60 THE ECHINODERMS OF TORRES STRAIT. 
actinal intermediate plates lacking.” I believe Koehler is mistaken in describing offreti 
as having such a series, for on comparing my specimen of ocellata with his description of 
offreti, I was at first led to believe such a series was present in the Murray Island species, 
but later I was struck by the close correspondence between this series and the adambulacral 
series and a little examination and dissection under the microscope soon showed that what 
seems like a series of actinolateral plates is really composed of the outer ends of the adam- 
bulacral plates. The supposed inner series of small actinolaterals is, I think, non-existent. 
A comparison of Koehler’s plate xv1, figure 3, with Fisher’s plate 106, figure 4, shows the 
series of supposedly separate plates as plainly in one as in the other. 
The only known specimen of glyptodisca was taken by the Albatross in Buton Strait, 
Celebes, in 24 fathoms, on a bottom of sand and broken shells. 
Ferdina cancellata. 
Scytaster cancellatus Grube. 1857. Arch. f. Naturg., 23, pt. 1, p. 340.—Nova Acta Acad. Leop., 27, p. 9; pl. ii," 
figs. 3, 3a. 
Ferdina cancellata Sladen. 1889. Challenger Ast., p. 780. 
The locality whence came Grube’s type is not known, but Liitken (1871) records a 
specimen in the Copenhagen Museum from the Fiji Islands. The holotype was a small 
sea-star only 2 inches across, and it is quite possible some of its peculiarities are due to youth. 
Ferdina ocellata? sp. nov. 
(Plate 6, Figure 5; Plate 31, Figures 1 and 2.) 
R =42 mm., r=13 mm.; br=14 mm.; R=3.2 rand 3 br. Disk and rays very flat 
and rather rigid; interbrachial arcs subacute. Abactinal surface covered by a pavement 
of plates, ranging in diameter from 0.5 to 2 mm., and in convexity from nearly flat 
to tubercular; these plates show no regularity of arrangement, but there is an indef- 
inite series of larger plates along the median line on the basal half of each ray and 
even less definite transverse series of medium-sized plates across the rays, 2 to 5 of 
these on each ray; the middle and distal part of each ray has 2 to 5 tubercle-like plates. 
Superomarginal plates 13 to 15, in an irregular series, made up of alternating larger, swollen, 
and smaller, flatter plates; the alternation is not perfect and the smaller plates are of 
quite diverse sizes and form. As a rule, the transverse series of medium-sized plates on 
abactinal surface of ray lie opposite small superomarginals; the first superomarginal of 
each series is very large, but nearly flat; it lies close against its fellow of the adjoining 
series of the next ray, the two together forming a conspicuous area on the interradial 
margin of disk, 5 mm. wide by 3.5 mm. along the interradial suture. Terminal plate 
rather small, diamond-shaped, somewhat swollen, bare and smooth, with a single terminal 
tubercle. Madreporic body conspicuous, elevated, about 3 mm. across, nearer center of 
disk than margin. Anus central, conspicuous because of the group of 10 to 12 coarse gran- 
ules by which it is surrounded. Whole abactinal surface, except madreporite, tubercular 
abactinal plates, swollen superomarginals and terminal plate, covered by a very close and 
fine granulation, 200 to 400 to a square millimeter, which is coarsest around the papule 
but is coarser at the center of each plate than it is near the margin; the tubercular abactinal 
and swollen marginal plates have their margins concealed by this fine granulation, but the 
larger part of their surface is bare and the boundary line between bare and granulated 
portion is very sharply drawn. Papule large, isolated, numerous, with no regularity of 
arrangement whatever, each surrounded by a circle of coarse granules. 
1 Tn the text and in the explanation of plates, plate ii is called plate i and vice versa. This has led to erroneous 
references to these figures. 
? Ocellatus =having little eyes, in reference to the eye-like red spots on the superomarginal plates. 
