EXPERIMENTS ON VERTEBRATES 137 



cord, but are only in contact with them. 1 For the 

 transmission of the impulse this fact is of no import- 

 ance ; it is not necessary in either case that the gan- 

 glion-cells of the posterior horn and the sensory 

 nerves be grown together, they need only to be in 

 sufficiently close contact. Engelmann called atten- 

 tion to these relations long ago in his excellent article 

 on conduction in the ureter. 



2. Sufficient data exist for proving the segmental 

 localisation of reflexes in the spinal cord. In a dog 

 whose spinal cord has been severed somewhere in the 

 thoracic region the posterior part is entirely separated 

 from the anterior part as far as the motor and sensory 

 functions are concerned. Immediately after the oper- 

 ation severe shock-effects appear, but these are only 

 temporary, and we shall return to this subject later. 

 The interruption of the continuity is permanent, for 

 in the central nervous system of higher animals no re- 

 generation has been observed, but only a healing 

 together of the cut surfaces by means of connective 

 tissue. In such an animal the part located behind the 

 point of division shows all the reactions which are 

 possible in the corresponding segments. Goltz has 

 proved this for dogs. Rubbing of the skin produces 

 scratching movements of the hind-legs ; erection of 

 the penis and urination can be produced by stimulat- 

 ing the foreskin. The reflexes of the rectum and 



o 



bladder and the vasomotor reactions are intact. We 



1 Apathy's publications arouse suspicion as regards the results obtained by 

 Golgi's methods. 



