South African Clatved Frog. 



35 



as eight thousandths of a second (8 a), that is to say, the crossed is longer 

 than the same-side time by this interval, but sometimes there is a reversal 

 and the crossed time is shorter than the same-side time. In other cases no 

 appreciable difference of time is found between the two reflexes. 



It was hoped that the difference of time would be suitable for an investiga- 

 tion of the influence of temperature upon the delay at a synapse, on the 

 assumption that an additional synapse is interposed in the path of the 

 heteronymous reflex, but it is difficult to explain on this assumption the 

 complete disappearance of the interval in some experiments or the occasional 

 reversal. On the other hand a difference of 8 cr is quite consistent with the 

 view that the heteronymous path involves more synapses than the homo- 

 nymous and indeed is most easily explained in this way. One is inclined to 

 contemplate tlie possibility that as the conditions change, the path taken 

 through the spinal cord in the heteronymous reflex may alter. 



The following results were obtained by the use of the innocuous stimulus. 

 The times are given in thousandths of a second (tr) and the temperature in 

 degrees Centigrade (see Table on p. 36). 



The time difference between the reflexes on the two sides, when present, is 

 small. In the above experiments, which I have grouped together, the longest 

 time difference, 4*1 cr occurred at the lowest temperature used, 14°, but the 

 differences generally do not seem to have a simple relation to the tempera- 

 ture ; at any rate, from the data at present available such is not evident. 

 It may be mentioned that frog B, which gave the difference of 4"1 a, died 

 three days later (forty-five days after operation) so that presumably at the 

 time of experiment it was not in the best condition, and this may have 

 had an influence upon the time difference. 



I do not think that the above experiments taken by themselves would justify 

 us in assuming that in the spinal frog at the temperatures employed, and with 

 the frogs in good condition, the intraspinal path of the heteronymous reflex 

 involves a larger number of synapses than that of the homonymous reflex. 

 The fact that the former may sometimes have a shorter time than the latter, 

 which seems also to have been observed by Eosenthal (1), tells against such a 

 view. It would rather appear that we are dealing with mechanisms involving 

 the same number of neurones on the two sides. Even if they are similarly 

 constituted as regards number of synapses one could hardly expect that the 

 mechanisms would work absolutely synclironously. The fact that the 

 homonymous time is usually slightly the shorter may possibly be due to the 

 motor neurones responding rather more readily to stimuli reaching them 

 through afferent nerves from the same side of the body. Doubtless the 

 greater number of the reflex flexions of a leg will, in the ordinary course of 



D 2 



