1. For amputations of the thigh, it is important to distinguish between those stumps which are weight-bearing and those which are not. In the latter case, the success or failure of the artificial limb depends upon an accurate fit at the ischial tuberosity. Most brace-makers fail to realize that the tuberosity does not slant from above downward and forward but in the reverse direction, namely, from below upward and forward. This upward inclination, be it ever so slight, must be taken into account. The usual type of support given by the brace-maker, does not conform to this anatomical fact, but slants from above downward and forward, so that the patient slips downward on the support and almost invariably suffers pain anteriorly, near the pubic bone. The result is that the stump is rotated, and the artificial limb does not fit. In addition to the tuberosity of the ischium, the adductor muscles are capable of bearing great weight when they have been properly hardened. The pubic bone, however, cannot stand pressure and must be left free. The gluteal muscles and the vasti also help to support the body-weight.

When the stump is short, a pelvic girdle with a strong joint at the level of the trochanter is necessary; whereas in the long stumps, the pelvic band and trochanteric joint are unnecessary. In patients with marked atrophy of the muscles, unable to balance themselves securely upon their stump, the trochanter joint should allow flexion and extension only, since the pelvis would drop toward the opposite side of the body, were abduction permitted.

In applying the steel uprights which support the body, or, in case of a wooden limb, in joining the thigh-piece with the calf, it is advisable to give the calf about 2° of genu valgum position. This adds markedly to the stability of the artificial limb.

The type of knee-joint does not, so far as I can observe, play an important rôle. In general, the simpler the mechanism the more effective. Complicated screws, ratchets, or springs add merely to the likelihood of breakage and to the cost of keeping the limb in order. Besides, for the majority of patients, who live at a distance from the industrial centres where brace-makers are to be found, the entire construction

of the limb should be so simple as to permit the wearer himself to make the necessary repairs. In one European hospital there is an admirable custom of giving each amputated patient a 3 weeks' course in the brace-maker's shop, and discharge from the hospital is dependent upon ability to repair his own prosthesis.

An essential in the mechanical construction of the joint is the location of its axis posterior to the centre of gravity of the

FIG. 137.-Diagrams illustrating the importance of posterior displacement of the knee joint of the artificial limb. A, Body; B, hip; C, knee joint; D, ankle. In Fig. I, the axis of the artificial joint corresponds in position to the anatomical. A slight degree of flexion brings the body weight posterior to the axis and, as is evident from the figure, further flexion must result. For the patient this position of the axis causes insecurity since the least degree of flexion is almost certain to cause him to fall. In Fig. II, the axis of the artificial limb has been displaced posteriorly. The body weight, represented by the dotted line, now falls anterior to C (the axis) and tends to lock the knee instead of producing further flexion.

anatomical joint. If this demand is not complied with, the patient loses all sense of security, because the artificial leg tends to bend at the knee under the patient's weight. If the mechanical joint lies posterior to the normal, then the bodyweight tends to lock the joint as is seen by reference to the diagram (Fig. 137).

An artificial quadriceps does not, I find, add to the naturalness of the stride, but almost invariably tends to hold the leg fully extended, so that the patient walks as though the knee were ankylosed. A freely swinging joint with some simple rubber or spring device to prevent jarring in extension or flexion gives the patient the best opportunity to imitate the normal gait.

2. For amputations of the calf, the type of limb depends upon the length of the stump. If it is short-less than onehalf the length of the calf-there must invariably be a thighpiece and a knee-joint. If it is long, these may be dispensed with provided the stump is capable of weight-bearing.

As already indicated, when the stump is not capable of weight-bearing, the artificial limb must be so moulded as to grasp the tuberosity of the tibia firmly, not the condyles, as is usually taught. The patella-tendon also is capable of weight-bearing, as can be learned by observing any patient who has worn an artificial limb for many years.

Some difficulty is frequently experienced in bringing the leather socket of the artificial limb over the gastrocnemeii. This can be obviated by slitting the socket posteriorly and inserting eyes so as to lace it up when once it is in proper position.

The ankle-joint, like the knee, should be of the simplest type, allowing merely flexion and extension. In addition to the ankle-joint, there should be one corresponding to the metatarsophalangeal junction.

Types of Artificial Limb for Amputations of the Upper Extremity. The problem of dealing with amputations of the upper extremity is far more difficult than is the case with amputations of the lower limbs. The legs merely have to carry the body, but the arm has a great variety of functions to perform. Depending upon the nature of these functions, and also to a great extent upon the site of the amputation, the artificial limb must vary from one case to another. Thus, an artificial limb which might be of value to a lawyer or business man would be of little use to the farmer or mechanic; and of two farmers, one with an amputation of the forearm, another with an amputation above the elbow, the one would

have to be equipped with a type of limb differing markedly from that supplied to the other. There is no universal artificial limb applicable to all cases.

1. Types of Artificial Arms Designed for Amputations of the Fore-arm. For the farmer and artisan, a simple and effective prosthesis has been designed by August Keller. Amputated

FIG. 138. The Keller artificial hand. The picture illustrates Keller's method of inserting a small knife, with which he is sharpening his pencil. Note also the piece of cork attached to the pencil. This enables him to grip the pencil between the claws and to write with it. The lower arm socket is held firmly in place by a broad strap which makes a figure-of-eight turn about the elbow.

himself, some nine years ago, he constructed an artificial limb of the simplest materials, so well adapted to the needs of the farmer that the amputated scarcely note the handicap under which they are compelled to work. Keller's device consists of a leather socket reinforced by two longitudinal steel bars, held in place by a figure-of-eight strap which passes just above

the elbow (Fig. 138). The hand-piece, made of wood, can be removed from the socket if desired (Fig. 141). Inserted into the wooden hand-piece are three strong steel hooks. These are not adjustable. They aid the patient in two ways: first, small objects, such as pencil or knife, can be inserted between them, second, they furnish the leverage for larger instruments. To hold these latter in place, a leather strap, attached to the

anterior portion of the apparatus, is made to take a double turn about the handle of the article used (see Fig. 143) and then passing backward between the hooks, is fixed to the posterior aspect by means of a steel pin. The illustrations indicate how Keller uses his own device. The speed, accuracy and power which he exhibits are scarcely inferior to that of the normal individual.

A large number of other contrivances have been evolved to replace the fingers. These consist of hooks, rings, clamps, and