250 é Triassic Echinoderms of Bakony. 
from that of the Pentacrininae, since in place of the radial ridge-group, there is 
here only a single radial ridge. 
Among the Jsocrini, the joint-faces in I. candelabrum appear to be the most 
primitive. The petal-floors are still small and unstable, and the relatively long 
crenellae as they near the periphery curve upwards so as to be more nearly parallel 
to the perradius, as they are in «Pentacrinus venustus>. Between these two forms 
there was, we must suppose, a stage in which the crenellae or ridges, instead of 
remaining parallel to the perradius, tended to fan out more from the petal-floor. 
Thus the first stage in the formation of a radial ridge-group would have been the 
loss or alteration of the single radial ridge, and its replacement by two ridges 
meeting in an angle on the perradius at the periphery. In the next stage two 
pairs of ridges would meet in an angle on the perradius, thus forming an incipient 
radial ridge-group; and this stage is actually found in the smaller columnals of 
I. candelabrum (pp. 38, 41). 
Isocrinus scipio (p. 48) also has very few radial ridge-groups, and it is quite 
possible that the single short radial ridge often found in this species is the remains 
of the long radial ridge in «Pentacrinus venustus» rather than a secondary modi- 
fication of the adcentral pair of ridges. 
It is easy to see how these primitive stages progressed towards the more 
developed stage with a number of pairs in the radial ridge-group, and the history 
need not here be followed further. 
No sooner was the Pentacrinine type of columnal established, than it began 
to diverge in three directions, characteristic of the three genera Jsocrinus, Pen- 
tacrinus, and Balanocrinus. Such a form as Isocrinus Hercuniae (p. 48) is of much 
interest because it combines in a partly developed condition characters of all three 
genera. When the radial ridge-groups are composed of a series of ridges at right 
angles to the perradius, it would only need the suppression of the radial triangle 
and the shortening and equalisation of the peripheral crenellae to produce a joint- 
face like that of Balanocrinus. On the other hand, the increase of the radial 
triangle, with the elongation of the petal-floors, would produce the plan of Pen- 
tacrinus, as it appeared shortly afterwards in P. versistellatus ScnaFu. 
Among the Echinoidea, the chief evolutionary problem that confronts us is 
the origin of the Diademoida, with which is probably bound up the origin of the 
whole Sub-class Regularia Ectobranchiata, if not the origin also of all the Irregularia. 
So far as any negative can be firmly established in palaeontology, it appears 
certain that none of the Ectobranchiata or of the Irregularia existed before Triassic 
times. They must therefore have had some Triassic or Pre-Triassic ancestors. 
Apart from the doubtful Tiarechinus, all Triassic Echinoids that are not Diademoids 
are Cidaroids, and in fact belong to the one family Cidaridae. The other Orders 
of Palaeozoic Echinoidea scarcely need consideration. The peculiar Ordovician and 
Silurian genera Bothriocidaris, Palaeodiscus, and Echinocystis may have had des- 
cendants, but were certainly not the immediate ancestors of any Triassic genera. 
The gradual specialisation of the ambulacral plates in the Melonitoida prevents us 
from considering that Order as ancestral to forms with far simpler ambulacrals. 
That the Cidarid type had been evolved by Permian times, we know from the 
presence of Miocidaris Keyserlingi in the Magnesian Limestone. In Carboniferous 
