244 DISCOVERY REPORTS 



edge is gently convex. The pores have a well-developed tentacle scale (Fig. 3 a). Dorsal 

 arm plates in the basal part of arms fairly regularly divided into two parts (Fig. 3 b). 

 Arm spines two on the joints within the disk; beyond the disk there are three spines. 

 In the proximal part of the arms the lowermost and the uppermost spine enlarged, 

 somewhat flattened and sometimes a little curved. In the distal part the two uppermost 

 spines transformed into hooks (Fig. 3 c). The scales of the disk as usual with a glassy, 

 concentrically striated margin. Rather large mainly three-radiate spicules in the 

 stomach wall (Fig. 3 d). About the colour in life there is no information; the preserved 

 specimens are white. 



It is evident that this species is nearly related to the South African O. dentatus, 

 Lyman, and it might, indeed, seem questionable whether the small differences to be 

 observed (cf. fig. 3 with the figures of O. dentatus given in my paper on the South 

 African Echinoderms, pp. 310-11, figs. 32-4) are really of suflnicient value for specific 

 distinction. Of this there cannot, however, be the slightest doubt, for I find that the 

 South American form is viviparous and— partly— hermaphrodite, whereas O. dentatus, 

 as well as the closely related O. purpiireus, are not viviparous and have separate sexes. 

 Thus it is certain that the South American form is a quite distinct species, the more 

 interesting as this is the first viviparous species of the genus Ophioscolex to be known. 



The statement that this species is "partly" hermaphrodite means that some speci- 

 mens are not hermaphrodite, but purely males or females. Thus the species is a 

 facultative hermaphrodite. This is of interest in connection with the fact that the two 

 known viviparous species of Ophiomyxa have separate sexes. It would seem to be a 

 beginning of hermaphroditism that we are witnessing in this primitive Ophiurid, herm- 

 aphroditism in Ophiurids being apparently a specialization acquired in connection 



with viviparity. 



As regards the arrangement of the gonads in the hermaphrodite specimens it seems 

 to be the more usual condition that the male gonads are situated at the adradial side, the 

 female gonads along the interradial side of the bursae, but there is no regularity either in 

 the arrangement or the number of the gonads. I have found only one or two young ones 

 at a time in the bursae. 



No species of Ophioscolex was known till now from the Antarctic or sub-Antarctic 

 region, the "Ophioscolex Coppingeri" of Bell being nothing but Ophiomyxa vivipara 

 (cf. above, p. 242). It is thus a rather surprising fact that no less than two species of the 

 genus, Ophioscolex nutrix and the following species, O. marionis, have been discovered 

 by the ships of the Discovery Committee in the sub-Antarctic region. 



Ophioscolex marionis, n.sp. 



St. 1563. 7. iv. 35. 46° 48' S, 37° 49' E, off Marion Island, 113-99 m- i specimen. 



Diameter of disk 5-5 mm. Arms all broken close to the disk; they are 1-5 mm. broad, 

 somewhat flattened. Disk covered by thick skin, through which some scattered short 

 spines protrude all over the dorsal side, a single one also here and there on the ventral 

 interradii. Apparently no radial shields. The buccal shields are small, rounded tn- 



