40 PSYCHOLOGICAL EFFECTS OF ALCOHOL. 



RECORDING DEVICE. 



On grounds which are already indicated, we believe that adequate 

 records of the human knee-jerk must be either direct records of quad- 

 riceps thickening or galvanometric records of the muscle-current of 

 action. While the latter are probably preferable to the former, and 

 should be used for the final analysis of the phenomenon, they are much 

 more expensive of time and material, and are practically more difficult 

 to manage. In the present investigation the records were produced 

 by direct recording levers which wrote on a Blix-Sandstrom kymograph 

 the reflex thickening of isometric quadriceps muscle. 



The adaptation of our recording levers to the different subjects 

 proved an unexpected source of difficulty. It was proved in the case 

 of Dodge that neither the size of the lever terminal which rested on the 

 muscle nor the pressure which it exerted on the muscle at the point of 

 contact, had any considerable influence on the latency of the reflex. 

 In various subjects, however, a new and somewhat serious source of 

 error was discovered. The blow of the hammer upon the tendon 

 always sets up within the muscle a mechanical wave-like disturbance. 

 We depend on this wave to record the moment of stimulation. Under 

 certain combinations of stimulus intensity and tonic contraction of the 

 muscle, which are otherwise undefinable, this mechanical disturbance 

 consisted of a succession of damped oscillations, which occasionally 

 seriously complicated the curve and rendered the true beginning of 

 reaction uncertain. Two devices seemed to lessen these vibrations: 



(1) The area of contact between the lever system and the muscle should 

 be relatively large. In all except the earliest experiments we used a 

 rectangular base, 13 mm. by 70 mm., placed lengthwise of the muscle. 



(2) The elastic pressure of the lever system against the muscle should 

 be relatively intense as well as quick acting. An elastic band was used 

 for this purpose which exerted a pressure of about 500 gm. Though 

 this varied somewhat from individual to individual because of the 

 variations in diameter of the respective thighs, it remained practically 

 constant for each individual throughout the series. Our lever system 

 magnified the muscle thickening by the proportion of 6 to 1. This 

 proportion was found by preliminary experiment to be the most favor- 

 able with our particular lever arrangements. 



For recording the knee-jerk we used the Blix-Sandstrom 1 kymo- 

 graph, which was run at a peripheral rate of 100 mm. per second. 

 While this form of kymograph is one of the most accurate and con- 

 venient available, it may not be used without constant watchfulness 

 and occasional readjustments of the regulator. Even the most careful 

 regulation at the beginning of an experimental session proved to be 

 inadequate. Except in the earliest experiments, we consequently used 

 a]control time-record throughout. Unfortunately for psychological 



ix, Archiv f. d. ges. Physiol., 1902, 90, p. 405. 



