362 ROBERT TRACY JACKSON ON ECHINI. 



adradial, being perpendicular to the surface, but on account of the thickness of the plates the 

 sides are far from parallel, being often almost wedge-shaped in some of the thick-plated species. 

 Interambulacral plates bear small secondary tubercles with corresponding small spines as in 

 the ambulacra. 



In the lower genera of the family the species are largely based on the number of interambula- 

 cral columns in an area; but in Melonechinus there is a large range in the number of ambulacral 

 columns, and this is considered a more important character, taking precedence over the number 

 of interambulacral columns, which is relegated to a secondary place as a species differential. 

 It is worth noting that the highest species, M. giganteus, has the highest number of interambula- 

 cral columns known in the family (Plate 60, fig. 3; Plate 61, fig. 8). 



The interambulacra of Melonechinus ventrally have two plates in the basicoronal row 

 (discussion, p. 66), and three plates in the second row (Plate 56, fig. 2). With rare exception 

 the fourth column originates in the third row, above which additional columns, if added, come 

 in with considerable regularity and with a perfectly definite system, as first shown by Jackson 

 and Jaggar (1896). This system is the same as in all this family, and is shown in a represen- 

 tative species in Plate 57, fig. 1 and text-fig. 246, p. 382 (p. 70). 



Mr. Agassiz (1881, pp. 78, 79, 95) considers that the numerous interambulacral plates of 

 Melonechinus and other genera of the Perischoechinoida are derived from the splitting up of 

 primary interambulacral plates, but I cannot agree with this view (pp. 28, 64, 378). Mr. Agassiz 

 (1881, p. 79) also considers that the large number of coronal plates in the Palaeechinidae is a 

 feature in which they " show most unmistakably their systematic affinity to the Crinoids." In 

 thus expressing himself he supports the view that he earlier published (1864, p. 16), in which he 

 considered the genera of the Perischoechinoida as "only synthethic and prophetic Crinoids." 

 These statements are in striking contrast to what Mr. Agassiz says in the Revision (1874, 

 p. 645), where he gives structural grounds for widely separating the Perischoechinoida and the 

 Crinoidea. I give reasons (p. 200) for rejecting any close affiliation of the Echini with the 

 crinoids. 



The peristome of Melonechinus is known in only one specimen (Plate 50, figs. 7, 8), but 

 in this there are many rows of ambulacral plates. In each radial area there are two relatively 

 larger primordial ambulacral plates adorally, and, passing aborally, the number of plates 

 increases to many plates in a row on the outer border of each ambulacral area of the 

 peristome. Three non-ambulacral plates occur interradially in probably each area. This is 

 the only case of a peristome known in the family (pp. 80, 378). 



The apical disc in Melonechinus is small measuring proportionately about 13 to 25 % 

 of the diameter of the test. It is interesting to note that in Melonechinus springeri which is a 

 primitive species in the genus, the apical disc is relatively large ; its diameter being about 23 % 

 of the diameter of the test; whereas in M . liratus and M. multiporus which are more specialized 



