Dr. F. A. Bather — Studies in Edrioasteroidea. 163 



crushed, but the arched cover-plutes, instead of being puslied into the 

 groove, have resisted the pressure and given way at the summit 

 of the arch. 



Here we may turn to the differences presented by tlie subvective 

 skeleton of JE. levin. T}ie floor-plates are not separated from the 

 interradial plates by a continuous line of curved sutures, but join 

 them by irregular angular sutures, varying in accordance with 

 the outline of the plates against which they abut. And, just as 

 they merge in this way into the interradials, so are they not 

 distinguished from them by so sudden a rise above the general 

 level as is the case in E. hUjshyi. Thus the rays of E. levis are less 

 conspicuous, and, by reason of this as well as of their more regular 

 curvature, have not that curious appearance of independent life. 



The relations of the floor-plates to the cover-plates are also slightly 

 different. A single cover-plate, instead of abutting on a single floor- 

 plate, abuts as a rule on two. Consequently each floor-plate, instead 

 of a single curved facet for its cover-plate (with a possible minute 

 facet at each side thereof for an accessory plate), has one long curved 

 facet for the cover-plate hinged to it, and a ver}' short straight facet 

 where the adjacent cover-plate plays against it. The longer curved 

 facet, where distinctly observed, is on the distal side of the floor- 

 plate. There are no accessory plates. 



On the floor of the groove, the perradial zigzag channel is less 

 marked, and the side-branches from it less pronounced. The surface 

 of the plates betweeii these branches is gently rounded, almost flat, 

 and not ridged. The outer pore-depression and the inner, more 

 adradial, depression are well marked, but not very distinct from 

 one another. The specimens do not yield convincing evidence for 

 or against the passage of pores to the interior. 



The appearance of additional cover-plates along the median Hue is 

 very strongly marked in certain regions of E 15900 (Text-fig. 2). 

 These regions are just where the ray curves more sharply and passes 

 along the periphery, and is therefore most subjected to both natural 

 torsion of the cover-plates and post-mortem pressure. The former 

 force seems to have induced an adcentral bending of the cover-plates 

 as they approach the median line, and this is best seen towards 

 the proximal end of the disturbed region (Fig. 2, II). Nearer the 

 periphery the latter force has compressed the grooves so that the 

 cover-plates have been raised, as in E 16054, and pushed together. 

 Consequently their adcentrally directed ends have been broken across, 

 and in some places seem to form a double row of small plates, 

 alternating with each other and with the main cover-plates from 

 which they are derived (distal end of Fig. 2, II). Occasionally each 

 of these small plates is again divided by an apparent suture transverse 

 to the median line of the groove, so that two small plates go to each 

 main cover-plate (Fig. 2, III). The regularity of these appearances 

 is diminished in proportion as the specinaen is cleaned and carefully 

 examined. None the less, the divisions between the plates look like 

 true sutures, and they may represent a structure that arose naturally 

 during life rather than the consequence of some accident or of post- 

 mortem pressure. 



