8i8 PEOE. P. il. DUNCAN ON THE MOEPHOLOGT 



the Stage-reflector, is necessary. The specimeu, fixed on wax, 

 should be moved on its axis by the rotating-stage so as to cast 

 shadows in diff'erent directions. This proceeding is especially 

 necessary in carrying out some of the following observations. 



On looking at the actinal edge of the interradial plate next in 

 vertical succession to that just mentioned, it will be found to be 

 covered with a considerable number of minute, well-defined, blunt, 

 conical projections of the reticulate calcareous tissue of the test. 

 These knobs can be seen from the outer surface of the plate, 

 extending beyond it. They are in rows ; and when the two plates 

 are in accurate and normal contact, the knobs fit into the sockets 

 of the lower plate (Plate YIII. fig. 4). The only place where 

 this knob-and-socket arrangement is deficient is quite at the end 

 of the edge of tlie plates, where the pit ruus in for a short dis- 

 tance. Each interradial plate has a series of knobs on its actinal 

 edge, and sockets on the abactinal edge. 



The ambulacral areas are occupied by two vertical series of 

 plates, separated by a zigzag vertical suture along the median line, 

 and they are flanked by the numerous small, pored plates of the 

 tentaculiferous series. Each large ambulacral plate has an actinal 

 and abactinal horizontal edge, an angular suture uniting it with 

 two neighbouring plates along the median line, and an opposite 

 end, to which are attached the small plates of the poriferous zone. 

 The abactinal edge of each of the large ambulacral plates is 

 covered with minute knobs of the reticulate tissue, and its actinal 

 edge has rows of sockets ; so that the lower or actinal edge of 

 one plate receives the knobs of the abactinal edge of the plate 

 below, an arrangement exactly the reverse of that which exists in 

 the interradial areas. The pits interfere but slightly with this 

 arrangement, as they are almost out of the way and close to the 

 outer angle. 



The vertical sutures of both areas being in relation to the large 

 pits of the median lines, are very different in their construction to 

 those just noticed, and must be considered in detail. 



First. The median zigzag suture of the interradial area, which 

 unites the broad interradial plates side by side. 



The edge of a plate in this position forms the sides of an angle ; 

 and, as before noticed, these large pits excavate its surface. Each 

 expanded pit is separated from its neighbour on the same plate- 

 edge by a projection, which is broad toward the outer surface of 

 the plate and narrower inwards. There is also a shelf-like pro- 



