XI. Balanocrinus. 3893 
in those that are swollen or ridged, but is more marked im 
lobate or stellate examples (cf. d’Archiac’s fig. 16a). At 
first one observes a simple alternation in height, in E 22019 
the heights are 2°5 mm. & 2:0 mm., mean diameter 6°7 mm.; 
then an increase in the diameter of the higher ossicles, thus 
in E 21961 the measurements are 2 mm. & 7°3 mm., and 
1:7 mm. & 7:0 mm., respectively ; finally, columnals of 3 or 
more, perhaps 4, orders, distinguished by both height and 
diameter (EH 21964). In quinquelobate fragments the larger 
projecting columnals are not so depressed radially as are 
the smaller ones, and this produces a somewhat scalariform 
surface (E 21964). 
The Number.of Columnals in an internode, as observed 
in the dozen intersyzygia available, varies from 15 to 82. 
The material is too limited to permit of satisfactory con- 
clusions being drawn, but the following facts may be noted. 
The highest columnals occur in the internodes with only 15: 
in E 21981 the height is 2°55 mm., in E 22027 it alter- 
nates from 3°2 to 2°5 mm. The lowest columnals are in the 
internode with 32 (EH 21983), where they are in three orders, 
from 1°7 to 0°8 mm. This last has a stellate section and 
certainly comes from a more proximal region of the stem. 
Three orders are also manifest in the specimen with the 
next number of internodals, viz., 31 (EK 20982), but the 
height of the ossicles is rather greater, 2°6 to 1:5 mm. The 
two fifteeners are plane and quite smooth (E 22027) or with 
a slight ridge of sparse granules (EH 21981). Those with 
the higher numbers have more or less swollen and ridged 
ossicles as a rule. EH 21975 is exceptional in that, though 
ridged, it is plane and has equal internodals of 2°5 mm. in 
height. On the whole, the facts suggest that the height of 
the internodals was greater in the distal region, but that 
the whorls were at more frequent intervals. This applies 
only to full-grown individuals. While height of columnals 
and the extent of their alternation in size must depend 
mainly on position in the stem, it appears that ridging and 
swelling have not that connection—at all events, in so great 
a degree,—but are characters of the individual. Stellation 
or marked lobation were probably most common in young 
stages and in the proximal region of adult stems; the facts © 
are consistent with that view, but there was certainly some 
individual variation as between circular, pentagonal, and 
slightly lobate. 
The Radial Pore lies at the bottom of a depression, which 
is well marked in some specimens, generally smooth ones 
