PHOLIDOCIDARIS. 439 



ambulacral plate is in the basicoronal row in area C, and this specimen was one of the first of 

 Palaeozoic Echini in which this structure was discovered. I correlated it with the single plate 

 at the ventral border in Bothriocidaris and the j'oung of Recent Echini (Jackson, 1896, pp.212, 

 230). In the second row there are two plates as a second stage in development, and in the third 

 row three plates. This is all as usual in the development of all Palaeozoic Echini in which the 

 primordial interambulacral plate is retained (Plate 25, fig. 1), excepting Bothriocidaris, in which 

 there is only a single column of plates in an area (Plate 1, fig. 1). In this area of this specimen 

 of Pholidocidaris there are also three plates in the fourth and fifth rows (Plate 73, fig. 6). In the 

 sixth row there are four plates, column 4 originating on the right of the center as usual. This 

 is, however, a very unusual case in that column 4 originates so late. It ordinarily originates 

 in the next row after the introduction of column 3, so that there is typically in Palaeozoic Echini 

 only one row ventrally with three plates (text-figs. 25, 30, p. 70). In Palaeechinus quadriserialis 

 in area C (Plate 30, fig. 3), and in Lovenechinus septies, also in area C (Plate 45, fig. 1), there are 

 two rows of three plates before the introduction of column 4. Again, in a specimen of Melon- 

 echinus multiporus, figured by Jackson and Jaggar (1896, Plate 3, fig. 12; Plate 4, fig. 18) there 

 are at least four rows of three plates before the introduction of the fourth column (p. 377). 

 These are all rare and exceptional variations of retarded development, and with Pholidocidaris 

 irregularis are the only cases seen in which the fourth column (when existent) did not originate 

 in the next row after the third (p. 349). In Pholidocidaris irregularis (Plate 73, fig. 6) the fifth 

 column originates in the seventh row, immediately after the fourth, and the sixth column origi- 

 nates in the eighth row, immediately after the fifth, thus, as regards these two, showing an 

 accelerated development. The plates of interambulacrum A are so confused that they do not 

 show much structure. Fragments of primary spines occur scattered over the test. 



A fragment of a large crushed test from Montgomery County, Indiana, in Mr. Braun's 

 collection, is very instructive. This specimen (Plate 74, fig. 1) is a portion of a test extending 

 from the apical disc nearly or quite to the mid-zone. The dorsal plates are seen from the 

 exterior and are not shaded. Part of the dorsal plates are wanting, so that ventral plates are 

 seen from the interior, and are shaded to differentiate them. From this peculiar condition 

 of preservation the ventral plates can be directly compared with the dorsal plates of the same 

 specimen. The ventral ambulacral plates are very large, from 5 to 7 mm. in width, with pore- 

 pairs about in the middle of each plate, as seen well in area D. In contradistinction, the 

 dorsal ambulacral plates are small, about 2 to 3 mm. in width. This extraordinary difference 

 in the ventral and dorsal ambulacral plates (p. 434), is shown more strikingly in this than in 

 any other known specimen. The adradial dorsal interambulacral plates are wide and high as 

 usual. Dorsally, there are two genital plates which are wide and rounded in outline, each with 

 many pores. A fragment of a lantern shows in part a brace, two epiphyses, and part of a 

 pyramid. 



