South African Clawed Frog. 



37 



events, be aroused by sensory stimuli playing upon the same limb. This may 

 have led to slightly greater facility and shorter delay in transmission at the 

 same-side synapse. 



The actual time of the reflex varies, of course, with temperature, size, and 

 condition of animal. 



The temperature in the room in the Physiology Department where the 

 frogs are experimented upon rises in Summer as high as 26° C. If we select 

 a reflex time recorded at a room temperature of 23° C. from a frog in good con- 

 dition, as for example frog B, Experiment 5, flrst response, thirty-five days after 

 operation, we find an interval of ll'-i cr for the total reflex time. The length 

 of nerve from toes to cord and from cord to thigh was found j^ost mortem in 

 this frog to be 16 cm. The rate of conduction in the nerve of Xenopus at 

 23° C. was determined in the usual manner by electrical stimulation at two 

 points, with the electrical change in the muscle as indicator, and was found 

 to be 37 metres per second. The latency of the muscular response to nerve 

 stimulation close to the muscle was found to be 2'4 a. If we allow 1 a for 

 the latency of sensory nerve endings we may suggest tlie following analysis: — 



Total reflex time ll"4o- 



Nerve conduction 4"3 



Muscle latency 2 "4 



Latency of sensory endings I'O 7'7 



SynajJse time 3"7cr 



The conclusion to which one is led by these results is that, at the higher 

 range of temperatures, and when tlie frog is in good condition, the homo- 

 nymous and heteronymous reflexes each involve a path through the spinal 

 cord which contains one synapse or set of synapses having a delay of 3'7 a. 

 The reflex in the spinal frog is of the nature of a jerk, and we may compare 

 this synapse delay of 3'7 a in Xenopus with the delay of about 2 a found for 

 the knee jerk in the mammal (3), (4). 



There is however evidence that at lower temperatures, or when the frog is 

 not in good condition, the short path to the crossed motor neurones, which we 

 have described as containing one synapse, is not open. I have grouped 

 together the remainder of the experiments performed on the spinal frog as 

 examples of this. The figures are as in Table on p. 38. 



In this second group of experiments there is a considerable time elapsing 

 between the beginning of activity in the two limbs. The extra delay of the 

 crossed reflex amounts to about 8 a, wliich suggests that a heteronymous 

 path is being traversed here which contains more synapses than the 

 homonymous path. We have seen reason to conclude that the delay at 

 one synapse is almost 4 a, and we have to consider whether this difference of 



