188 PROE. p. MARTIN DUNCAN ON THE 



the actinal surface of the test with the first tubercles of the 

 interradial area (fig. 10). 



Although the sutures of this species are so readily separable, 

 and the plates can be isolated so easily, still no separation wiU 

 take place down the median line of a ridge (figs. 11 & 12). On 

 the other hand, it is noticed that one of the zones (6) of the in- 

 terradium has a large plate coming to the base of the ridge, and 

 separated from the plate which forms the bulk of it by a trans- 

 verse line of suture, whilst the other zone {a) has a small plate 

 which forms only a small part of the ridge in advance of the 

 plate of zone h. The succession of large interradial plates, and 

 the presence of a very low plate in one zone, and of the great 

 plate of the ridge, which has no median or other suture, are very 

 constant peculiarities in this and other specimens (fig. 11). The 

 direction of the inner or peristomial sutures of the plates at the 

 base of the ridge is from above, inwards and downwards obliquely ; 

 and the relatioa of the large plates to their tubercles on the 

 actinal surface of the test can be easily seen, but that is not the 

 case with the small plate, for usually it is too high up (fig. 12), 

 nevertheless it may have a vestige of a primitive tubercle. The 

 great plate (zone a) which forms all the rest of the I'idge is 

 evidently placed over the first large tubercle of its zone of the 

 interradium. Both plates in zone h have tubercles actinally. 



In a smaller specimen of the same species the ridges were easily 

 separated from the adjacent processes at the lines of suture, and 

 the separated faces of the ridge showed lines of sockets and in- 

 termediate lines of depression and furrowing * (fig. 12). These 

 corresponded with knobs and ridges on the separated face of the 

 process ; and when both surfaces were studied, it became evident 

 that a third and upper plate entered into the composition of the 

 ridge (figs. 12 & 13). 



Were it not for the presence of the relics of a suture on the 

 sides of the ridge there would be no reason why the undivided 

 plate of a ridge should not be named plate 1, and be made com- 

 mon to both zones, as iu so many edentulous Echinoidea. Then 

 the plates in zone a would be JS'os. 2 & 3, &c., and in zone 5, 2 & 3 

 (fig. 11). But this cannot be correct. Plates 2 cannot be thus 

 numbered, for there are at least three plates. (See further on.) 



The examination of the ridges proves that they are composed 



* See Duncan, Journ. Linn. Soc, Zool. vol. xvi. p. 353. 



