The brachialis anticus and triceps should be considered as pure flexor and extensor of the elbow respectively, though the supinator longus acts powerfully as a synergist flexor when the elbow is slightly bent and the forearm is in mid-pronation. The earliest training should be performed with the patient recumbent and the forearm vertical (see Fig. 41, p. 88). In this position the muscles can act from the vertical with the aid of gravity and slightly against gravity through the few degrees on each side of the vertical. When an angle of more than 25° or 30° is passed, the strength required is very considerable to elevate up to the vertical once more. The "zero" position is secured by supporting the arm and forearm on a board whose top is at the level of the armpit. The forearm is then moved to and fro across its surface. A the distal edge of the board is dropped resistance is offered to flexion, assistance to extension (see Figs. 43 to 45, pp. 89 and 90).

In practice, elevation or depression need not be carried further than about 6o°, though theoretically the process should be continued till the vertical is reached. The final stages can be dealt with by raising the patient from the half-lying to the sitting position. Using the weight and pulley, the middle pulley should be employed, resistance to the brachial anticus being afforded when facing the apparatus and to the triceps when standing with the back to it, the cord passing over the shoulder (see Figs. 98 and 99, pp. 232 and 233).

The position of rest for the triceps is full extension of the forearm while it hangs loosely to the side of the body. To secure rest for the brachialis anticus flexion to something short of a right angle is essential. This can be given by supporting the forearm across the chest, e.g., by passing the hand through the opening in the waistcoat above the top button but one. A loop of bandage serves the same purpose.

The biceps should, for the purpose of early training, be considered as a pure supinator of the forearm, though it undoubtedly serves as a synergist flexor of the elbow. It should be trained at first while the patient is recumbent, the elbow flexed to a right angle with the forearm vertical. By raising the trunk slight extra resistance can be given, and then rotation should be practised with varying degrees of flexion and extension of the elbow. It should be noted that, in full extension, the power of the biceps as a supinator is very poor (if indeed it acts at all), the action being performed by the supinator brevis. The pronators of the forearm - radii teres and quadratus - are the antagonists of biceps and supinator brevis, and are trained to contract in similar manner. To train either supinators or pronators by means of apparatus is not very easy, as graduation of resistance has to be very carefully regulated. The first stage is to wind up a few turns of the rope on the rotator and then release both catches. The weight is made to assist the muscles we wish to train, and to resist the antagonists. Reduction of weight reduces the assistance till finally, when no weights are attached, movement is practically "free." The addition of weights to act in the reverse direction then affords resistance. The position of rest for the rotators of the forearm is secured by resting the forearm across the chest as suggested for resting the brachialis anticus. The mid-position between pronation and supination rests both supinators and pronators alike. It is worthy of note that MacKenzie regards the supinator longus as an accessory supinator only and not as a synergist flexor of the elbow.

Flexion and extension of the wrist should be regarded as the functions of the long carpal flexors and extensors. When the hand is in full supination gravity will extend the wrist quite easily, if the flexors are relaxed, and vice versa. As pronation takes place gravity plays less part, until in the mid-position its action is zero and movement becomes truly "free." Further pronation means extra resistance from gravity to dorsi-flexion. For some time I was under the delusion that the use of the roller apparatus (the patient turning the instrument towards himself) was an exercise for the dorsi-flexors. A patient with a complete posterior interosseous paralysis can, however, perform it with ease. The only muscles that are really involved in the movement are the flexors of the fingers and the triceps. So unexpected are some of the actions revealed by muscle-training! The only incentive to the dorsi-flexors to act in performing this exercise is that of their synergistic affiliation towards the flexors of the fingers. This supplies a key to another method of early training of the carpal extensors, namely, to train the grip of the finger flexors. It can readily be done in any use of the weight and pulley provided the patient stands with his back to the apparatus, and also by ladder exercises provided the grip is taken in supination or on the side bars of the apparatus. The carpal flexors rarely need any definite precaution as regards rest, unless indeed only one is weak. Then a lateral splint should be applied to prevent over-action of its fellow. The dorsi-flexors are best protected by use of the short cock-up splint (see Fig. 161, p. 463). The lateralising action of the carpal extensors and flexors on the wrist can be trained by placing the forearm and the palm of the hand flat on a smooth board.

It is well to consider the training of the flexors and extensors of the fingers together. The muscle action involved is intricate and is best studied by consideration of the ordinary deformity seen so often (quite unnecessarily of course; its existence indicates neglect in treatment) after paralysis of the ulnar nerve. The metacarpo-phalangeal joints of annularis and minimum digitus are hyper-extended, and the inter-phalangeal joints are fixed in flexion. The extensors of the metacarpophalangeal joints are therefore over-acting, their flexors are paralysed: the extensors of the inter-phalangeal joints are paralysed, their flexors are over-acting. The only possible conclusion is this. The extensors of the metacarpo-phalangeal joints are the extensors communis, indicis and minimi digiti; the flexors are the lumbricals. The extensors of the inter-phalangeal joints are not the long extensors, but only the interossei, the flexors of the proximal joints being the flexor sublimis, and of the distal the flexor profundus. All training of the flexors of the fingers should be performed with the wrist dorsi-flexed. The carpal extensors act powerfully as synergists of lumbricals and long flexors equally. The lumbricals are trained while the inter-phalangeal joints are kept in full extension, the long flexors while the metacarpophalangeal joints are kept fully extended - the golf club grip.