BOTHRIOCIDARIS. 241 



plate in each area excepting 2, where it is wanting, doubtless from erosion; in the third and 

 succeeding plates to the mid-zone there may be one, two, or three tubercles varying with flu- 

 areas, the highest number occurring in area 4. Dorsally, for one or more plates in an area, 

 there is only one central tubercle to a plate like that in the basicoronal plates. In the youngest 

 interambulacral plates in contact with the oculars, there are no tubercles, a common charu<-t.- r 

 of plates next the apical disc. The characters of the tubercles on the intcniinlnilacra <x press 

 radial variation, ventral development, and dorsal localized stages. The distribution of tubercles 

 on interambulacral plates is striking from the difference in different areas. Of course some 

 tubercles may have been worn off, but those shown are probably essentially the numlx-r that 

 existed when the animal was alive. Dorsally (Plate 1, fig. 2) the interambulacra abut agum-t 

 two oculars, a condition otherwise known only in spatangoids (text-fig. 175, p. 149), and in 

 aberrant variants of regular Echini (Arbacia, Plate 4, figs. 11, 12; Eucidaris, text-fig. 185, p. 

 167). The single column of interambulacral plates is considered a primitive character, because 

 in the young of all later Echini there is a single plate in the basicoronal row representing a 

 single column as a stage in development (compare text-figs. 22-31, p. 70). There is no resorp- 

 tion of the base of the corona, a character of the young in all later Echini and of certain adult 

 types of later Echini (p. 70). 



The peristome of this ancient type (Plate 1, fig. 1; text-fig. 40, p. 80) consists of t\\. 

 rows of ambulacral plates. The first row represents the primordial ambulacral plates, and 

 examining them, we find that the la, Ha, III6, IVa, V6 are larger plates, and the 16, 116, 1 1 In. 

 IV6, Va are smaller. This is in accordance with the law worked out by Loven in Recent 

 Echini (Plate 2, fig. 1), and it is most striking that it was already at work in those ancient 

 times. By means of it I oriented the specimen as shown in the figures. I could distinpui-h 

 no tubercles on this first row of plates. The second row of the peristome surrounds the fir>t 

 and aborally is in contact with the basicoronal interambulacral and ambulacral plates. The 

 plates of this second row are distinguished from the coronal plates only by their position and 

 interambulacral relation, 1 but the same thing is true of the young of some Recent types 

 (Plate 2, fig. 6). The second row was probably derived by flowing down from the corona 

 (pp. 79, 86). On some plates of this second row one or two tubercles were observed. 



The oculars are very large relatively, and are unique in that they meet in a continuous 

 ring. Ocular III measures 1.4 mm. in height, but the others somewhat loss. All have three 

 tubercles excepting II, which has only two. Ocular III has in addition two tiny sinuous ele- 

 vations which appear to be part of the plate. Jaekel thought these indicated a madreporite; 

 this may be correct, but it is doubtful. The small plates in the dorsal interradial angles of 

 the oculars I consider as genitals (as discussed, p. 88). Genitals 1, 2 and 4, 3 are in contact 



' By an error in making my drawing in Berlin, from which original Mr. Blake drew Plate 1, fig. 1, the initial plate of 

 area 1 rests against the third plate of 16, instead of the second plate as it evidently should. 



