412 



SCIENCE 



[N. S. Vol. XLl. No. 1055 



suppose that the primitive nervous mechan- 

 ism is the automatic one seen at work in 

 the movements of narcosis progression. 

 These vrould serve in the simplest animals 

 the purpose of progression which may rea- 

 sonably be regarded as among the first 

 functions of coordination a nervous system 

 would be likely to serve. In other words, 

 the nervous mechanism of locomotion, like 

 the nervous mechanism of respiration, is 

 fundamentally an automatic mechanism. 

 Later on afferent neurones are added to it, 

 comparable to those of the pulmonary 

 branches of the vagus. In this connection 

 it is most significant that in general the 

 same conditions so frequently referred to as 

 stimuli of the respiratory center — lack of 

 oxygen, excess of carbon dioxid, etc. — are 

 the very conditions found to favor the 

 movements of narcosis progression. 



If, then, to the respiratory center, which 

 has thus far stood in lonely glory as the 

 one fully established example of automatic 

 nervous action, we must add the funda- 

 mental centers of locomotion, the thought 

 at once suggests itself that renewed investi- 

 gation may find the same thing true of 

 other actions which in the past we have too 

 complacently catalogued under the head of 

 reflexes. The field thus opened up is a 

 large one. 



AXON REFLEXES 



Text-books of physiology usually record 

 two observations, one by Langley, the other 

 by Bayliss, which were not suspected of 

 bearing any relation to each other and both 

 of which have been difficult to fit into the 

 orthodox scheme of nervous action. So far is 

 this true that Langley 's axon reflex has 

 been relegated to the inglorious place of a 

 laboratory curiosity which plays no role 

 in normal life, while Bayliss's proposed 

 theory of antidromic impulses has been 

 treated with a polite but uncompromising 

 skepticism. 



The axon reflex is a reaction made pos- 

 sible by the branching which generally 

 takes place at the end of an axon. Inas- 

 much as nerve fibers can conduct impulses 

 in both directions, it follows that stimula- 

 tion of one of the terminal branches will 

 start an impulse traveling up to the point 

 of union of the two branches, and then down 

 the other branch to the end organ. The 

 axons to a frog's sartorius, for example, 

 branch soon after entering the muscle and 

 it often happens that one branch will go to 

 one side of the muscle, while another branch 

 of the same neurone will pass to the oppo- 

 site side. If, now, the lower third of the 

 muscle be divided longitudinally, it is 

 found that a stimulus applied to one half 

 so as to excite its nerve fibers will cause 

 contraction of the opposite half of the 

 muscle. The same thing is rendered pos- 

 sible whenever a preganglionic efferent 

 neurone passes through several sympathetic 

 ganglia, giving off collaterals to postgan- 

 glionic neurones in successive ganglia; in 

 this case stimulation of the terminal branch 

 of the preganglionic neurone will start an 

 impulse centripetally and excite, through 

 the collaterals, the cells with which these 

 collaterals are connected. It is also well 

 known that one must be on his guard 

 against axon reflexes in testing the regen- 

 eration of nerve flbers, for it often happens 

 that in the process of regeneration an axon 

 of the central stump may branch before 

 entering the peripheral stump; if these 

 two axon branches find their way into dif- 

 ferent branches of the peripheral nerve 

 trunk, stimulation of one of these branches 

 may give an apparent reflex, which, how- 

 ever, is only an axon reflex. 



These and other examples that may be 

 cited are, however, only laboratory curiosi- 

 ties. Where the two branches of the axon 

 end in a muscle or a gland neither branch 

 can be stimulated at its ending except by 



