428 PKOE. p. M. DFNCA]!^ ON THE STETJCTTJEE OF 



that the whole subject of the classification ought to he reinvestigated, 

 the morphology of the ambulacra being considered of primary im- 

 portance. It became easy, after the examination of weathered speci- 

 mens, to decide that whilst some recognized species oi Pseudocliadema 

 were evidently Oligopores and closely allied to the modern Diadema^ 

 others were Polypores, having sometimes as many as fi.ve or six pairs 

 of pores to an ambulacral plate. Again, some species are allied to 

 the recent forms by having the optic pores at the actinal margin of 

 the radial plate, and by having decided branchial cuts and even tags 

 arising from the cuts. Moreover the structure of CypJiosoma being 

 known to me, I could hardly consent to so close an alliance as 

 Dr. Wright suggested between it and Pseudodiadema. 



Pseudodiadema hemisj)hcericum is well drawn by Bone in Dr. Wright's 

 Monograph of the Erit. Foss. Echinodermata, pt. 1, 1855, plate viii. 

 The shape of the radial jDlates and the position of the optic pore at 

 the very margin of the plate are clearly indicated, and the drawing 

 of an ambulacrum (fig. ] d) shows the relation of three pairs of pores 

 to each tubercle-bearing plate. The exact relation of the pairs is 

 not shown ; for the specimen was so perfect that no sutures probably 

 were visible. But in the British Museum there is a specimen 

 (No. 23329) from Malton, named, as of old, Diadema pseudodiadema, 

 and the lines of the sutures may be seen here and there. 



In the great majority of the ambulacral plates there are three 

 pairs of pores. Each pair is in a primary plate, and the three pri- 

 maries have become fused, as it were, into a geometrical compound 

 plate (fig. 5). 



Fig. 5 (see p. 451). 



The first or aboral pair of pores of this compound plate has its 

 adoral pore in contact with the adoral suture of the low broad 

 primary plate which forms the first or apical portion of the com- 

 pound plate ; and this suture is directed to the median or vertical 

 suture in a course which is somewhat curved, the convexity being 

 adoral. 



The suture crosses the boss of the tubercle just abactinally to the 

 mamelon. 



The pair of pores which belongs to the middle plate of the 

 combination has its adoral pore in contact with a suture that unites 

 its adoral edge with the aboral edge of the third plate. The 

 direction of this suture is towards the median suture, and it has 

 a path from the interradial end of the poriferous zone to the 

 adoral pore just noticed, and thence with a curve directed apically, 



