522 A. MILNES MARSHALL. 



tact with the ambulacral epithelium, get carried by the action 

 of its cilia to the mouth. 



Irritation of the ambulacral groove at any part causes the 

 adjacent pinnules to be at once turned forwards, i.e. with 

 their tips towards the free end of the arm, and folded alongside 

 the irritated part, apparently to protect it from further injury. 

 Slight irritation of a pinnule or of an arm causes corre- 

 spondingly slight and local movements ; stronger irritation 

 causes movements of the whole arm, which may spread to other 

 arms, or lead to the animal detaching itself and swimming 

 freely. Irritation of the oral pinnules, however slight, causes 

 them to be firmly closed over the disc, and stronger or pro- 

 longed irritation causes the arms to be flexed strongly, so as to 

 cover the disc, or else the -whole animal to detach itself and 

 swim away. 



If an Antedon be detached and placed with its oral surface 

 downwards, it will right itself almost at once. If the surface 

 on which it is placed be a rough one, the righting movement is 

 effected in a few seconds or almost instantaneously. In a glass 

 vessel it takes longer to perform, but with an active specimen 

 I have never seen more than two minutes spent over the 

 operation. In righting itself an Antedon first flexes all the 

 arms slightly, so as to raise the disc a little above the ground ; 

 then follows a moment of apparent uncertainty as to which 

 arm to use. One arm is then flexed more strongly than the 

 others, so as to slightly lift the disc on that side, the pinnules 

 of the flexed arm being extended and apparently used to push 

 against the ground. Then, after another pause, a rather 

 sudden and violent flexion of the arms immediately adjacent to 

 the already flexed one causes the animal to turn on its side, 

 when a few energetic swimming movements place it right way 

 up. An active animal has apparently the strongest objection 

 to being placed mouth downwards, and will right itself again 

 and again if so inverted. When attached by the cirri^ how- 

 ever, they may, as noticed above, remain in the inverted 

 position for days or weeks. 



If an arm be cut off from an active Antedon, the detached 



