saegent: the optic reflex apparatus of vertebrates. 237 



than when the electrodes were applied directly to the sciatic nerve. If, 

 instead of the cerebral hemispheres, the cut surface of the medulla 

 spinalis was stimulated after its severance from the medulla oblongata, 

 the delay is only 0.017 second greater than when the sciatic nerve is 

 stimulated. They therefore concluded that the delay between the cere- 

 brum and the cord is about 0.020 second. The whole time lost 

 between the surface of the brain and the gastrocnemius muscle is 0.050 

 second. Of this, 0.020 second occurs in the brain, 0.015 second in the 

 cord, and the rest in the nerve and muscle. 



Of much greater value in this discussion are the results of Wilson ('90, 

 p. 504). He found that by direct stimulation of the optic lobe of 

 the frog from which the cerebrum was removed the m. triceps femoris 

 reacted in 0.044 second. When the spinal cord was stimulated at its 

 anterior end, the reaction time was 0.020 second, showing a delay of 

 0.024 second in the passage of the impulse from the optic lobe to the 

 cord, which according to Wundt's estimate would indicate six synapses 

 in the conduction path. 



The results of Exner, Langendorff und Krawzoff, and Wilson agree 

 fairly well. They show a delay in the brain proper of respectively, 

 0.025, 0.020, and 0.024 second ; and a delay in the cord of approxi- 

 mately 0.015, 0.017, 0.020— second. 



Similar reaction-time experiments have been made on the dog by 

 Frangois-Franck ('77), Pitres ('83), Bubnotf und Heidenhain ('81), Novi 

 e Grandis ('87), and Fano ('95). These investigators find that the delay 

 of the nerve-impulse in the cerebral cortex is from 0.0i5 to 0.020 second, 

 or a little less than in the frog. Fano found that the extirpation of the 

 motor zone of the cortex causes a shortening of the reaction time, whereas 

 electric stimulation of the cortex results in a lengthening of the reaction- 

 time. From his experiments Fano concludes that the ganglion cells of 

 the cortex perform an inhibitory function. Schafer ('88) got results 

 with the monkey similar to the above. 



It will be seen that the results of Exner, of Langendorff und Krawzoff, 

 and of Wilson on the frog are in substantial agreement, and may be 

 briefly summarized as follows: The time which necessarily elapses 

 between the stimulation of the brain surface and the contraction of the 

 leg muscles is about 0.050 second. Half of this delay, 0.025 second, 

 occurs in the brain, about 0.015 second in the cord, and the remainder, 

 0.010 second, in the sciatic nerve and muscle. Bernstein ('82) finds the 

 time lost in the muscle-plate to be 0.0032 second. If we allow 10 cm. 

 for the length of the sciatic nerve in the frog, the delay in trans- 



