528 A. MILNES MARSHALL. 



specimen was removed with scissors and the central capsule 

 carefully scooped out with a small scalpel. The animal on 

 being released fell to the bottom of the water, where it lay on 

 its side with the arms very strongly extended ; it remained in 

 this position for several hours without any attempt to move. 

 If taken from the water and thrown in again the arms moved 

 fairly actively, but there was no attempt at swimming, each 

 arm apparently acting quite independently of the rest. Finally, 

 if placed on its oral surface it remained there for an indefinite 

 time without making the slightest attempt to right itself. 



Experiment 8. — The preceding experiment was repeated 

 on an eviscerated specimen, the results being in all respects 

 the same. 



These two experiments are of very great importance. They 

 show that removal of the central capsule completely destroys 

 the co-ordinating mechanism between the arms as tested (a) 

 by the power of executing the normal swimming movements, 

 (b) by the power of righting itself when inverted ; both these 

 powers being permanently destroyed by the operation. To 

 obtain definite results it is necessary to completely remove the 

 central capsule, and this I have found cannot be eff'ected by 

 simply cutting away the centrodorsal plate ; besides this the 

 capsule must be either scraped out with a fine scalpel or else 

 destroyed by free painting with strong acid. Specimens in 

 which the centrodorsal plate has been simply snipped ofi", 

 though they lose temporarily the power both of swimming and 

 of righting themselves, yet regain these more or less com- 

 pletely after an interval of half an hour to an hour. If, 

 however, sufficient care has been taken to entirely destroy the 

 central capsule the loss of power is absolute and permanent. 



Experiment 9. — The centrodorsal plate of an active speci- 

 men was removed, and the central capsule entirely destroyed ; 

 the cavity was also very freely painted with nitric acid so as to 

 expose and destroy the pentagonal commissure connecting the 

 axial cords together at their roots (cf. fig. 3). After being left 



