Qn 
Echinoid Radioles, Cidaris alata. ; 170 
First, it should be noted that the cotypes of AGassiz by no means. repres- 
ent the norm of the species as developed at St. Cassian: they are distinctly 
more claviform, and bear pustules more equal on the two faces but less regular 
in distribution. The original specimens were found at Buchenstein; therefore they 
may not be of the same age as the Cassian specimens, but probably come from 
Upper Wengen beds.* 
It was no doubt some such consideration that led A. p’Orpicny to separate 
from the C. alata of Acassiz (as a distinct species, C. swbalata), the Cassian 
radioles described by Minsrer. The name is unfortunate, since the Cassian radioles 
are more alate rather than less; but it may be of some service to use it in either 
a varietal or mutational sense, calling the Cassian forms «Cidaris» alata subalata. 
Secondly, it appears that the radioles of «Cidaris» alata from Bakony may 
be separated into two sets. Those from the Cassian beds of Cserhat are either 
like C. alata typica or C. alata subalata. 1 am not clear as to the age of bed 7 
at Section XI, but one radiole from here is apparently subalata while the other is 
of the usual Raiblian pattern. Radioles of the latter type, as found at Cutting | 
and at Jeruzsalemhegy differ in various features from both typica and subalata, and 
must be regarded, if not as pertaining to a distinct species, at all events as repres- 
enting. a mutation, which | shall name poculiformis. Now this form departs 
‘from the claviform type far more than does C. alata subalata. 
The suggestion may therefore be hazarded that C. alata typica, subalata, and 
poculiformis constitute an evolutionary series, which has branched off from the 
line represented. by C. dorsata. Evidently the C. dorsata of the Cassian and 
Raiblian beds cannot itself be the ancestor; but it may be regarded as having 
maintained the radiole characters unaltered. A form with similar slightly flattened 
radioles probably preceded C. alata typica, and was in its turn descended from a 
form with regular claviform or subcylindrical radioles. 
The frequent appearance of ridges does not, so far as I can see, point to an 
ancestor with longitudinally ridged radioles. It is obvious that these ridges either 
are composed of fused pustules, or tend to break up into pustules. The next stage 
either preceding or succeeding, consists of pustules arranged in longitudinal rows, 
sometimes with a trace of transverse rows as well. In the next stage, the pustules 
are irregularly dispersed. That the direction of evolution was at first from the 
irregular, arrangement, through the linear, up to the ridges, and not vice versa, is 
the view that accords better with the supposed order: typica, subalata, poculi- 
formis, and with the supposed descent from the C. dorsata type; it is also the 
view more in harmony with the stratigraphical distribution of the forms. But, 
accepting this view, it must be admitted that the tendency to ridges did not progress 
far: in C. alata poculiformis one sees rather an increase in size of the adapical 
pustules with a return to irregularity or even a tendency to transverse rows. 
The Microstructure of the Radioles throws some light on the inter- 
relations of the species and subspecies, and a more exhaustive study than has been 
possible would doubtless lead to still more definite results. Sections of a typical 
peripheral ‘radiole of C. dorsata from St. Cassian (fig. 438) show no external modi- 
* Dr. Maria Ocitvie-GorvoN in litt. 26. Dec., 1907, suggests as their probable horizon bed 
No.7 ef the table facing. p: 16 in her paper, Quart. J. Geol Soc. XLIX; 1893. 
