NEEVOUS SYSTEM OF ANTEDON ROSACEUS. 527 



slightest irritation caused very active and violent flexion of the 

 arms. 



Experiment 6. — An active uninjured specimen was held 

 under water, and the dorsal half of the centrodorsal plate 

 removed by a single snip with a large pair of scissors so as to 

 expose and partly remove the central capsule (cf. fig. 1). On 

 being released the animal fell to the bottom with the arms very 

 strongly extended, but in about twenty minutes gradually 

 righted itself and assumed the normal position. The exposed 

 central capsule was then irritated, first with a needle and then 

 with strong nitric acid applied by a small brush ; the effect of 

 irritation was to cause very strong and spasmodic flexion of the 

 arms, which in the first case ceased on removal of the stimulus, 

 but in the case of the acid persisted for several hours. 



The three preceding experiments show that irritation of the 

 central capsule, whether mechanical or chemical, causes strong 

 flexion of all the arms, which persists as long as the stimulation 

 is continued. Experiment 4 is the same as Dr. Carpenter's 

 Experiment A, though the results are not quite identical ; for 

 while Dr. Carpenter describes sudden and consentaneous 

 flexion of the arms as following irritation of the central cap- 

 sule from the oral surface, I have found that swimming move- 

 ments quite as often result. The difference is a slight one, 

 and may, I believe, be accounted for by the oral pinnules being 

 accidentally irritated in some of the experiments. If these 

 were clipped off I found that swimming movements of the 

 arms almost invariably followed irritation of the central cap- 

 sule from above. 



The experiments prove in the most positive manner that the 

 central capsule is in direct physiological connection with the 

 muscles of the arms ; and the further fact that the experiments 

 yield identical results, whether performed on eviscerated or on 

 unmutilated specimens, proves that the subepithelial bands form 

 at any rate no part of the central mechanism. 



Experiment 7. — The centrodorsal plate of an active 



VOL. XXIV. NEW SER. N N 



