CARADOCIAN CYSTIDEA FROM GIRVAN. 



447 



an average height of '57 mm,, and are approximately equal. Distally they become 

 lower, and seem to alternate in height. In the median region several of the low 

 columnals seem to fuse to form compound columnals, at first retaining the transverse 

 rings of the original columnals, but in the distal part of the region becoming smooth. 

 At the same time they increase in height from 2 mm. to 4*4 mm. 



§ 329. The structure of the proximal region is really more complicated, but it is 

 rather difficult to reconstruct its exact arrangement. The columnals are flanged 

 outwardly, but in some cases the flange seems to be at the top, in other cases at the 

 bottom. In Gl33 one counterpart shows the columnals imbricating distalwards, the 

 other shows a proximal imbrication. 



§ 330. The accompanying diagram (text-fig. 57) gives my idea of the structure of the 

 proximal region : a large flange at the top, sometimes marked off" from the lower part 

 of the columnal by a slight groove ; a small flange at the bottom, normally fitting inside 



muscle-fossa 

 f ulcral ridge 



— _ ligament-fossa 



Fig. 56. Fig. 57. Fig. 58. 



Text-fig. 56. — Brachiole-facet of Chcirocrinus constrictus. This is borne by two proliferated plates at the 

 top of radial 19 in G133. Enlarged about 40 diameters. Cf. PL IV. iig. 51. 



Text-fig.s. 57, 58. — Solid sections of Proximal region of Stem in Cheirocrinus constrictus. Fig, 57 shows the 

 proximal region proper ; fig. 58 the tract where it merges into the median region. These diagrams are 

 provisional interpretations of the appearances presented by squeezes. Cf PL IV. figs. 51 and 56. 



the large flange. On the interior the ossicles have a slight transverse groove. It is 

 possible that the large flange represents an original independent flanged columnal, and 

 that the intermediate region represents an originally inner columnal (" Schaltglied," 

 Jaekel). Naturally it is almost impossible to decide in these casts whether a certain 

 appearance represents a suture or a mere impressed groove. 



Where the proximal region merges into the median region the upper flange becomes 

 very large, the groove between it and the lower part of the columnal accentuated, and 

 that part of the columnal lower but swollen ; thus is produced the appearance of high 

 and low columnals alternating (text-fig. 58). 



§ 331. Occasionally in the proximal region there seems to be a slight indication 

 of tubercles on the flanges. In the median region these have developed into longi- 

 tudinal ridges, well seen in G 133, where there are about eight in a width of '8 mm. 



§ 332. The distal end of the stem is not preserved in any of these specimens. I 

 am therefore unable either to confirm or disprove Prof Jaekel's suggestion ; but, as 

 previously explained (§ 306), all the probabilities are against it. 



TRANS. ROY. SOC. EDIN., VOL. XLIX. PART II. (NO. 6). 59 



