112 F. A. BATHER, CRINOIDEA OF GOTLAM). 



The posterior interradius is considerably modified by the large anal tuhe. The 

 madreporite, which in some genera is a heart-shaped plate as large as, or larger than, the 

 deltoids, is liere pressed in and partly covered by the anal plates, so tliat on the surface 

 it appears of triangulär outline with the base of the triangle bounded by the large cover- 

 ing-plates that roof in the mouth. In some genera the madreporite is, as its narac implies, 

 pierced by numerous pores; in other genera, however, pores, though they may have been 

 present, are not to be distinguished, and the surface of the plate is merely crinkled or 

 curiously rugose. In the latter state it very closely resembles the various stages in the 

 development of the madreporite of Asterias glacialis as figured by Prof. Loven in his 

 Etudes sur les Echinoidées, Pl. LM, tigs. 266 — 271. In the present species the madre- 

 porite is crinkled rather less than in Lovens fig. 268, and no pores can be made out. 

 (Pl. V, figs. 156, 157). 



On either side of the madreporite, and supporting like it the covering plates of the 

 ambulacra, is a plate of about the same size as the madreporite and of irregular shape. 

 These plates abut on the adjacent deltoids, but do not appear to touch the radials, in 

 which respect they differ markedly from the deltoids. They abut on two other large plates 

 one on either side of the anal tube; and it is these latter plates that rest on the radials. 

 These latter plates also serve to support the lines of small perisomic plates that pass be- 

 tween the radials and brachials. 



In specimen a, and possibly in b, the ventral tube appears to rise rather abruptly 

 from the tegmen, so that the large lower plates of the tube are at right angles to the 

 plates that abut on the madreporite; but in specimens c, d, and possibly g, the ventral 

 tube rises gently, so that all the plates in the posterior interradius rise up towards the 

 vertical wall of the tube. 



In specimens c and <1 the ventral tube has been broken off close to the tegmen; 

 and, in each, nine plates are seen to enter more or less into the base of the tube (see 

 Text-figures 14, 15). These are (1) the madreporite; (2) a tiny plate abutting on the 

 left posterior ambulacrum and hardly forming part of the tube; (3) a plate abutting on 

 the madreporite and the left posterior ambulacrum, and forming part of the tube-wall 



(4) a large plate resting on 1. post. R. and bounded internally by the lumen of the tube 



(5) a plate resting on anal x, and bounded by the lumen; (6) a plate corresponding to 4 

 (7) a plate corresponding to 3; (8 and 9) two plates within the tube, of which only a 

 minute portion is seen from the exteriör, and which are immediately behind the madre- 

 porite. In text-fig. 14 another plate is represented as resting internally on 6 and 7; this, 

 however, is the lower part of the ensuing tube-plate, and appears to have a broken sur- 

 face. The surface of all the other plates is granular showing that they were suturallv 

 united with the plates above them. 



The Stem is quite short. It is composed of ossicles of alternate size. Their arti- 

 eular surfaces have marginal radiating striae but are granular in the ccntre. There are 

 5 radial sutures. Distally the stem ends in a large swollen massive base. The stem, as 

 here described, is only known in one specimen, and in that it is much weathered; it is 

 therefore possible that its shortness is merely abnorma!. This is rendered more probable 

 by the alternate size of the ossicles, characteristic of only the proximal stem region in 



