SARGENT: THE OPTIC REFLEX APPARATUS OF VERTEBRATES. 239 



retardation of 0.03 to 0.04 second in the respiratory-centre inhibition, 

 whicli is produced by excitation of the vagus trunk." 



The evidence obtained by reaction-times as to the number of synapses 

 traversed by a motor yeflex to the peripheral nerves, though reliable only 

 witliin wide limits, shows that the simpler motor reflexes involve trans- 

 mission through a chain of from three to six neurons and consume a 

 considerable time, — from 0.03 to 0.05 second. 



Our morphological knowledge of the chains of neurons involved in 

 such processes as the last is not sufficiently exact to enable lis to tell 

 just how many neurons are included in this chain, but there is reason to 

 believe that there are at least three, for even the simpler reflexes. 



4. Time of OjJtic Reflex Reaction. 



We have seen that the average time of delay at a synapse in the, 

 central nervous system may be conservatively estimated at 0.006 second. 

 The delay in trausmission through a cell body cannot be considered as 

 known, nor do we know in how many of the neurons the impulse passes 

 through the cell body. We may accept it as determined that the 

 impulse is in some cases transmitted through the cell, and that in such 

 cases there is a delay. Taking tliree neurons as a conservative estimate 

 of the number in the chain by which ordinary i-eflexes in the frog (the 

 animal in which these values have been best determined) are transmitted 

 from the ending of the optic fibres to the musculature, we may estimate 

 the time involved in transmission, and compare it with the proliable 

 time over the short circuit path of Reissner's fibre. The time of a 

 reflex through the cord may then be estimated as follows : three 

 central synapses (0.006 X 3 = 0.018), plus muscle-plate arborization 

 (0.003), plus transmission over 0.2 m. nerve (0.007), plus delay in two 

 nerve-cell bodies (2 x) gives for the total time 0.028 + 2 x second. 

 Subtracting from this the time of peripheral delay, we have approximately 

 0.025 + x second, which corresponds f\xvorably with Wilson's result of 

 0.024 second for central delay in the frog. 



5. Time saved by the OjMc Reflex- Apimratus. 



In the transmission of a reflex over the Reissner's-fibre path a single 

 neuron suffices to bridge the gap (Fig. M), so that but two synapses and 

 one cell body must be traversed. jNIoreovcr, the rate of transmission 

 along the conducting substance may, as we have seen, be safely taken as 

 twice as great as ordinary transmission, giving the following estimate of 



VOL. XLV. — NO. 3 8 



