PROTEROCIDARIS. 411 



The ambulacra are narrow, at the mid-zone composed of four columns which consist of 

 occluded and demi-plates in each half-area (Plate 67, fig. 5). In this figure the plates of the 

 two half-areas are mechanically separated. It is a view from the interior, and the occluded 

 and demi-plates, at least on this side of the test, are of about equal width, and pore-pairs lie 

 near the middle of each half-area, doubtless due to the fact that it is an internal view (com- 

 pare Plate 43, figs. 3, 4). About three ambulacral plates equal the height of an adambulacral, 

 and they bevel under the adambulacrals on the adradial suture (or over them, as seen in internal 

 view). DeKoninck says that in the type the ambulacra are not visible, but, in the flattening 

 of a form with interambulacral plates imbricating laterally, it is mechanically natural that this 

 area should be overridden and thus shut out from external view, as occurs in Perischodomus 

 biseriaUs (Plate 29, fig. 5). 



The • interambulacra are very broad, with numerous culunuis of plates in each area. 

 De Koninck says that the type appears to have 65 columns; this would be an average of 

 thirteen columns to an area, which is just the number shown here in Plate 67, fig. 4, after 

 Fraipont. In his description, Fraipont says that there are eleven or twelve columns of plates 

 at the middle [mid-zone]. In a section seen from the exterior (Plate 67, fig. 4) at the mid-zone 

 there are seen to be thirteen columns of plates. The thirteenth column originates on the left next 

 to the adradial column. The initial plate of this column is a pentagon, and instead of having 

 a heptagon on its ventral border as usual, the adambulacral plate X has attained an additional 

 compensating side, so that it is hexagonal instead of pentagonal as usual. The interambulacral 

 plates imbricate strongly dorsally and from the center laterally and over the ambulacrals. 

 Each plate bears a small primary tubercle with numerous secondary tubercles. The primary 

 spines are small, slender, and swollen at the base. De Koninck says that the primary spines 

 have a length of 10 mm. and a diameter at the base of 1 mm. The primary spines shown in 

 Fraipont's figures are smaller, apparently about 4 mm. long, but his specimen is much smaller 

 than the type. The secondary spines are more slender than the primaries, and measure about 

 2 mm. in length, judging from the figure (Plate 67, fig. 7). 



The interambulacra as seen from within are very interesting for comparison. In this 

 internal view (Plate 65, fig. 3, area I ; Plate 67, figs. 5, 6) the plates imbricate adorallj^ and toward 

 the center from the sides, largely covering up the median column of the area; also the ambul- 

 acral plates laterally bevel over the adambulacrals. All this is just the reverse of what is seen 

 on the exterior, as shown in different areas of Plate 65, fig. 3 (also compare text-figs. 32 to 35, 

 p. 75). Fraipont's fig. 2 on Plate 5 is inverted; that is, the dorsal portion is toward the bottom 

 of the plate. This gives the impression that the direction of imbrication is the reverse of 

 what it really is, but I have corrected the orientation in my copy of part of this figure (Plate 67, 

 fig. 5). In the specimen figured on Plate 65, fig. 3, there are twelve columns of plates in inter- 

 ambulacrum I, and in the view of a section from another specimen (Plate 67, fig. 5) there are 

 twelve columns. In one of Fraipont's figures (his Plate 5, fig. 1) there is a low, wide genital 



