In these cases, fortunately comparatively rare, in which there is reason to believe that the glenoid is chiefly involved in disease, and yet that the disease can be removed without amputation, access will be gained most easily by an incision (Plate III. fig. B.) on the posterior surface of the joint, corresponding in size and direction to the linear incision in front. This gives a much easier mode of access to the glenoid. I have seen this practised in one very remarkable case by Mr. Syme, in which the glenoid cavity and neck of the scapula were extensively diseased, while the head of the bone was quite sound. _After-treatment_ is exceedingly simple; for the first day or two the shoulder is to be supported on a pillow with a simple pad in the axilla, if there is any tendency for the arm to drag inwards; after this the patient should be encouraged to sit up and move about with his arm in a sling, the elbow hanging freely down. _Results._--Hodge records ninety-six cases in which this excision was performed for gunshot injury, of which twenty-five proved fatal, and fifty for disease, of which only eight died,--results which are more encouraging than those of amputation at the shoulder-joint for disease; though for injury the mortality is much greater than Larrey's famous Statistics of Amputation, _q.v._ p. 65. Spence had thirty-three cases, with three deaths. He generally made a counter-opening behind to get rid of discharges, and inserted a drainage-tube. Gurlt's statistics of excision for gunshot injury give of 1661 cases 1067 recoveries, 27 doubtful results, and 567 deaths, the mortality being 34.70 per cent. EXCISION OF THE ELBOW-JOINT--_In what cases should it be performed?_--1. For disease of the elbow-joint which has resisted ordinary remedies, and is wearing down the patient's strength, including caries, ulceration of cartilages, and gelatinous synovial degeneration. 2. For wounds of the elbow penetrating the joint, the prognosis both as to the patient's life and the usefulness of his arm is much better after excision than after endeavours to save the joint without excision. This is especially the case when the wound of the joint is small and punctured, but if the case is seen early and treated by free drainage, with antiseptic precautions, excision may not be required. 3. For anchylosis, in cases where after disease or injury the limb has stiffened in a bad position, especially when, with a straight elbow, the hand is rendered almost perfectly useless. _How much should be removed?_--In the elbow-joint, more than any other joint in the body, complete excision is absolutely necessary; any portion of the articular surface being left proves a source of unfavourable result. The surgeon is apt to err rather in removing too little than too much. For the removal of too little bone is, on the one hand, apt to result in long-standing sinuses, on the other, to induce anchylosis.