IV. Ophiuroidea. 17 



connective tissue of the gullet. - - Both nerve ring and radial cords are formed 

 as in the Asterids of an epithelium of filamentous cells , the nuclei of which have 

 been taken for nerve cells by Kohler and others [see Bericht for 1887 Ech. p 7]. 

 Among the bases of these high cells lie the nerve cells and fibres. The radial 

 cords give oft" nerves to the iutervertebral muscles, 2 pairs per ossicle in Ophiocoma 

 scolopcndrina, but only 1 pair, which soon forks, in other species. The extension to 

 each tentacle retains its structure of supporting epithelium and fibres, and forms a 

 ring round the base of the tentacle. The tentacular nerve starts from the inner 

 side of this ring ; while the outer side gives off a bundle of fibrils which enters 

 the peripheral layer of the test and supplies the spines. One or two small 

 branches pass to the muscles of each spine, but a principal one forms a ganglionic 

 swelling containing nerve cells and then enters the spine , piercing its articular 

 face and finally dividing into numerous small branches. The tentacular nerves 

 proceeding from the oral ring form a ring round each tentacle but give off no 

 peripheral branches. The nerve ring of Ophiothrix lies obliquely in the oral ring 

 sinus and on its upper and inner side gives off the process connecting it with the 

 digestive tube. From the lower side 2 nerves come off in each interradius, a larger 

 one supplying the interradial muscle and a smaller one which gives off branches to 

 the dental papillae. In the Euryalids the nerve ring is almost horizontal and gives oft' 

 two nerves from its outer edge which supply the interradial muscle and dental papillae 

 respectively ; while from its inner side there come off large nerves which ramify 

 in the connective tissue forming the wall of the gullet, but are nowhere in relation 

 with the glandular cells. - - The coelom, vascular and ambulacral systems contain 

 identical corpuscles, or amoebocytes. They have long pseudopodia, readily 

 forming plasmodia, and contain refringent coloured granules of haemoxanthin [see 

 supra, p 2]. Respiration and excretion are carried on by osmosis through the 

 walls of the genital bursae or hydrospires, the water entering by the adradial 

 and leaving by the abradial side of each slit. Its circulation is effected by move- 

 ments of the aboral face of the sac and by patches of ciliated cubical cells. In 

 Ophiothrix rosula a prolongation of the hydrospire enters the interradial muscle 

 and ramifies in its substance. The muscular fibres described by Kohler in the 

 walls of the Polian vesicles [see Bericht for 1887 Ech. p 7] are only of a 

 connective tissue nature. The vesicles are lymphatic glands, the inner wall of which 

 consists of a fibrous network forming irregular alveoli, in which the amoebocytes 

 develope and become charged with the albuminogenous ferment haemoxanthin, 

 eventually falling into the cavity of the vesicle , from which they creep into the 

 ambulacral ring. The madreporic ampulla of the water-tube is situated in the 

 thickness of the plate and not free , as described by Ludwig ; and it is lined by 

 the same palisade epithelium as the water-tube. The water-pore of Ophiurids is 

 simple as in young Asterids and almost entirely covered by the buccal plate ; but 

 in Astrophyton there are several pore -canals all leading into a small ampulla, 

 and lined like it by cubical cells ; while the water-tube which follows a nearly 

 straight course is lined by a palisade -epithelium. The only true blood- 

 vessels in the rays are the so called inner perihaemal canals. The outer canal 

 or supra -neural space is morphologically only a portion of the external medium 

 which becomes isolated in the early developmental stages, though virtually re- 

 taining connection with it in those forms in which the nerve ring is continuous 

 with the digestive epithelium . The lining of the radial sinus is not a continuous 

 epithelium, but consists merely of scattered embryonic cells. The mass of fibrils and 

 nuclei occupying the mid-ventral line of the nerve is not a vessel as supposed 

 by Ludwig and Kohler [see Bericht for 1887 Ech. p 7], and its nuclei do not undergo 

 the changes necessary for the formation of amoebocytes. The radial sinus com- 



