72 TH. MORTENSEN, (Schwed. Siidpolar-Exp. 



To the description of A. cavernosus given by Philippi and, especially, by De 

 Loriol (under the name gallegosensis) the following remarks may be added. 



The hinder part of the labrum does not reach beyond the first ambulacral plate. 

 The labrum carries small spines and pedicellariae. There are generally 5 well devel- 

 oped subanal tubefeet, the first of them occurring on the 6th — 7th ambulacral plate. 

 The peripetalous fasciole passes over the 15th — 16th plates of the posterior ambu- 

 lacra. Genital pores generally 3, sometimes only 2. The test is often unequally 

 developed, the right side projecting distinctly beyond the left. The petals are stated 

 by De Loriol to have »une tendance marquee a s'effiler a l'extremite»; they are 

 certainly slightly narrower at the lower end, where they join the fasciole, but they 

 are still distinctly petaloid in their whole length, in marked difference from Tripylus 

 excavatus. The latero-anal fasciole may be rather distinct in young specimens, and 

 also sometimes in larger specimens. 



The spines are rather coarse, as in all the Abatus-s-peci&s. — The spicules are 

 irregular spinous or ramose rods, well developed in the tubefeet of the frontal am- 

 bulacrum, very feebly developed in those round the mouth. The sucking disk is 

 well developed in the tubefeet of the frontal ambulacrum, the edge of the disk not 

 lobate; the rosette plates are broad and well developed, though not reaching the 

 edge of the disk. — The sphaeridiae are generally much grooved. 



The pedicellariae occur in three forms, viz. globiferous, rostrate and tridentate. 

 The globiferous pedicellariae (PI. XIX Figs. 28 — 29, 33), which do not occur on all 

 the specimens, are rather large, the valves enclosed by a thick, brown, evidently 

 glandular skin. The valves have, as in Sckizaster, an interior gland cavity, opening 

 at the point of the valve. The opening is surrounded by 2—3 teeth on either side, 

 sometimes, however, there may be only one tooth on the one side, 2 — 3 on the 

 other. The sides of the basal part rather narrow; the blade is rather strongly 

 curved. There is no neck; the stalk is simple, without distinct thickening at either 

 end. — The rostrate pedicellariae (PI. XIX Figs. 30, 38, 45) have a rather deep blade 

 with more or less serrate edges; the outer end is narrowed. The valves are generally 

 not much curved. The neck is well developed, the stalk simple. — The tridentate 

 pedicellaria; (PL XIX Figs. 37, 46, 50 — 51) occur with two or three valves indifferently; 

 the larger samples, however, are all three-valved. They reach a size of about 1,3 

 mm. length of head. The blade is simply leaf-shaped; in the larger ones there may 

 be some meshwork on the outer side of the blade. In the smaller ones the edges 

 join in their whole length, in the larger they are apart for some distance below, 

 up to half the length of the blade, this part being accordingly narrowed. The edge 

 is finely serrate in the outer, joining part, with more or fewer coarser serrations in 

 the lower part. The valves of the 2-valved specimens are somewhat compressed, 

 otherwise not differing essentially from those of the 3-valved; the apophysis is dis- 



