SIMPLEST NEURAL ARCS. 



was mounted at one end on a vertical axis, so that it could move freely 

 in a horizontal plane at about the height of the subject's patella r 

 tendon. The free end of this rod was attached by a flexible cord to a 

 point concentric with the axis of the pendulums, and was adjusted to 

 such a height that both pendulums struck it at their center of gravity 

 when they reached a vertical position. The height of the whole 

 system could be changed for each subject, without changing any of 

 the instrumental constants, by raising or lowering the sliding base BB, 

 so that the rod B' rested against the middle of that particular subject's 

 patellar tendon. Once adjusted, the base was securely clamped against 

 the vertical frame, and the rod B' was given an even tension against 

 the tendon by the pressure of an elastic band which was stretched 

 between the rod and a fixed point on the upright support. When once 

 fixed for any subject, this system remained unchanged throughout an 

 afternoon's experiments. It could be afterwards reset for the same 

 subject by the use of a scale which was attached to the side of the 

 frame. But for obvious reasons the scale alone was never depended 

 upon. On each day the position of the system was carefully verified 

 for each subject. 



The blows of the pendulums were thus transmitted to the subject 

 through a light horizontal lever which was adjusted as above indicated. 

 This secured identity of the point of application of the two blows. 

 Since a lever neither increases nor decreases energy, the effect on the 

 tendon must be practically identical for both pendulum hammers, 

 even though one pendulum strikes the horizontal transmitting-rod 

 somewhat nearer its axis than the other. This simple theoretical 

 relationship is somewhat complicated by the fact that in practice the 

 lever will have a certain amount of weight and elasticity. To reduce 

 the error of transmission to a minimum, our transmitting rod is as long 

 as it can conveniently be (60 cm.), while the two percussion hammer 

 pendulums strike it as near together as practicable (25 mm. apart). 

 The consequent discrepancy in the energy of the successive blows is a 

 small fraction of 1 per cent, and is negligible in practice. In compara- 

 tive measurements this discrepancy can play no role at all, since it is 

 an instrumental constant. 



Two checks on the constancy of the stimulus blows are included in 

 our records. (1) If the blows of the pendulums are exactly equal, 

 the extent of the mechanical disturbance to the muscle incident to 

 stimulation should be equal after each blow. To be sure, this can not 

 be a very fine measure of the relative energy of the pendulums, but it 

 served to disclose any accidental differences in the weights. (2) We 

 took what is probably excessive precaution in arbitrarily omitting the 

 first blow at least once in every series of experiments, to see if the blow 

 from the second pendulum produced an appropriate reaction. There 

 were no measurable discrepancies. 



