Echinoid Tests, Diademoida. 121 
confluent. Its transverse diameter is always greater than its meridional diameter, 
sometimes nearly twice as great (e. g. specimen c); but it is not necessarily greater 
than the greatest height of the plate (e. g. specimen f, fig. 203), for the plate may be 
reduced in height in the neighbourhood of the tubercle, or in some part the rebate 
on the adoral margin may project considerably. 
The extra-scrobicular surface is covered with miliaries, numerous but distinct, 
and varying slightly in size; the larger of them in the larger plates (e. g. in a) 
tend to become tubercles, i. e. to be mamelonate. In specimen a they are about 
4 to 1.4 mm., but rather closer in the adradial tract and more variable in size. 
In specimen b about 4 lie within the height of the plate, viz., 1.3 mm.; here they 
tend to lie in transverse and meridional rows (fig. 201). In specimen m some trans- 
verse rows, corresponding with the denticles, are clearly seen in the adradial tract 
(fig. 207). Although the miliaries come close up to the scrobicule, not any are so 
distinguished from the others in size or arrangement as to be called scrobicular 
tubercles. 
The adradial margin is distinctly scolloped. The plate, as has been said, is 
bent down to this margin; but the curvature is more pronounced on the outer 
surface, with the consequence that the plate thins out gradually towards the margin. 
This produces the effect of an exceedingly acute bevel on the inner surface, as 
in m (fig. 205). That surface may however have a distinct though slight additional 
bevel, as in e and g (figs. 200, 202). In some specimens, as e, the limit of the latter 
bevel is marked by the usual ridge parallel to the adradial margin; this, however, is 
always faint, and may be entirely absent in other specimens. 
The number of denticles on the adradial margin of each plate is 5 or 4, 
according to the height of the plate. Though not very sharply cut, they are perfectly 
distinct, and the grooves between them run far back (being often traceable over the 
ridge) and die away into the surface of the plate. Sometimes the denticles are quite 
simple, as in e, g, and m; sometimes they are slightly excavate on the inner surface 
near their adradial ends, as in +; sometimes this excavation is more pronounced 
at a distance from the margin, where it may give rise to hollows which alternate 
with the grooves and seem to supersede them, as in o (fig. 198). These depressions, 
like those of Triadocidaris immunita (p. 81.) are similar to those frequently seen 
on the main ridges of loose or of articular unions in various Echinoderms, and 
were probably for the attachment of stroma-strands, either muscular or ligamentar. 
The transverse margins are strongly bevelled, but each bevel is stopped by 
a ridge. In the bevel facing outwards the ridge is on the outer edge; this is taken 
to be the adoral margin. On the adapical margin, which has its bevel facing inwards, 
the ridge is on the inner edge. This combination of ridge and bevel may be described 
as azgroove; or, considering the great projection of the bevel, it is better called a 
rebate. The bevel projects more in the interradial part of the plate, while it almost 
disappears at the adradial end. If two plates are fitted together, it will be seen 
that the bevelled surfaces do not meet, but that the bevel of each rests against the 
ridge of the other. Thus there can have been no regular imbrication, nor can the 
plates have been tongued together; but there was a loose union, and a space of 
roughly rectangular section between adjoining plates was presumably filled with stroma. 
The interradial margins are bevelled in the same direction as the adjacent trans- 
verse margins; but the bevel is not so strong, and there is no ridge (figs. 202, 203). 
o, 2 
