THE NERVOUS SYSTEM OF ANTEDON ROSACEUS. 295 



enclose the radial cords, but merely form calcareous plates between the 

 cords and the integument, which later on thicken, grow round, and 

 enclose the cords completely. 1 In this early stage the relations of the 

 radial cords are very similar to those of the ambulacral nerves of an 

 adult Ophiurid or Echinid, 2 and as the latter have certainly acquired 

 their adult condition by becoming detached from the epidermis and 

 shifting inwards, so also may the same process be supposed to have 

 occurred in the Crinoid. The subepithelial bands of the Crinoid retain 

 their primitive positions, but the delicate connective-tissue lamella 

 that sometimes separates them from the overlying epithelium in 

 Antedon 7vsaceus } and is a far more evident structure in Antedon 

 Eschrichtii and in Actinometra, probably represents the earliest stage 

 in the process by w T hichthe nerve becomes detached from the epidermis 

 and shifted inwards. Again, the external and internal plexuses of 

 Echinus, with their connecting fibres in the substance of the calcareous 

 test offer us a condition of things in some respects approaching that 

 of the Crinoid. 



Concerning the morphology of the central capsule, I feel in much 

 more doubt. Dr. Carpenter's observations lead to the belief that, at 

 any rate in its present form, it is connected with the change from the 

 pedunculate to the free-swimming condition; and it is worthy of notice 

 that the two actions with which it has been found to be specially 

 concerned physiologically, i.e. the movements of swimming and of 

 righting, are ones that the pedunculate form, from the very nature of 

 things, can never exercise. 



While, however, this theoiy of the derivation of the system of the 

 central capsule and axial cords of a Crinoid, by concentration from the 

 antambulacral portion of a continuous nerve-sheath, renders a compari- 

 son between the Crinoids and the Echinoderms possible, it still leaves 

 the gap between the two groups a very wide one. Crinoids are some- 

 times compared with Asterids or Ophiurids, but they differ from both 

 these groups in a great number of points of fundamental importance. 

 In the absence of any representatives of ambulacral ossicles, the 

 convoluted character of the alimentary canal, the position of the anus, 

 the permanent communication between the ambulacral system and the 

 coelom, the replacement (functionally if not morphologically) of the 



1 That this condition is a primitive one is shown by its occurrence in some of the Palseo- 

 crinoids in which the axial cords often lie in grooves, and not in canals in the calcareous 

 plates (Carpenter). 



3 Of course they do not correspond to these. 



