METAZOA ECHINODERMA. 177 



pels them to be drawn out and to diminish in number 

 while at the same time the plates themselves become 

 more or less rhombic in form. 



The lines x, y, z in PI. 321, fig. 5, bisect eight rows, 

 and indicate by their narrowing angle the stringing-out 

 arrangement of the plates in this area (Jackson). This 

 reduction does not seem to be comparable to*the dying 

 out of parts or organs so characteristic of gerontic forms. 

 As Dr. Jackson aptly says, it may be compared to a flock 

 of sheep coming through a narrow pass. The small 

 number in the pass does not mean that the flock is les- 

 sening, but that no more can get through at once. If 

 this were a gerontic condition we should expect to find 

 the middle or latest formed rows disappearing first and 

 not the lateral or primary rows. This is the case in the 

 few gerontic specimens observed (see below). 



If now a graphic summary of our knowledge of the 

 development of the ambulacra of the Palaeozoic Echini 

 be given (PI. 322, A-G) it will show at a glance that the 

 primitive and fundamental simplicity of Bothriocidaris 

 (A) has given rise through progressive steps represented 

 by Rhoechinus (B), Palaeechinus (C), and Oligoporus (D 

 at ventral border, E at ambitus) to the complexity of 

 Melonites (F at ventral border, G at ambitus). Dotted 

 lines are drawn through the primary plates a, b in each, 

 and also through the secondary plates a\ b 1 . New plates 

 begin to appear between these in Oligoporus (E) and 

 probably constitute the rows c, d in Melonites near the 

 ventral border (PL 321, fig. 3), while at the ambitus ten 

 rows are found. The remarkably large and fine speci- 

 men of Melonites gigantcus Jackson (Pi. 323, photo- 

 graph) shows still greater specialization than Melonites 

 multiporus. There are twelve rows of ambulacral and 

 eleven of interambulacral plates. The interambulacral 

 area (PI. 323, fig. 2, at the right) is especially interesting, 

 since it shows a tendency toward specialization by reduc- 

 tion. The last formed row of plates (n) has died out 



