152 F. H. SCOTT. 



were cut and laid on the electrodes, but even with this arrangement the 

 root usually responded with a slightly weaker current than the nerve, 

 but if the frog has been well cooled the difference is not marked. 

 Tracings were made by extending the hind legs with elastic bands and 

 then attaching threads to each cut tendo achilles. The threads were 

 passed over pulleys and attached to levers. Signal magnets under the 

 levers marked the point of stimulation. A second battery was used to 

 operate the magnets. In most cases the two sides responded equally 

 well and gave the same number of contractions before failing to respond 

 after 30 seconds rest (Fig. 1). In a few cases the ganglionated or the 

 severed failed a little before the other. If there was very marked 

 inequality the experiment was abandoned, as probably some injury had 

 been done in the preparation. The one side was usually stimulated for 

 30 seconds and then the other. When one side failed a longer rest was 

 allowed. In all cases, finally, the ganglionated side was the better. 

 The tracings show some of the records obtained in the manner above 

 described. The levers were made as equally sensitive as possible, and 

 were tested by interchanging the levers of the two muscles. 



Fig. 1. Upper line from stimulation of cut root, lower line from corresponding nerve in 

 plexus. The fall of the signal indicates stimulation. The time of stimulation equals 

 30 seconds, when the commutators were reversed into the other electrodes and magnet. 



Fig. 1 is a tracing in which the responses are much more regular 

 than usual. It represents the second set of stimulations, one hour after 



