1072 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
bb. Disk not so high ; form nearly pentagonal; marginal plates conspicuous. Erect conical tubercles 
on both surfaces Nidorellia 
aa. Marginal plates hidden or inconspicuous; not visibly defining the contour of body. The abactinal 
plates are not superficially distinguishable in the adult. Form thick. 
b. A pair of large marginal plates at the end of the ray. Form substellate to stellate. Test covered 
with globose or acorn-shaped tubercles. Papulae evenly distributed, not in definite areas. 
Asterodiscus 
bb. Form pentagonal, biscuit-like; no large plates at end of ray. Papulae distributed in large, 
definite areas _ ..Culcita 
Genus PENTACEROS Schulze. 
Pentaceros Schulze, Betrachtung der veisteinerten Seesterne u. ihrer Theile, Warshau u. Dresden, 1760, p. 50. Gray, Ann. 
N. H., ser. 1, vol. VI, 1840, p. 276. 
Goniaster (pars) Agassiz, M<5m. Soc. Sci. Nat. Neuch&tel, t. I, 1835, p. 191. 
Oreaster Muller and Troschel, System der Asteriden, 1842, p. 44. Bell, Proc. Zool. Soc. London, 1884, p. 57. 
The specimens of Pentaceros from the Hawaiian Islands have given considerable trouble, as might 
naturally be expected. Relying chiefly upon Prof. F. Jeffrey Bell’s useful and valuable revision of 
the genus, “ The Species of Oreaster” (Proc. Zool. Soc. Lond., 1884, p. 57), I have separated the form 
as a distinct species, whose affinities appear to be with the Pentaceros orientalis section._ Among the 
species of this group it most resembles orientalis and troscheli. On the other hand, there are certain 
points of similarity with P. occidentalis. One of the difficulties lies in the fact that some specimens 
have the adambulacral armature disposed in 3 longitudinal series (triplacanthid), while others have 
but 2 series (diplacanthid). Just this state of affairs exists in P. occidentalis (from Mazatlan, Mexico). 
Bell places occidentalis in the diplacanthid section, but implies that occasionally a third furrow series 
is present. But hawaiiensis differs from occidentalis in a most important respect — that is, in its growth 
stages. In occidentalis, as pointed out by Bell and as exemplified by specimens at hand, the young 
are more spiny than the adult, while in hawaiiensis exactly the reverse is the case. In this species, if 
any spines are present in the young they are inconspicuous, and are confined to the median-radial line. 
Hawaiiensis, then, is probably not derived from the American form, but from the Asiatic group. In 
the following diagnosis a description of the type is given, but under each category of characters the 
more important variations are indicated. 
Pentaceros hawaiiensis, new species. 
PI. XXXII, figs. 1, 2, 3; pi. XXXIII, fig. 1; pi. XXXIV, fig. 3. 
Rays 5. R = 125 mm.; r = 50 mm. R=2.5 r. Dimensions measured on ventral side. (In 2 
larger specimens the rays are relatively shorter, R equaling 160 mm. and r about 78. In still another 
specimen, the largest of all, R = 188 mm. , and R = 2. 35 r. ) 
Disk large, elevated, and regularly convex ; about 55-58 mm. high, measured from actinal plane. 
This dimension varies considerably according to the inflation of disk, several specimens being much 
depressed. Rays well produced, rather broad at base. Actinal area plane on rays, sloping upward 
toward actinostome on disk. 
The whole abactinal surface is marked off by trabeculae into large triangular, or on the ray often 
rectangular papular areas. On disk they are fairly regular, but on ray are irregular. At each angle 
is a large conical tubercle, there being a definite medio-radial series extending to tip of ray, and on 
either side 2 parallel longitudinal series, the first extending about half the length of rays, and the 
second not so far as the first. In interradial angle formed by the outer series of adjacent rays are 3 to 
6 tubercles. The pentagon marked off by the conspicuously enlarged primary radial tubercles (apical 
area) is also divided into triangles, often very regularly into 10, but again irregularly. There is a 
variable number of smaller tubercles in this area (7 in type) . All tubercle's decrease in size toward tip 
of ray and margin of disk. Granulation consists of variously sized polygonal granules crowded 
together, often increasing in size as they ascend the tubercles, the spinous tip of the latter being always 
free from granules and the line of demarcation well defined. Granulation is finer than in either 
P. reticulatus or P. occidentalis, the only species with which I have been able to make direct compari- 
sons. Numerous small slit-like pedicellarise, flush with the general surface, are thickly scattered over 
