AMBULACRA OF THE EECENT DIADEMATID^. 105 



Still nearer the peristome the compound plates diminish in 

 vertical dimensions, and the pairs of pores become very remote 

 from the interradial suture on account of the presence of the 

 ascending process of the auricule ; but although the arran2;ement 

 of the triplets is a crowded one, the triplets of every compound 

 plate can be distinguished, there being no accessory plates. 



The principal alteration in the arrangement of the triplets is 

 not externally, but near the median line. The first plate of the 

 combination (plate TV. of the series) is low, aud as it approaches 

 the median liae it becomes lower, and finally it does not reach 

 the median line. It is a low demi plate. The second plate com- 

 prises the whole, or nearly the whole, of the expansion of the 

 plate at the median line. The third plate is small, and sometimes 

 it may reach the adoral angle of the median suture, but with 

 increasing crowding it is shut off from it as well as from the outer 

 edge of the compound plate. (Fig. 5, «?' is the line of the auricule.) 



The characteristic arrangement of the triplets in this species, 

 and probably it is generic, is that of small, close, low, and more 

 or less rectangular primaries near the radial plate. More 

 actiually the first and third plates of each triplet become low 

 broad primaries, having their smallest height at the median line. 



Then below the ambitus the third plate is crushed out ex- 

 ternally at the interradial sutural edge. Still more adorally the 

 first plate becomes lower at both ends, and finally it is excluded 

 at the median line and becomes a demi plate in the ordinary 

 acceptation of the term. In no case is there an arrangement as 

 in Strongylocentrotus. 



In Diadema this crowding out of the third plate does not 

 occur, but close to the peristome the first and the third plates 

 barely reach the median line, and are in one or two places demi 

 plates. 



Tlie Median Suture of tlie Ambulacra. — The line of this zigzag 

 is not composed of straight lines forming a succession of angles 

 at their points of union or contact. 



The line of suture is marked by a series of symmetrical curves, 

 and til us where there might have been an angle at the median 

 end of a coronal plate there is a convex process fitting into 

 a corresponding concave line between the two opposed plates. 

 (PL y. figs. 7 & 8.) 



It is perfectly evident that the junction of the edges of the 

 plates along the median line is not by simple apposition along 



JA'S^. JOUEN. ZOOLOGY, VOL. XIX. 8 



