THE AMBULACRA OF FOSSIL ECHLN'OIDEA. 429 



reaching the boss adorally to the mamelon. The course is then to 

 the median suture, and the termination is close above the adoral and 

 inner angle of the compound plate. The middle plate is low at the 

 poriferous zone, nipped in vertically at the boss, where it includes 

 the mamelon, and expanded towards the median line. 



The adoral plate of this combination has an arched aboral edge 

 and is a low and broad primary, smaller in vertical measurement at 

 the median line than at the poriferous part. 



The pairs of pores are in peripodia, and the amount of arching is 

 slight. This description would suffice for a compound plate near 

 the ambitus of a recent Diadema. 



But the fossil form has some compound tubercle-bearing plates at 

 or just below the ambitus, which are polyporous ; for there are 

 distinctly four pairs to a compound plate, and not three only (fig. 6). 



Pig. 6 (see p. 452). 



There are therefore four plates in the combination, and all are not 

 primaries, there being a small demi-plate (6) amongst them which 

 does not reach the median line. The aboral or first plate (a) of 

 the set resembles th^ corresponding plates of the combination 

 already described, and is a low and broad primary with the adoral 

 edge bent actinally. The second plate {h) is a demi, and it reaches a 

 little way up the tubercle, and is bounded aborally by the edge of the 

 first plate, and adorally by part of the suture of the third plate in its 

 path to reach the median line. Part of the third plate (a) has the 

 shape of the middle plates of the combinations in which there are 

 only three pairs of pores, but it is rather lower, and the fourth plate 

 («") resembles the adoral plates of the triple compound plates. The 

 second plate is the relic of a primary which has undergone absorp- 

 tion owing to that growth-pressure which is so easily traced in some 

 recent forms of Diadematidae. The recent species of Diadema do 

 not, however, present this phenomenon, and there are no demi- 

 plates in them. 



The simplest form of fossil Diadematid is a species which, had it 

 doubling of the pairs of pores close to the peristome, would fall 

 within the specific diagnosis of Pseudodiadema dejoressum, Agass. 



The specimen in my possession was obtained by Prof. J. Morris, 

 M.A., from the Cornbrash of the Chippenham district. It has 

 nearly straight rows of pairs of pores, the outer pores being larger 

 than the inner. There is but sHght obliquity of the pores, and the 

 pairs are not close. There are three pairs to each tubercle-bearing 

 compound plate, and the three plates are primaries of the true Dior- 

 dema type (fig. 7). The commencement of the compound plates is 

 very close to the radial plate, and there are only one or two solitary 



