NERVOUS SYSTEM OF ANTEDON ROSAOEUS. 521 



A. The Movements of Uninjured Specimens. 



The normal position ofAntedon rosacea s, the species on 

 which all my experiments were made, is a fixed one, the animal 

 being attached by the dorsal cirri to some foreign body, and 

 the arms spread out horizontally with their tips slightly flexed. 

 The oral pinnules are bent over the disc, crossing one another 

 above it; the other pinnules are spread out nearly at right 

 angles to the arms. 



In an aquarium containing a large number of specimens the 

 great majority will be found attached either to the bottom or 

 sides of the tank, i. e. with the oral surface directed either up- 

 wards or more or less obliquely ; some specimens, however, 

 are almost certain to be found, if there be foreign bodies in 

 suitable positions for attachment, inverted, with the oral 

 surface downwards. 



An Antedon when once attached exhibits very little ten- 

 dency to alter its position, and may remain fixed in the same 

 place for weeks. If detached, either spontaneously or by 

 force, it can, and usually does, swim actively until it reaches a 

 suitable place of rest, to which it anchors itself by its cirri. 

 The normal swimming movements, which are peculiarly grace- 

 ful, consist in strong flexion of the proximal half of the arm, 

 which is raised vertically over the disc, and then extension of 

 the whole arm, the distal half of which is thrown out some- 

 thing like a whiplash or the line of a flyrod. During flexion 

 the pinnules are folded alongside the arm ; during extension 

 spread out so as to expose as great a surface as possible. 

 Usually two or three arms are raised simultaneously, some- 

 times as many as five, and the only rule I have noticed is that 

 the two arms of each pair are always flexed alternately and not 

 simultaneously. 



When attached by its cirri the arms of Antedon exhibit but 

 very slight movements ; they are usually spread out widely, 

 apparently to expose as large a surface as possible for the en- 

 tanglement of food particles, which, if they once come in con- 



