138 Triassic Echinoderms of Bakony. 
of A. testudo are absolutely indistinguishable from a few of A. Buchi. This, say 
the objectors, is enough to show the invalidity of A. testudo. I do not agree. A better 
method is to compare the whole assemblage of A. testudo with that of A. Bucht. 
Setting aside the facts, above alluded to, that young radioles, or radioles from the 
actinal surface, may be similar in distinct species, it is admitted that the limits of 
variation in one species may encroach on the limits of another species. But two 
circles are not identical because their circumferences cut one another. If the var- 
iations of A. testudo and A. Buchi be represented graphically, the curves of the 
two species will be distinct. So it is with other species. Such assemblages indicate 
differences of time or environment, and to call them by the same name is to predicate 
an identity that does not exist. 
A similar argument applies to assemblages occurring at the same time and 
place, but generally distinct in form. Cidaris alata and C. dorsata, as they occur at 
St. Cassian, are excellent instances. Normal examples are perfectly distinct, and yet 
there occur radioles about which no two authors are agreed. But if a large number 
of radioles be compared, it is seen that such intermediate forms are comparatively 
rare, and that the graphic representation of the whole material results in a markedly 
double curve. 
The consideration of all these modes of variation therefore justifies the retention 
of many of the old divisions and names, not merely on the ground of practical 
convenience, but because the concepts actually are distinct. Where I have merged 
two of the old species, it has been because examination of the material, and 
especially of the type-specimens, has failed to reveal characters capable of verbal 
or numerical expression. 
It has been the custom to refer nearly all Triassic radioles to Cidaris. Now 
that the genus Cidaris has been split up, this course is inadmissible. Whatever 
Cidaris s. str. may be — a question not yet satisfactorily settled by zoologists — 
it certainly is not known to occur in the Trias. Further than this, many of the 
radioles do not even appear to belong to Cidaridae. Except in the case of Anaul- 
ocidaris, there are no good grounds for referring any particular radiole from Bakony 
to one genus more than another. It therefore seems safest to use the word Radiolius 
in place of any accepted generic name, just as, in the case of columnals, one is 
occasionally driven to the non-committal word Entrochus. In referring to species 
previously described, one may of course use, without prejudice, the name under 
which each is best known. 
Anaulocidaris. 
1879. Anaulocidaris K. A. ZirreL, Handb. d. Palaeont., Palaeozool., I, p. 486. . 
The genotype is A. Suchi (MUNst.), concerning which the literature is as follows: 
1829. Cidarites Buchit MONSTER in GoLDFuss, Petref. German. I, p. 121, pl. XL, f. 5a, 4. 
1834, > » Munster, Neues Jahrb. f. Min., 1834, p. 2 and p 8. 
1841. Cidaris Buchti MGwsteER, Beitr. z. Petrefactenk. 1V, p. 48, pl. Ill, f. 11. 
1841, > remifera MUNSTER, op. cit. p. 43, pl. III, f. 12. 
1852. » Buchiit Mtnst., F. A. Quenstept, Handb., d. Petrefactenk., p. 579, pl. XLIX, f. 10. 
1855. » » Muwnst., E. Desor, Synops. Ech. foss., p. 20, pl. Il, f. 8. 
1855, > remifera MUnst.,? = C. Buchiz var., E. Desor, op. cit. p. 20, pl. Il, f. 11. 
1865. » Buchit Mtwsr., G. C. Lause, Denkschr. Akad. Wiss. Wien, Math.-Naturw. Cl. XXIV, 
Abth. 2, p. 288, pl. X, f. 2 (includes C. rem7fera, as is the case in all subsequent authors). 
