INTRODUCTION 207 



Omitting the seven species the sexual character of which is unknown, we find thus 

 that nine species of viviparous Ophiurids have separate sexes. Two or three appear to 

 be parthenogenetic, and thirty-six species are hermaphrodites, one of them, Ophioscolex 

 nutrix, a facuhative hermaphrodite, some specimens having separate sexes. 



The overwhelming majority of the viviparous Ophiurids thus are hermaphrodites. 

 Since not a single non-viviparous Ophiurid is known to be hermaphrodite, there must 

 be some connection between viviparity and hermaphroditism. One might suggest the 

 reason for the hermaphroditism of the viviparous forms to be to facilitate fertilization ; 

 but the fact that several of these species are protandric hermaphrodites and others 

 apparently parthenogenetic rather tells against such a suggestion. The fact that the 

 species with separate sexes are mainly found among the more primitive forms, Ophio- 

 myxa, Ophiacantha, may indicate that hermaphroditism represents a condition acquired 

 by the more specialized forms ; but since there are also forms with separate sexes among 

 the morphologically highest types, e.g. Ophiozonella and Ophiurolepis, this reasoning 

 loses its weight. Indeed, the whole matter seems inexplicable from the facts at present 

 available. 



Ophioceres incipiens is a rather intricate case. It seems fairly certain that it starts as a 

 male, changing then to female, returning again to the male condition and finally to a 

 pure female condition. 



The very interesting fact that a sort of copulation takes place in the viviparous 

 Astrochlamys bnineiis might also represent an eff^ort to facilitate fertilization; but the 

 three other Ophiurids in which a similar copulation takes place, Ophiodaphne materna, 

 Koehler, Ophiosphaera insignis, Brock, and Amphilycus androphonis , Mortensen, are not 

 viviparous (cf. my paper quoted above. Biological observations on Ophiurids, pp. 178-88), 

 so the above suggestion does not apply equally to all four cases of copulation. 



It appears that there is a tendency towards an intra-ovarial development in the 

 viviparous Antarctic Ophiurids. The only case hitherto known was Ophionotus hexactis ; 

 I have now found it to occur likewise in Amphiura microplax and monorima, and almost 

 certainly also in A. Lymani and Belgicae, Ophiomages cristatus, and Amphiophiura 

 Rozvetti. The most remarkable of these cases is that of Ophionotus hexactis, in which the 

 embryos even pass through a stage as a sort of " pelagic " larva within the ovary (cf. my 

 Studies of the development and larval forms of Echinoderms, 1921, p. 179, pi. xxxii). 



I cannot suggest any reasonable explanation of this tendency to give up using the 

 bursae as a marsupium and to let the embryos develop within the ovaries themselves. 

 It would seem that fertilization must be less easily practicable within the ovaries than 

 within the bursae. On the whole, there are a number of perplexing questions in con- 

 nection with this matter: why should there be such a high percentage of viviparous 

 forms in the Antarctic-sub-Antarctic region, when in the Arctic-sub-Arctic region, 

 with corresponding low temperatures, there are relatively much fewer viviparous forms ; 

 why that pronounced tendency to hermaphroditism among the viviparous forms, and 

 why the tendency to intra-ovarial development? Perhaps the study of other animal 

 groups in the same region may lead to the solution of these problems. 



