310 Comparative Animal Physiology 



nerves cut eventually rights itself but shows poor coordination.^*' If all ele- 

 ments of the ring are removed, an isolated arm travels toward its base; if a bit 

 of the ring remains, the arm moves toward its distal end; if the ring is cut in 

 two places, the starfish pulls itself in two. Ophiuroids show no fixed central 

 control but a "plastic reciprocal relation between peripheral conditions and 

 central functioning".-^*^ In two monographs Smith^^^. 386 j^^s described the in- 

 terrelations between peripheral plexi, and radial and ring nerves as a series 

 of reflex arcs. In general, the central ring of echinoderms is directive, the 

 radial nerves are necessary for true locomotion and righting, but alone the 

 peripheral system can permit a certain amount of coordination between 

 pedicellariae, spines, etc. 



The nervous system of balanoglossids (Enteropneusta) consists of a sub- 

 epidermal plexus, together with dorsal and ventral cords and collar connec- 

 tives.'^'^ Experiments in which the cords were cut showed that the subepi- 

 dermal network can conduct both circularly and longitudinally but conducts 

 more readily circularly. Conduction in the proboscis is entirely by way of the 

 subepidermal network. The dorsal, ventral, and collar trunks seem to be 

 essentially conducting cords and show little integrative activity. 



In tunicates a single ganglion lies between the two siphons. Experimental 

 work has been done largely on Ciona and Ascidia, in which a peripheral 

 nerve net has not been adequately demonstrated. When the ganglion is re- 

 moved or cut transversely, intersiphonal reflexes are abolished.^ ^^ •^^*'' ^^^ 

 Strong protective reflexes are also abolished. Local responses of a single 

 siphon and of the body to mechanical stimuli remain, but reaction time is 

 increased and excitability diminished. Opinions are equally divided as to 

 whether tonus is increased or decreased by extirpation of the ganglion.^^^ A 

 contraction wave was conducted around a cut which slit a siphon. ^^^ It 

 seems certain that a direct connection exists from epidermal receptors to 

 underlying muscles in ascidians, but conduction and tonus properties of the 

 muscle need elucidation. 



Among higher chordates peripheral nerve networks persist only for coor- 

 dination of smooth muscle movement, particularly in the gut. On the other 

 side of the phylogenetic tree (worms, molluscs, arthropods), peripheral nets 

 are less important in controlling locomotion than on the echinoderm-chordate 

 side. 



Among flatworms dependence on central neurones is striking, in view of 

 the primitiveness of these worms in other respects. In free-living flatworms 

 the nerve trunks (2 to 8 in number) run posteriorly from the brain and 

 contain nerve cells which are not grouped into ganglia. As long as a por- 

 tion of one of these nerve cords is left in a piece of a planarian, spontaneous 

 movement and some coordinated responses persist.^^ A bit of remaining nerve 

 cord is necessary also in nemerteans.^^^ The only example of independent 

 function of the peripheral system is in the planarian proboscis where, if the 

 two connections to the central nerve trunks are severed, the proboscis auto- 

 amputates and shows food-seeking reactions.^^'^ 



Annelids, being tubular and containing a movable coelomic fluid, might 

 well be expected to show peripheral coordination. In the earthworm a sub- 

 epidermal plexus of branching fibers and scattered nerve cells lies outside 

 the circular muscles; each central ganglion gives rise to three pairs of seg- 



