Nervous Systems 823 



descending tracts; also the spinothalamic tracts are differentiated and the long 

 ascending tracts of the posterior columns appear. Associated with these mor- 

 phological changes there is a striking change of function of the spinal cord. 

 In the transition from water to land, locomotion changed from swimming, a 

 mass movement, to the much more complicated quadripedal locomotion. 

 Lower vertebrates are characterized by behavior involving mass movements 

 of a total-pattern type as distinguished from localized reflexes of higher ver- 

 tebrates.^^" 



There has been much disagreement over the autonomy of the spinal cord 

 and the extent to which locomotor behavior is fixed by sensory stimulation 

 rather than following a set pattern of neurones in the cord. What the cord 

 can do seems to vary from animal class to class and from one related genus 

 to another. In a general way the cord carries out more independent action 

 in the lower vertebrates than in higher forms. Also the shock resulting from 

 spinal transection is less in lower vertebrates. In higher groups (mammals 

 particularly), the more primitive part of the central nervous system, the 

 spinal cord, is less sensitive to asphyxiation but more sensitive to strychnine 

 than is the cerebrum.^'^ Evidence regarding the potentialities of the cord is 

 obtained by isolating all or part of the spinal cord from the rest of the cen- 

 tral nervous system and removing sensory influence by cutting dorsal roots. 



A swimming or crawling animal with many segments often locomotes by 

 undulatory waves which show remarkably precise timing. In an eel, for ex- 

 ample, the phase of movement of each segment is just behind the phase of 

 the segment ahead, and the muscles on the two sides differ in phase by one- 

 half cycle.^"^^ Is the nervous basis for such a coordinated locomotor wave in- 

 herent in the spinal cord or is it dependent on chain reflexes? It has been 

 suggested that the central pattern for walking evolved from this basic loco- 

 motor plan. 



In Amphioxus any local stimulus elicits a general avoiding response. Le- 

 sions to the cord show that propagation of strong undulations over the whole 

 body and end-to-end startle responses require the median giant neurones but 

 that the rest of the isolated cord can reflexly evoke secondary superficial 

 waves.^**® 



Spinal hagfish CPolystotrema') are inactive for long periods but when 

 stimulated they swim normally. Normal undulating waves start at the an- 

 terior , end ; and | the point | of ' transection i of the cord hecomes a pacemaker for 

 initiating swimming waves. When the stimulation is vigorous, a wave may 

 pass a cut by reflexly exciting the region beyond.®.^ Similarly an elasmo- 

 branch made spinal by transection behind the head swims much like a nor- 

 mal fish. Ten Cate^°^ maintained that a wave of contraction could pass a 

 cut in the cord by virtue of overlap of proprioceptive fields in several seg- 

 ments and stretch of one segment by the preceding one. Thus the propa- 

 gation would be a chain reflex. Gray and his associates,^'^"- ^^^ on the other 

 hand, found that transmission could proceed in an intact cord even through 

 12 denervated segments, and that transmission failed when the cord was cut. 

 So long as the spinal dogfish is free from contact, a locomotor rhythm at 

 about 40 waves per minute occurs, but a gentle tactile stim'jlation, particu- 

 larly of the ventral surface, increases or inhibits this rhythm (Fig. 307). 



The rhythmic movement of body and fins of a dogfish is abolished by 



