46 



so tlial this form is probably to be regarded as a primitive globiferous pedicellaria. 

 Finally in Cœnopedina the latter form has been developed into a curious form of 

 globiferous pedicellariæ, with no glands on the stalk; there are true ophicepbalous 

 pedicellariæ. The stalk of the pedicellariæ is a single rod, or two rods connected by 

 crossbeams in the triphyllous and tridentate pedicellariæ, irregular, complicated in 

 the ophicepbalous ones. In Cœnopedina it is irregular and complicated, also in the 

 triphyllous and globiferous pedicellariæ. (Lissodiadema is imperfectly known as 

 regards the pedicellariæ.) There are thus several differences found in the pedi- 

 cellariæ, but scarcely any of higher classificatory value. We are thus referred to 

 the structure of the test and spines for characters from which to group the genera, 

 and there we note as the most important fact that in Cœnopedina the tubercles 

 are smooth, whereas in all the other genera (except Lissodiadema) they are crenulate. 

 The apical system of Cœnopedina is small, with all the ocular plates excluded from 

 the anal system, very ditïerent from the large apical system of other Diadematids, 

 in which the ocular plates are more or less directly in contact with the anal system. 

 Finally the stout solid spines are very different from the hollow spines of Diadema etc. 

 There are no blue spots on the test. — There can certainly be no doubt that 

 Cœnopedina must be referred to a different family from the rest of the Diadematids, 

 viz. to the fam. Pedinidœ; and to this family I would also refer such fossil genera as Pe- 

 dina Ag., Pseudopedina Cotteau, Mesodiadema Neumayr, Micropedina Cotteau, Leiopedina 

 Cotteau, Echinopedina Cotteau, Hemipedina Wright, Echinopsis Ag:, Diademopsis Desor, 

 Phymopedina Pomel , Hecistocgphus Pomel , Orthopsis Cotteau , Miorthopsis Pomel, 

 Gijmnodiadema Loriol, Palœopedina Lambert. — Among the genera referred by 

 Duncan to the Pedinidœ the genera Stomechinus, Polycyphns and Codechinus have 

 imperforate, noncrenulale tubercles, and Heterocidaris perforate, crenulate tubercles; 

 these evidently cannot belong to this family. I thus quite agree with Pomel in his 

 view of the family Pedinidœ. 



The other recent genera of Diadematids must certainly be referred to one 

 family, the Diadematidœ. To be sure, they are not all very nearly related; thus 

 Astropyga and Chœtodiadema form one group, Diadema and Echinothrix another 

 group, and Centrostephanus again stands apart from the other genera. But to refer 

 these groups to different families would certainly not be correct, the characters 

 distinguishing the groups being only of secondary importance (form of the test, the 

 blue spots, the pedicellariæ). The genus Lissodiadema, I think, must be referred to 

 the same family, in spite of its smooth tubercles and spines; as I have pointed out 

 in the description of this form it suggests Echinothrix in several important charac- 

 ters, and it must probably be regarded as an aberrant form of that type. — 

 Among the fossil genera Placodiadema Duncan and Helikodiadema Gregory must 

 probably be referred to this family. 



A number of fossil genera have been united by Pomei, in a family Pseudo- 

 diudematidæ, dilTering from the true uiadematidœ in having solid, smooth spines. 



