46 Mr. W. A. Jolly. Reflex Times in the 



■mammal as given by Sherrington (6). This author regards the reflex path 

 when the skin of the shoulder is stimulated as consisting of three neurones, 

 and as entering the grey matter twice, that is to say, it is a disynaptic arc. 

 But Sherrington states that this does not mean that there are necessarily only 

 two synapses, but that there may be other synapses due to the interposition 

 of one or more short intraspinal neurones. Tlie diagram given by this author 

 would be consistent with the results here obtained for the flexion reflex in the 

 frog on stimulation of the hand, if we assume that a short intraspinal neurone 

 occurs between the long descending propriospinal neurone and the motor 

 neurone, and that this intraspinal neurone connects not only with the motor 

 neurone on one side of the cord but, by a branch, with that of the other side 

 also. Tlie chain would thus consist of four neurones with three synapses. 



A number of delayed responses were obtained from the hind-limb on hand 

 stimulation, similar in time to those recorded from stimulation of the foot. 

 Thus homonymous hand stimulation gave a delay of 42-5 a, and heteronymous 

 stimulation delays of 57"5 a, 70 a, and 72 a. 



In one experiment where the stimulation was given to the homonymous 

 hand a reflex from the leg was recorded 16"9 tr after stimulation and another 

 deflection followed 58 a after stimnlation. Experiments such as these suggest 

 that the delayed reflexes involve higher centres than those of the spinal cord, 

 and that the path through these centres in the lower part of the encephalon 

 involves a large number of synapses. In a double response the segmental 

 spinal centre has discharged with its brief delay, while the afferent impulse 

 has also ascended to the brain and has led to a second reflex with a long 

 delay. If we assume the same delay at a synapse in the encephalon as we 

 have seen reason to believe affects the spinal synapse, then in the experiment 

 last mentioned the long path through the brain would contain ten or eleven 

 more synapses than the spinal path. 



We have assumed that the heteronymous hind-limb path, and also the 

 fore-limb to hind-limb path, contain three synapses, and we have been led to 

 this conclusion by using for comparison the reflex in the spinal frog with its 

 single synapse and the group of decerebrate reponses marked Group A on 

 diagram (fig. 5), which we assume to be reflex in Jiature and to involve one 

 synapse. The possibility, liowever, is not excluded that homonymous reflexes 

 in the decerebrate preparation may, in response to some forms of stimulation, 

 employ a path involving two synapses. It must be borne in mind that the 

 stimulus used to evoke reflexes in the decerebrate frog has been the powerful 

 squeeze. This is desirable in the case of the crossed reflex, because it 

 requires a strong mechanical stimulus to elicit the crossed reflex with fair 

 regularity, but the homonymous reflex is more easily obtained. The light 



