THE IIVI)|;os|'||;i:s AND SIMK.U I.Ks. 99 



really the lateral portions of wide and somewhat petaloid ambulacra. In well-pre- 

 served specimens they are crossed bj fine lines, continuous with, but less distinct 



than, those which start from the median groove. The latter separate the inner ends 

 of the large triangular side plates, while the former separate their outer ends and are 

 usually entirely obliterated (PL XI. figs. 8, 9; PI. XV. tig. 11). It will he seen 

 from the first of these figures that the lancet-plate is broad and nearly fills up the 

 radial sinus. Its sides slope downwards rather steeply from the narrow median 

 groove, and upon them rest the side plates, the section of which at the proximal 

 ends of the ambulacra is nearly an equilateral triangle. The upper side is slightly 

 incurved; and that portion of the curve which is immediately next to the food-groove 

 is all that is usually represented as side plate in figures (PI. XI. figs. 8, '•' ; PI. XV. 

 fig. 11). In reality, however, the whole surface between the food-groove and the 

 cleft is formed by the side plates, but the divisions between them are much more 

 marked near the food-groove than they are near the cleft. In fact, the broad outer 

 portions of the plates seem to coalesce so completely that they look like parts of the 

 calyx-plates intercalated between the sides of narrow ambulacra and the clefts, as 

 implied in the quotation given above (PI. XV. fig. 11). But in one specimen we 

 have found that the side plates are readily separable (PL XI. fig. S) ; and it is then 

 apparent that their outer portions really belong to the ambulacra and are not parts 

 of the calyx-plates. An approach to this condition occurs in the Belgian species 

 0. Orbignyanus, in which there is a sort of thickened rim to the wide ambulacrum 

 (PI. XL fig. 10; PL XIV. figs. 16-18). 



This enlargement of the proximal side plates of 0. stelliformis is much more 

 marked in some specimens than in others (PI. XL figs. 8, 9 ; PL XV. fig. 11), and 

 it is occasionally so very pronounced as to give quite a bent outline to that part of 

 the hydrospire-cleft which remains open as the spiracle (PL XV. fig. 11) ; but it rarely 

 extends to any great distance, and in some cases the clefts close up owing to the 

 under lancet-plate meeting the radials but a very short distance beyond the radio- 

 deltoid suture (PL XL fig. 9 ; PI. XV. fig. 13). In other cases the radial sinus seems 

 to be somewhat deeper, and the lancet-plate relatively narrow. The hydrospire-cleft 

 remains at the bottom of the sloping wall of the sinus, and the side [dates rest in the 

 angle between this wall and the lancet-plate. But they do not come in contact with 

 the radial wall till someway down the ambulacrum, and the external opening of the 

 radial sinus, which is the real spiracle, is thus somewhat longer than the internal 

 one by which it communicates with the hydrospire-sac beneath it, as is well shown in 

 fig. 8 on PI. X. 



It may be further noted that in 0. stelliformis (PL XL fig. 9 ; PI. XV. tig. 11) the 

 proximal ends of the spiracles are at a greater distance from the mouth than in 

 0. gracilis (PI. XVIII. fig. 9), or in either of the British species (PL XV. figs. 2, 

 5, S). In this respect they resemble the openings in Schizublastus Sayi (PI. 111. figs. 



o2 



