432 PROF, p. it. DrXCAX ON THE STErCirEE OF 



of the compoimd plate at the median line is formed bj' the large 

 primary (no. 3). 



It Tvill be obserred that this arrangement of the component 

 plates is not like that of the species already noticed, in which an 

 occasional fourth plate has been sometimes produced : for in the 

 present instance there is no demi-plate to be seen, and all the plates 

 are primaries. 



The compound plate at the ambitus (fig. 8) has a larger tubercle 

 than the one immediately above, and is larger altogether. It has 

 five plates entering into its composition, and there are therefore five 

 pairs of pores and five peripodia ; these are in an arc, and the 

 third pair from the abactinal edge of the compound plate is the 

 most remote from the median line of the ambulacrum. 



The first, second, and third plates from the abactinal edge are 

 formed after the model of the abactinal plates 1 and 2 of the compound 

 plate above, and they are low and broad primaries with a curved 

 adoral edge. The fourth plate is the largest, and corresponds in 

 shape to the third plate from the abactinal edge of the compound plate 

 above ; it is on the type of the middle plate of a Diadema, and is 

 expanded towards the median line, and is lower on the tubercle, 

 the actinal half of which, and sometimes more, it carries. 



The last and actinal plate of the combination (no. 5) is a primary, 

 low at the median suture, slightly higher at the poriferous zone, and 

 with an arched aboral edge which curves towards the mamelon and 

 just reaches the tubercle. The adoral edge of the plate is trans- 

 verse and straight, and it is the actinal boundary of the compound 

 plate. 



Thus the three actinal plates of this compound plate resemble in 

 shape and in position the three adoral plates of the compound plate 

 above, and it appears that the additional plate of the ambital com- 

 pound is either the fourth or fifth from the actinal edge. Gfreat as 

 the downward growth-pressure must have been, there are no demi- 

 plates. I have not had access to some of the most important poly- 

 porous species figured in Dr. Wright's monograph, but which do 

 not belong to his collection, and I can therefore only assert that in 

 all the forms of the group which have been examined by me there 

 is an absence of the demi-plate. 



Considering the first two groups of forms hitherto named Pseudo- 

 diadema, it is evident that the ambulacral structures unite them, 

 and at the same time separate them from the polyporous group. 

 It is interesting to note that these simple forms are the oldest; 

 they differ from the recent species of Diadema in shape and size 

 at maturity, and in the comparative height of their plates. More- 

 over the occasional demi-plate in the tubercle-bearing plate con- 

 stitutes a distinction ; for this is not seen in the modem forms. 

 The details of the peripodia of the ancient and modern forms 

 are not quite the same. 



With regard to the spines there is much difficulty in making very- 

 definite distinctions, and there are many loose spines found in the 

 Secondary series of rocks which are comparable with those of Diadema. 



