EXPERIMENTS ON WORMS 87 



resemble locomotion. These motions are presumably 

 produced by the passive stretching of the skin on the 

 ventral side of the hip-joint by the weight of the legs. 

 These motions are comparable to the reflex contrac- 

 tion of the longitudinal muscles of the earthworm, 

 which is due to the stretching of the skin. Because 

 of this reflex, coordinated locomotion would be quite 

 possible in a dog with divided spinal cord, if the dog 

 only could remain standing on its hind-legs. The 

 walking movement of the fore-legs would cause the 

 stretching which is necessary in order to bring about 

 the walking movement of the hind-legs. The differ- 

 ence in the behaviour of a dog with divided spinal 

 cord and of an earthworm with divided ventral nerve 

 cord in regard to coordinated progressive movements, 

 is not caused so much by differences in the functions 

 of the central organs as by differences in the develop- 

 ment of the peripheral organs of the skin and of the 

 organs of locomotion. If the dog had short stumps 

 instead of its long, jointed legs, we should have, after 

 dividing the spinal cord, the same phenomenon of 

 progressive movements that we have in the earth- 

 worm. The irritability of various parts of the peri- 

 pheral organs and the simple segmental arrangement 

 of the nervous elements suffice to preserve the loco- 

 motion when it has once been started. The correct- 

 ness of this conclusion is confirmed by experiments 

 on Nereis, which were made in my laboratory by S. 

 S. Maxwell (5). In these animals, the coordination 

 of the movements of the oral and aboral pieces is 



