144 F. A. BATHEB, CBINOIDEA OF GOTLAND. 



high, with a wide shallow ventral groove. The axial canal can be traced from the radials 

 some way down the arms (fig. 237). The ossicles show no trace of ornament. 



The covering-plates consist of an outer row of pentagonal pieces on each side, and 

 two median rows of quadrangular pieces, which alternate with one another and with the 

 outer or side pieces (fig. 236). The arrangement is not always verv orderiv, and it seema 

 as though the median pieces had originally formed the apices of compound conical covering- 

 plates, as in C. acinatubus, and had been shifted ont of line with the side pieces; at any 

 råte the side pieces can hardly be called anambulacral, or even adambulacral, for their 

 relations to the ventral groove are precisely the same as those of the pieces lettered c 1 

 in C. acinotubus (tig. 238). When the ventral surface of the arm is viewed directlv from 

 above, only a small portion of the brachials can be seen on either side of the covering- 

 plates (tig. 236). The number of side-pieces to a brachial is 3V 2 — 4 (fig. 235). The 

 covering-plates are quite smooth. 



IBr 3, 4 or ö. IBrj is sometimes almost semilunar in outline. 



LIBr 3, 4 or 5. 



HIBr; all numbers between 4 and 10 have been observed. 



IVBr may reach 14. VBr and VI Br have also ma ny ossicles. 



In specimen c, from which the above measurements were taken, the height of the 

 calyx is 19 mm., the length of the arms 03 mm. Specimen a has a rather smaller calyx 

 but stouter arms; the height of the calyx is iö.5 mm.; the arms have been traced to 

 90 mm. 



Anal structures: x is hexagonal, supporting three plates of the sac, of which the 

 median one is of exactly the same shape as x, and is rather larger than the other two. 

 The sac is composed of longitudinal rows of normally hexagonal plates, alternating with 

 one another; it is best seen in specimen b. This specimen in no way possesses the smooth 

 and regular appearance shown in Angelins figure; but it is considerably crushed and, 

 especially on the posterior side, very much battered, rubbed and weathered. A more nor- 

 mal state of the plates may be observed on the anterior side of the sac (fig. 242). The 

 plates, we thus see, were plane or possibly convex; there is no reason to suppose that they 

 were ever concave as stated by Angelin. This is contirmed by such parts of the sac as 

 are visible in specimen a, where the sac is by no means so perfect as represented in 

 Angelins figure. Where the plates are well-preserved the suture-lines are perfectly 

 straight; there is not the slightest waving either apparent or real: this indeed is correctly 

 expressed in Tab. XX, tig. 7. The apparent crenelation that is produced when the plates 

 are slightly weathered is due to the intimate structure of the plates. Where some of the 

 plates are broken right across they are seen to have had their stereom disposed in a 

 radiating manner, as in the cup-plates of so many Crinoids. When, therefore, weathering 

 has unequally affected the original stereom and the infilling calcspar, this radiating struc- 

 tnre is very clearly brought ont (tigs. 239 — 241). 



A folded structure of the ventral sac is so common among Fistulata in one förin 

 or another, that it is probable that the plates of the sac were invariably deposited as 

 radiating bars of stereom in a, retieulate perisome. Circumstances would lead to a greater 

 exhibition of the structure in some speeies than in others. That it is not discernible in 



