254 BULLETIN" 100, UNITED STATES NATIONAL MUSEUM. 



contact and are not separated throughout the ray, as the heading 

 of the synopsis would lead one to believe. 



In the specimen from station 5520 there is some irregularity in 

 the form of the plates, after being reduced to a single series on the 

 ray. At first hexagonal plates alternate with squarish ones, and then, 

 the latter die out and llie hexagonal plates become lozenge-shaped, 

 and are interrupted by the superomarginals meeting medially be- 

 tween them. The central granules of the abactinal plates are less 

 globular than in Sladen's figures and description. Eather they are 

 polj'^gonal, subtruncate, with a slight eminence in the center like a 

 pimple or mamelon. This specimen has usually 10 or 11 compressed 

 furrow spinelets, less often 9. Sladen gives 10, but his figures show 

 12 and 13. 



The specimen from station 5367 has slightly shorter rays and the 

 series of abactinal plates reaches the terminal plate. It agrees with 

 Sladen's variety hreviradiatus in these respects and in having a 

 pair of enlarged pedicellarialike granules adjacent to the actinal 

 intermediate pedicellariae. but diifers in having only a few (fre- 

 quently only 2) depressed subglobular granules on the central part 

 of the tabulum of radial abactinal plates. There are two kinds of 

 abactinal pedicellariae. One is the usual slender, entrenched forceps 

 described by Sladen. These have compressed toothed jaws, and 

 when the pair is closed they have a broadly elliptical form viewed 

 from the side. The other form is not common. It consists of 2 

 slightly modified elongate curved granules. In this specimen the 

 compressed furrow spines are commonly 9. 



The intestinal coecum is very large and consists of a central irregu- 

 lar pentagonal sac about 5 mm. wide, from which 5 slender coeca 

 extend interradially to the margin, each being divided at the middle 

 into 2 distal portions. The stomach is small and the hepatic coeca 

 extend about half the length of ray. The gonads are in 10 oi- more 

 independent tufts, which extend along the ray about as far as the 

 first 6 superomarginals. This condition recalls that in several Astro- 

 pectinidae, as, for example, Dipsacasfer. There are 2 Polian vesicles 

 in the hydropore interradius and 1 in each of the others; ampullae 

 i-trongly double; tube feet with strong sucking disk. Interbrachial 

 septum membranous. No superambulacral ossicles, or rudiments. 



The form of the abactinal skeleton as seen from the coelomic side 

 is very curious. The plates are bound together b}?^ slender, irregular 

 rods, much as in Mediaster^ but the rods are often once branched and 

 have tiny spinelike processes growing downward, which give to tho 

 coelomic surface of the abactinal integument a spinulose appearance, 

 suggesting remotely the abactinal surface of some species of 11 en^ 

 rif'/a having few sharp spinelets. The plates of the radial series are 

 -lightly 6-lobed. with 6 to 8 of the rods connecting it with neigh- 



