THE wil'.i LA< B \. 57 



the siiU- plates lie against the lateral borders of the Lancet-plate, but do not rest upon 

 it so as to conceal it in any way. In all genera with limar ambulacra, on the other 

 hand, the side plates actually rest upon the lancet-plate, which is moulded foi their 

 reception, and nothing of it is visible externally except the crenulated food-groove 

 along its median line (PI. IV. fig. 4; PI. VI. fig. 10 ; PI. VII.figs.14, L5; PI. Nil), 

 fig. 15 ; PI. IX. figs. 7, 16 ; PI. X. figs. 11, 12, 17 ; PI. XI. figs. 8, 11, 14. 15 ; 

 PI. XIII. fig. 14). Sometimes, indeed, there is no food-groove visible on the lancet- 

 plate (PI. IV. figs. 12, 15; PI. V. figs. 3, 4, 7, 19 ; PL XI. fig. 4 ; PI. XII. fig. 12; 

 PI. XIII. fig. 19 ; PI. XIV. figs. G, 9, 11). 



The broad and more or less petaloid ambulacra of Pentremites are shown in PI. I., 

 and the relation of the side plates to the lancet-plate is also seen in PI. XII. figs. 13, 

 14, and in PI. XVI. fig. 21. The radial sinus is so wide that even the large lancet- 

 plate is insufficient to fill it, and there is therefore a more or less extensive gap on 

 each side of the lancet-plate which is filled up by the side plates. Each of the latter 

 consists of two parts — a somewhat squarely cut inner end or body, which rests against 

 the lancet-plate, and a narrower outer portion or handle. This is continuous with 

 the distal part of the body, the proximal edge of which is cut away into a more or 

 less marked curve. Hence, although the bodies of the side plates are in contact, the 

 handles are not, being separated by a series of pores which lead down into the hydro- 

 spire-canal, but are partly blocked at the distal margin by the outer side plates, well 

 shown in PI. I. figs. 1-3. They have fallen away in the specimen represented in 

 fig. 8, so that the pores appear unnaturally large. 



In all the true Pentremites the sloping side of the radial sinus is marked by a 

 series of transverse ridges with intervening furrows (PL I. figs. 6, 7 ; PL V. fig. 28). 

 Each of the latter receives the handle of a side plate, and thus corresponds to a pore 

 leading down into the upper part of the hydrospire-sac, the folds of which may 

 be partially exposed, as seen in PL I. figs. 6 & 7, and in PL XII. fig. 13, or completely 

 covered by the under lancet-plate, as shown in PL XII. fig. 16. 



Each ambulacrum is thus traversed by two longitudinal sutural lines, which mark 

 off the two rows of side plates from the lancet-plate between them, as is also seen in 

 Cryptoschisma Schulzi, Phamoschisma caryophyllatum, and Orophocrinus OrMgnyanus 

 (PL I. figs. 2, 3, 5, 6, 8-10 ; PL XIII. fig. 20; PL XIV. figs. 3, 18 ; PL XVI. 

 fig. 21). But these lines are sometimes almost or quite invisible (PL I. figs. 1, 

 4, 11) ; and as a rule they are not very distinct, except in weathered specimens 

 (PL XVI. fig. 21). 



Each side plate is thus wedged in between the sloping side of the radial sinus and 

 the lancet-plate, so that its upper surface is more or less level with that of the lancet- 

 plate against which it rests ; while the sutures between successive side plates lie in 

 faint grooves, which exactly correspond in position with the alternating lateral 

 branches of the food-groove in the median line of the lancet-plate (PI. 1. figs. 1-3). 



I 



