages and with great difficulty.

Fractures of the acromion should be treated on lines similar to those outlined for fractures of the clavicle. When the body of the scapula is broken the bony injury is of minor importance when compared with that done to other structures. There need be no fear whatever of displacing the fragments, as they are held together by the muscles and their aponeurotic attachments. Mobilisation, however, should proceed apace to the limit of pain, as otherwise dense adhesions may form and the general utility of the whole arm may thus be in jeopardy.

A fracture of the upper third of the humerus rnay be impacted. Here, as elsewhere, impaction should often be respected by the surgeon; and, if this is done, the impaction may be regarded as the first stage in repair and the limb treated as if union had just taken place. Unimpacted fractures in this situation, if treated by mobilisation and massage, usually unite about the eighth to twelfth day. By this is meant that the granulation tissue, the precursor of true callus, that is forming between the fragments has become sufficiently organised to bear the strain of ordinary relaxed movements in all directions. From this point on there should be no movements of the fragments during our manipulations. If they occur we are doing too much. The patient should always be treated sitting up, preferably on a chair, which is so arranged that he leans the opposite side of the chest on the back of the chair, supports the arm on the top of the chair-back, and rests his head on his hand. The hand of the injured arm rests in a sling or on a cushion, which is placed on the knee of the same side, this being crossed over the opposite knee (see Fig. 74). Massage is carried out as for fracture of the clavicle, but a greater area of the back should be treated so as to include the latissimus dorsi. Mobilisation is required as for fracture of the clavicle, particular attention being paid to the hand movements. It is well, unl the surgeon orders otherwise, for the patient to sleep in an armchair or lounge cane chair with the feet up, and not in bed. If he sleeps in bed, it is essential that the space between the trunk and the limb should be well padded and that fixation should be secured adequately by flannel bandages. No movement should be given to the shoulder for eight or ten days unless specially ordered, and then all other movements should precede rotation. If ordered, slight movements, in every direction excluding rotation, may be given from the outset, provided the limb receives adequate support. Championnirre used to say that, thanks to mobilisation, fracture of the surgical neck of the humerus might be classed as a trivial injury. Weight and pulley exercises may be commenced in a fortnight to three weeks, and full use be allowed during the fourth week. Full strength should be regained about the eighth week.

Fig. 74. - To show the position of patient during massage after a recent fracture through any portion of the upper two-thirds of the humerus. The weight of the limb acts as an extension. Note that the whole of the neck, chest, and back can be treated without change of position. The arm has been rotated from the position in which it rests when supported by a sling. This represents a later stage in treatment than that referred to in the text.

If the fracture includes separation of the greater tuberosity, it occasionally happens that the callus formed, or even the tuberosity itself, may impinge on the acromion during abduction. The evidence consists of sharp pain just below the acromion during passive abduction. If this is noted, the surgeon should be warned so that he may have the opportunity of placing the patient on an abduction splint (see Fig. 75), in which union is allowed to take place. This ensures that there will be little interference with the movement of abduction. For some reason or other, these fractures are always more painful than those in which the shaft of the bone only is involved. It is possible that the reason is that fracture in this position involves the sub-deltoid bursa and that we have a bursitis to deal with in addition to the fracture. I know little about sub-deltoid bursitis personally, but reference to it by American surgeons is not infrequent (see Chapter XXV (Sacroiliac "Strain" And "Sub-Deltoid Bursitis")).

Fig. 75. - To show abduction splint applied.

Fractures through the middle of the shaft of the bone take a little longer to unite - about ten or fourteen day It is easier to give movement to the shoulder after this accident than if the surgical neck is broken, but elbow movements must be much more carefully guarded.