AMBITLACBA OF THE EECEISTT DIADEMATID^. 107 



absolutely a characteristic feature of the family as has been 

 supposed. It exists already well developed in Astropyga, but 

 with the important difference that the overlapping of the plates 

 is in the same direction in both areas. The lower edge o£ the 

 plate passes under the upper edge of the preceding plate " {I. c. 

 p. 71). Put in other words, this means that the adoral edge of 

 a coronal plate slopes from the outer surface inwards, in an 

 actinal direction, and that the aboral edge slopes from within 

 outwards in the abactinal direction. 



The Amhidacr a of Astro-p J ga, radiata.— The tubercles are large 

 perforate and slightly crenuiate, and the base of the boss is wide 

 and flattened out. The mamelon is small in relation to the size 

 of the rest of the structure. At the part chosen for illustration 

 (PI. Y. fig. 9), the great tubercle-bearing plate is apparently 

 followed, apically, by a smaller one carrying a small tubercle, which 

 is placed nearer the pores than the large tubercle. Seen from 

 within the test, the position of the large tubercles is shown by a cir- 

 cular depression (fig. 10) ; and when the clearing agent is applied, 

 it becomes evident that the large tubercle and the small tubercle 

 are not on separate geometrical plates. The large tubercle as 

 well as the smaller are really in relation with odo huge compound 

 plate; and the division of this plate into two is ai'bitrary, for the 

 transverse dividing line betAveen the two plates either abuts 

 against the salient angle of the zigzag in the median line or close 

 to it. Both plates, together with their aggregate of six com- 

 ponent plates, are included in one geometrical, compound plate 

 .(PI. V. figs. 9 & 10). It is best for the purpose of description 

 to consider the great plate as composed of two, each one being 

 made up of a triplet of pore-bearing plates. The edge of the 

 poriferous plates at the interradial suture externally is very 

 irregular, and some pores are quite out of the line (PI. V. fig. 9). 

 The three pairs and the apparently additional fourth are well 

 seen in relation, but the fourth pair of pores is really the lowest 

 of the triplet of the upper part of the combination. Seen from 

 within, the arrangement of the pairs is very easily understood, 

 especially after the study of Diadema and Ecliinothrix. 



The illustration, PI. Y. fig. 10, is a diagram founded on 

 magnified views of three great and geometrical plates ; and is 

 nearly true. It would be quite so were the effects of the bases 

 of the tubercles on the inside of the test shown, but that would 

 complicate the sutural lines. 



