116 Triassic Echinoderms of Bakony. 
genus. In the diagnoses and descriptions which follow, the presence ofall these 
characters is to be understood. 
It may, of course, be objected that the reasons for placing these species in 
the Diadematidae instead of the Cidaridae are insufficient. If placed in the latter 
Family, they would fall most naturally into Triadocidaris, and this would get over 
any difficulty that may be presented by the denticulate overlapping adradial suture 
or by the discontinuous perignathic girdle. Triadocidaris immunita is not so very 
different from Mesodiadema margaritatum but that the two might be placed in a 
single genus. It cannot, however, be denied that M. margaritatum is far removed 
from the normal Triadocidaris, or that its general external appearance is that of a 
Mesodiadema. It has already been shown that the denticulate adradial suture is no 
real bar to its position in the Diadematidae. Can one say the same of the peri- 
gnathic girdle? Undoubtedly this is of Cidaroid type in one of the specimens, and 
resembles that seen in Triadocidaris immunita. In the other specimen, however, 
there is only a ridge devoid of projections and not unlike that characteristic of 
Glyphostomata in general (see P. M. Duncan, 1885, J. Linn. Soc., Zool., XIX, p. 179). 
In the absence of ambulacra it is impossible to say whether ambulacral processes 
were present or not. On the assumption that the early Diademina were derived 
from Cidaridae, one would expect to find a gradual change in the perignathic girdle ; 
and a somewhat greater size of interambulacral processes, or a less size of ambulacral 
processes, would be only natural in any early Diademina. Since there is no published 
evidence on this point, I have prepared some_specimens of Diademopsis Bowerbanki, 
and find that the ambulacral processes are very slightly developed. 
It seems reasonable then to regard these Triassic Mesodiadema as intermediate 
between Triadocidaris and the later species of Mesodiadema. Just as we have 
already seen that the change from the Streptocidarid type to the Stereocidarid was 
a very gradual one, so we learn from these species that the change from Cidaridae 
to Diadematidae was also gradual. It may be that the new types of structure, when 
once developed, multiplied rapidly; but the more we learn, the more clearly we see 
that there was no sudden jump. ‘ 
The existence of these Triassic species confirms the view of those who have 
regarded this genus as primitive; but it does not prove that Mesodiadema was 
ancestral to such genera as Hemipedina and Diademopsis. On the contrary, in 
Mesodiadema the suppression of scrobicular and other secondary tubercles inherited 
from the Cidarid ancestor has advanced far beyond the stage reached in those two 
genera. In that respect Hemipedina (s. str.) is much more like a Cidarid. Hemipedina 
(Diademopsis) incipiens presents a stage from which the later Diademopsis may be 
derived far more readily than from Mesodiadema. Some of the ambulacra to be 
described later show that other primitive Diadematids existed in the Bakony Trias, 
and it is possible that they were the links that led to Hemipedina (s. str.). 
The conclusions of this discussion may be summarised, and the 
evolution of the genera therein mentioned may be provisionally set forth in the following 
hypothesis: — Beginning with Triadocidaris, there was a gradual change from Cida- 
ridae into Diadematidae, a change which may also have recurred at a later stage of 
Cidarid evolution. Mesodiadema, however, appeared early as a streptosomatous form, 
and gradually assumed stereosomatous characters. The main line of descent tended 
towards the paucituberculate Hemipedina; but, at an early period, some of the 
