should not be hindered. The motive has been to cure, but I am not satisfied that the expectations which have been time after time excited in this department of surgery have been fulfilled. I have only to point to the variety of plans which have been submitted to the profession during the past few years to justify my view.

The serum treatment has been abandoned, and in its place we have had the direct introduction of chemicals into the joint and the diseased synovial membrane.

Lannelongue's injections of chloride of zinc have not justified the claims made for them by that distinguished surgeon. Kolischer has admitted disappointment in the results obtained from his efforts to calcify the tuberculous structures by injecting neutral phosphate of calcium. Arsenious acid, tincture of iodine, and other substances have also had a trial. Iodoform is extensively used in Germany as a germicide. Krause reports twenty-three cures of various diseased joints out of sixty treated. I have had the most satisfactory results from the use of iodoform injections in cases of chronic tuberculous abscess. But this is an entirely different thing from injection into the substance of a greatly thickened tuberculous synovial membrane. Apart from the fact that many of these substances produce great pain, it is to be remembered that the injections must be repeated frequently at long intervals, and that the patient on these occasions has to endure some days of irritative fever, which in itself has a depressing influence on his general condition. Furthermore, the effect of the injection is limited to a small area around the point to which the needle has reached, and it is necessary to visit the already untouched parts at successive operations. At least two dangers seem to attend these proceedings. We can have no assurance that the diseased membrane has in every part been subjected to the claimed curative effects of the injection. then, certain portions have not been beneficially affected, they still remain as foci from which the disease will progress. The second danger is that outside the sphere of direct influence there must be an area of disturbance which may, and often does, advance to inflammation. In that area there are all the

If,

conditions which encourage mischief. In other words, the already diseased tissue in that zone beyond the reach of the germicide is stimulated into increased activity. So, as we ,should expect, in spite of all aseptic precautions, spots here and there inflame, cascate, and suppurate after the injections, and surely make the last condition worse than the first.

Meanwhile, what of the ligaments of the joint? One of the characteristics of this affection is the persistency and subtlety of the infective process. If we could say that disease is strictly limited to the synovial membrane, we should perhaps be encouraged in our efforts. But everyone knows that this is not so. Every ligament in the articulation is tainted. or infiltrated. Even the skin gives evidence of it, as we see so often as the result of an incision, in the unhealthy unhealing granulations that follow. Unless, then, we begin the injections at a very early stage, supposing them to be as potent as is alleged, they are not likely to do good; on the contrary, by their interrupted character, and the limited portion of tissue which each can affect, they are more likely to create a false security, while the disease is in truth making dangerous headway.

The opinion of German surgeons is in favour of this method. British surgeons have certainly not accepted it, although occasionally the results may be satisfactory; but the experience in these countries appears to be that it is an unreliable form of treatment, and that disappointment usually follows the practice of it.

But this having been tried and found wanting, we have gone a step further in England. Professor Wright, of Manchester, some twenty years ago led the way in what was deemed a further advance which would give better results than excision in the case of children. The suggestion that we might remove the disease, and at the same time preserve the mobility of the joint, is a very attractive one; and there have been many followers of his practice in all parts of the world.

Let us examine this method. It was believed at first that not only could the disease be extirpated, but that a movable

joint could be secured, and that shortening would not follow because the epiphyses were not interfered with. But experience has not found the promises fulfilled. What is the position to-day? It is quite true that movement is preserved in some cases, but the effort to secure it as a routine practice has been abandoned. We know that shortening does take place, as reported in eight of Professor Wright's cases, and that there are other drawbacks which, I believe, give the operation a secondary position when compared with excision. In erasion the surgeon removes the synovial membrane, say through two lateral incisions or one of horse-shoe outline; probably divides the ligaments to secure better access: perhaps takes away the semilunar cartilages; brings the articular surfaces of the tibia and the femur into apposition; fixes the limb, and awaits--not mobility, but rigidity. For, those who are most experienced in this operation do not make attempts at systematic passive movements, because they have learned. that they are attended by undesirable results. Now if we want to obtain union which will be solid and reliable, I do not think the best media for securing it are two articular cartilages. Mr. Howard Marsh has emphasised this by declaring that "the practical difficulty is that often no firm synostosis, but only fibrous ankylosis, can be obtained." And he says further, "While the hip, the ankle, and the elbow owe their security to the modelling of the articular ends, in the knee this security is entirely dependent on the presence of powerful ligaments. These ligaments, while they connect the bones firmly together, yet allow flexion and extension, sliding and rotation, as well as these movements in constantly varying forms of combination. By arthrectomy, not only is the largest of the synovial membranes completely removed, but the ligaments are divided. In fact, all the essential components of the joint, except the articular ends themselves, are sacrificed, and the bones are henceforth connected merely by cicatricial tissue. One is almost tempted to remark that if, after this, a useful joint remains, the original structures would appear to have been a needless arrangement" (Diseases of the Joints and Spine, pp. 335 and 337).

There can be no surprise, then, that a joint so treated, which depends for its subsequent stability upon the healing of its divided ligaments and the establishment of fibrous ankylosis. should disappoint our expectations. There being no true osseous union. the tendency to flexion is considerable, for the soft structures inevitably yield to the force of the flexor muscles. So serious is this tendency that Mr. Watson Cheyne says. that "a back splint should be worn for a long time, sometimes for years" (Tuberculous Disease of Bones and Joints, p. 189). Here the danger is that when the child has passed from direct surgical treatment, the deformity may go on increasing owing to the neglect of the parents. and the patient may returu to us with a limb bent almost at right angles, demanding excision of the bones.

Now there is no exaggeration about this picture. Professor Wright stated at the Oxford meeting of the British Medical Association, in 1903, that, of the ninety-seven cases which he analysed, twelve had "slight" flexion, and twenty-six "considerable" flexion-or more than one-third in all. I think I may say that a child with considerable flexion needs a crutch, and, although there may be no more active disease, for purposes of progression the patient is in no way improved.

In the second place, the procedure has a limited application. That is, it is specially reserved for children in whom the membrane only-or little more--is affected. There can be no objection whatever to adapting our operation to varying developments of disease. We do it every day. But we must

be sure as to what the condition is.

Now I need hardly say that we cannot always be sure that the disease is absolutely confined to the membrane. Opening a knee joint may reveal entirely unexpected conditions. There may be erosion of cartilages, pus pockets, and caries of the bone ends. Clearly, removal of the synovial structure will not suffice. The surgeon must develop his procedure. He must scrape out the caseous foci of disease; must deal thoroughly with the caries, and gouge out all dead or dying bone. It may be that he must use the saw, and his operation passes into the category of "modified" arthreetomy. When he does this, it is difficult to discern the line

which divides the proceeding from excision, so far as immediate shock or ultimate mobility of the joint go.

In the third place, let me notice the question of permanent shortening due to interference with the epiphyses in excision. My own experience does not support the allegation that this is one of the results of the radical operation. In the large number of children on whom I have operated, and examined subsequently at various periods, I have not noticed this defect, and I believe it seldom occurs. If a neighbouring lesion

has this effect as a necessity upon the epiphysary line, we should see it more frequently than we do. In the violence inflicted near the ends of young bones by fractures-even when there is a separation along the line of the epiphysiswe do not look for a stoppage of development. It may occur, and it is wise to take all chances into consideration in such circumstances, but I hold that we do not look for this as a matter of course. If the epiphysis were always as sensitive as undoubtedly it sometimes is, then we should expect to find this lack of development in many of the modified arthrectomies. For there the bone is entered, and the surgeon does not hesitate to use his gouge freely in clearing out carious masses, even if doing this carries him into the epiphysary line at more than one point. This invasion of a region of great physiological activity may be followed by all the dangers which are held by some to be inherent in the operation of excision. On the other hand, I wish to emphasise the fact that in the thorough operation we do not necessarily encroach upon the growing line at all, and that therefore we are not accustomed to see the disastrous shortening which we are told ought to follow.

I have now brought before you what I conceive to be the defects of the methods which I have enumerated, and I have no doubt whatever that all these ways of dealing with tuberculous joints are inferior in all respects to the radical operation of excision. In my own experience the mortality is not two per cent., and is at least as low as that of the simpler operations. The results are certainly better, for, having removed the joint and cleared away the infected soft