142 a ee __ Triassic Echinoderms of Bakony. 
In comparing these measurements of thickness with those in the spatuliform 
and trulliform radioles, it should be remembered that in the remiform radioles the 
greatest thickness is theoretically in the sagittal plane, whereas in the others there 
is sometimes a median concavity making the radiole thinner in this plane than on 
either side of it. 
Specimen a is the smallest and specimen f about the largest of the complete 
radioles that fall within the definition of «remiform» given above. Specimens a and 
b are represented on Plate X. (figs. 256—261), specimens e—h in the accompanying 
outline figures. It is easily seen from both measurements and figures that, whereas 
the thickness increases approximately in proportion to the length, the width increases 
much more rapidly. Thus if length be 100, then in a thickness is 12 and width 
28, while in f thickness is 12°9 and width 59. In this respect there is a gradual 
transition to the spatuliform radioles. 
Returning to a, however, we note that it differs very little from the secondary 
radioles of most Cidarids or from the adoral radioles of many. These radioles 
are always those in which the peculiar characters of the species are least manifest, 
and are therefore, one supposes, the nearest to the primitive ancestral form. At 
the same time they must themselves be modified in consequence of their position. 
Thus this radiole, in addition to being compressed, has its axis slightly curved, so 
that one of the flattened sides is convex and the other is concave. The former, 
as will be seen, is the adapical or outer face; the latter is the adoral or inner. 
The terms right and left will be used as coinciding with the right and left of the 
observer, when the radiole has its outer face upwards and its base towards the - 
observer. This radiole is not quite bilaterally symmetrical, but is a trifle longer on 
the left so that the distal margin slopes up towards that side; the plane of greatest 
thickness also appears to lie a little to the left of the sagittal plane. This asymmetry, 
however, does not appear to be a constant character of the more adoral remiform 
radioles. 
The downward curvature of the radiole was associated with a downward 
bending at its attachment, and this has produced its effects on the structure of the 
base. Thus the collar, as it passes from the outer to the inner surface, slopes in 
a distal direction, and so presumably remained parallel to the surface of the test. 
Again, the acetabulum is not merely transversely elliptical, but its rim is excavate 
on both the adoral and adapical margins; on the adoral margin the rim appears 
thrown into a slight fold, as though it had been pushed back; on the adapical] 
margin are traces of a similar fold, but the excavation of the rim passes through 
the fold to the groove limiting it; thus the outline of the acetabulum approaches a 
triangle. In the description of the interambulacral plates referred to this species, it 
has been explained that the: articular surface of the mamelon slopes adorally. In 
so far as this slope affects the whole structure and keeps pace with the downward 
bending of the radiole, it is not easy to see why the acetabulum should assume a 
triangular shape. But if the depression of the radiole were more pronounced than 
the slope of the tubercle, the central ligament would press against the upper edge 
of the acetabulum and would prevent the accretion of stereom at that point. In 
larger remiform radioles the fold on the edges of the acetabulum appears as a 
swollen ring. 
The whole surface of the radiole tends to be longitudinally striate. On the 
