PHOLIDOCIDARIS. 435 



adjacent fellows when in place. They also show the strongly curved outlines which the suture 

 lines of the several adjacent plates possessed. The dorsal adradial plates bear each a large 

 eccentric perforate primary tubercle with scrobicule, and, in addition, small secondary tubercles 

 (Plate 73, figs. 3, 4; Plate 74, figs. 1-5; Plate 75, figs. 1-3). The plates of the median columns 

 dorsally are strongly rounded on the suture lines, and bear secondary tubercles only (Plate 73, 

 figs. 3, 4; Plate 75, figs. 1-3). The ventral part of the test is known from Meek and Worthen's 

 figure (my Plate 75, fig. 4), also from the interior as seen in part in Plate 74, figs. 1 and 2, 

 and from the exterior as seen in Plate 73, fig. 6.^ This last specimen is somewhat confused 

 and difficult to study, yet it shows important features. The primordial interambulacral plate 

 is in place in the basicoronal row as seen in area C. There are two plates in the second row and 

 three in the third row; thus the third column originates in the third row as usual in the Palaeozoic 

 species in which the base of the corona has not been resorbed in the advance of the peristome 

 (p. 66; text-fig. 30, p. 70). There are three rows of three plates before the fourth column orig- 

 inates in the sixth row. This is very late for the fourth column to develop, the latest of any 

 case figured in this memoir, and it is very likely exceptional for the species, but on this point 

 there are no other specimens as yet known for comparison (p. 439). The fifth column origi- 

 nates in the seventh row (the next row after the introduction of the fourth) and the sixth 

 column originates in the eighth row. These plates all imbricate aborally and from the center 

 laterally and over the ambulacra on the adradial sutures. All interambulacral plates on the 

 ventral side are of about the same size as far as this figure shows (Plate 73, fig. 6), the adradial 

 plates not being strikingly different from plates of the median columns as they are on the 

 dorsal side (Plate 73, figs. 4, 5). The plates of this specimen are much worn, but the adradial 

 plates, and also the median plates bear a large perforate eccentric primary tubercle, doubtless 

 with, in addition, secondary tubercles, but these latter, if they existed, are worn off. There are 

 therefore primary tubercles on all interambulacral plates ventrally, whereas they exist only on 

 adradial interambulacral plates dorsally. The primary spines are rather stout, enlarged at 

 the base, terete. Meek and Worthen say that the larger primary spines attain a length of about 

 an inch and a thickness of 0.1 inch at the base, which is swollen (Plate 75, figs. 4, 5). The 

 largest that I have seen measure about 12 mm. in length (Plate 73, fig. 5). Secondary spines 

 are small, filiform, expanded at the base, and measure about 3 to 4 mm. in length (Plate 75, 

 fig. 4; Plate 73, fig. 5). 



The apical disc is only partially known, but it is evidently small in proportion to the 

 diameter of the test. A small plate dorsal to ambulacrum B, in Plate 73, fig. 3, has two pores, 

 and may be an ocular (p. 419), but other oculars, if present in any of the specimens, are not 

 recognized. The genitals are high and wide, with many genital pores, up to ten counted in one 



1 This figure is of the specimen which I earher described as P. meeki, but which is now referred to irregularis as a 

 synonym, as discussed on p. 438. 



