60 BULLETIN 76, UNITED STATES NATIONAL MUSEUM 



In prepared specimens : The carinal marginal and actinal plates are regularly four- 

 lobed. The carinals are distinguishable from the dorsolaterals mainly by reason of 

 regularity, the latter (as seen in the figures) being irregularly arranged and not 

 symmetrical in form. New plates are continually being added to the dorsolat- 

 erals. In very small specimens there is a single row. I have none small enough to 

 demonstrate their absence. A specimen of true typicus with R only S mm. has a 

 single series of three-lobed dorsolaterals. In big specimens the marginal plates are 

 lateral and bend upward a little at the interbrachium. The superomarginals form a 

 sharp line of demarcation between the irregular dorsolaterals and the regular trans- 

 verse ranks of actinals-plus-marginals which are characteristic of very large specimens 

 of this genus. The plates strongly overlap as shown in Plate 28, Figures 1, la, and 1 J. 

 There are 4 or 5 actinals in each transverse series at the base of the ray, increasing 

 to 8 to 11 at about the middle, and then decreasing again near the tip. The actinal 

 plates do not, therefore, stand in regular longitudinal series on account of this highly 

 peculiar arrangement. The actinals are rather thin around the edges. In a great 

 mass of small plates it is sometimes difficult to be sure of the marginal series, and it 

 is essential to be certain. The superomarginals, in addition to standing at the upper 

 end of the transverse rows, have both their transversely oriented lobes overlapping 

 those of their neighbors. 



The plates usually bear a single spinelet only, except the carinals and the larger 

 dorsolaterals, which commonly carry two or three. Over the disk the spinelets are 

 rather thickly placed, most of the larger plates having two or three, and I find some 

 specimens in which the proximal superomarginals have one or two smaller accessory 

 spinelets. The spinelets, averaging 0.45 to 0.6 mm. long, end in three to several sharp 

 divergent points, except the smaller secondaries which may have but 1. (PI. 27, 

 figs. 6, 6a, 66.) 



The adambulacral spines, two to a plate in an oblique transverse row, form 

 two regular series on the edge of the furrow. The inner spine is slender, slightly 

 tapered and pointed; the outer is a little stouter and longer, sometimes slightly tapered 

 and pointed, sometimes somewhat clavate and blunt. Occasionally there are three 

 spines to a plate either in a regular transverse series, or two are rather obviously 

 smaller, furrow, spines and the third a heavier subambulacral. The first pair of 

 adambulacral plates do not meet in the interradial line back of the mouth plates. 

 A very few small blunt ovate straight pedicellariae (0.25 to 0.27 mm. long and about 

 0.18 broad) occur here and there along the inside of the furrow. 



The mouth plates, the form and armature of which are best appreciated from the 

 figure, usually have a longitudinal series of four fairly long spines and at the inner end 

 of the plate, one, two, or three smaller actinostomial teeth, with, occasionally, a tiny 

 ovate straight pedicellaria, 0.25 mm. long. 



The papulae, usually one to an area, are small and in alcoholic specimens 

 rather hard to distinguish. 



The crossed pedicellariae are of two sorts and are of definite diagnostic value. 

 The larger kind (pi. 26. figs. 2, 2a-2e) are very abundant on the actinal and lateral 

 surfaces, and sometimes in big examples (station 4792) on the dorsolateral regions also. 

 In small specimens they are restricted to the lateral and actinal regions, the latter, of 

 course, very narrow. The smaller sort are found on the abactinal surface very 



