Morphological Results. 251 
rocks the nearest approach to the Cidaridae is presented by the Archaeocidaridae 
(or Lepidocidaridae), and there is no reason to doubt that Archaeocidaris (= Echi- 
nocrinus) is a true ancestor of Miocidaris. It has been fully shown to how great 
an extent the flexible union of the plates characteristic of the Archaeocidaridae is 
preserved in many of the Permian and Triassic Cidarids. The only point on which 
further evidence is needed is the transition from the multiserial to the biserial interam- 
bulacrum. Perhaps the apparently peculiar form from the Permian of the Saltrange, 
Cidaris forbesiana DE Kon, will ultimately elucidate this. 
Of the Order Diademoida, the first representatives appear, somewhat doubt- 
fully, in the Cassian beds, and more certainly in the Raiblian beds. All, so far as 
they are capable of exact determination, belong to the Family Diadematidae, and 
apparently to the genera Mesodiadema and Diademopsis (see pp. 102—128), though 
one plate from the Cassian of Cserhat is doubtfully referred to the Acrosaleniid 
Eodiadema, a genus in which Lampert has also placed Cidaris regularis Mbwst. 
(p. 101). 
The evidence available points to the derivation of these forms from the Cida- 
roida, and probably from the Cidaridae (see p. 116). As to their origin from Cida- 
roida there are no reasonable grounds for hesitation. In many features the Diade- 
moida agree with the Cidaroida, the main differences being these: The Diademoida 
possess external’ branchiae, sphaeridia, ophicephalous and trifoliate pedicellariae, and 
a lumen in the primary radioles. We do not know whether the earlier Diademoida 
had sphaeridia or pedicellariae like the later ones; but since these structures must 
have been evolved gradually, the precise date of their appearance is of no great 
importance. They may have arisen concomitantly with the branchiae, or they may 
have been developed subsequently. The detailed description of numerous radioles in 
the preceding pages has shown clearly that the lumen arose gradually by the 
breaking down and resorption of the axial complex; it has also shown that the 
regular arrangement of wedge-like septa had scarcely been reached by the most 
Diademoid of Triassic radioles (see especially p. 225). 
A more important difference is the possession of external branchiae. These 
are paired interradial diverticula from the peripharyngeal sinus, which pass to the 
exterior at the margin of the peristome, just where the interambulacrum meets the 
adjacent ambulacrals. In dry tests or in fossils the presence of these branchiae is 
to be inferred from the presence of the buccal or branchial clefts, which notch 
the peristome at these points. In many early species, however, which all writers 
agree to refer to the Diademoida, these notches are very feebly developed, and it 
is quite obvious that the branchiae may have existed without the notches, and that 
the circle of the peristome may still have been as complete as in the young specimens 
of Strongylocentrotus drocbakensis figured by Loven (1892, «Echinologica» pl. iv). 
In the ontogeny of living forms, the notches arise by resorption of the interambu- 
lacrals, and are preceded by the branchiae. The same course was probably pursued 
in phylogeny. We cannot therefore expect any definite palaeontological evidence 
as to the origin of the external branchiae. 
There are, however, other changes in structure that render the appearance 
of external branchiae about this time quite probable. The name «branchiae» implies 
that these organs are essentially respiratory. This may be the case, but they are 
not wholly so, and their inception may have been due to another cause than the 
