3- Gleba: proboscis free, projecting; fins with a continuous margin, without ventral lobe; on 

 the latero-dorsal margin of the fins are some ver\- complicated indentations ; the three 

 sets of muscles in the fins are only distinct at the margrins, while in the centre only one 

 of them, running dorso-ventrally, (so the one in which the bundles are all parallel to each 

 other) may be distinguished. — Shell ovoid, flattened, with hardly any cavity; its dorsal 

 margin is thickened, the ventral one is thin ; sometimes covered with a few tubercles. 



We shall now proceed to examine successively the two families, which have been collected 



by the Siboga. 



Cymbulia Peron et Lesueur. 



iSio. Cymbulia Peron et Lesueur, Histoire de la famille dcs Mollusques Pteropodes, Ann. 

 Mus. d'Hist. Nat. Paris, vol. XV, p. 66. 



For the characters of the genus, see above. 



The right explanation of the organisation has been given by Pelseneer ^). The elongation 

 of the dorso-ventral axis of the shell has led most authors to believe that the pointed extremit\- 

 was anterior and the truncated end posterior. An examination of the animal, however, without 

 its shell, soon shows that this opinion is wrong, and removes at the same time the difficulty 

 in comprehending Pelseneer's arguments, when these are heard for the first time. The 

 only real difference between Cavoliniidae and Cymbuliidae is the presence of an internal 

 cartilaginous shell in the latter. Both these families have descended from Limacinidae; the 

 Cymbuliidae are more specialised with regard to pelagic life. — The figure of Boas ■) (though 

 this author did not recognize the true phylogenetic affinity of the Cymbuliidae) may serve to 

 make Pelseneer's opinion better understood. 



I shall not enlarge upon the various "species" referred to the genus Cyiiidiilia, as they 

 have already been discussed by Pelseneer. In some cases figures and even description are 

 entirely wanting, as in Cymbjilia odiusa Lesueur '*) (not mentioned by Pelseneer). Up to this 

 time it is only one species we know fairly well. But even this form, Cymdiilia pei'oni, has 

 never been sufficiently figured. I have had occasion to stud)- some specimens from the 

 Mediterranean. After careful examination ot the shell, I observed that the rows of spines were 

 not quite similar in the different specimens. The large forms (62 — 43 mm.) one of which is 

 figured on PL III, figs. 82 — 84, showed on the lateral side three parallel rows (fig. 83, «, b, c), 

 while other specimens (42 — 26 mm.) possessed only one of these rows (fig. 86, a). In addition 

 to this, the spines at the aperture which are of unequal size in the large specimens, especially 

 at the right side where they are very strongly developed, are almost uniform in these small 

 individuals. Are the latter young forms which have not reached a definite size ? Or do they 

 represent perhaps a new species, not hitherto observed" I can scarcely admit this last opinion, 

 as the difference, above noted, among the specimens has never been recorded by any author 



i) Challenger Report, LXV, p. gC — 97. 



2) Spolia Atlantica, p. 25, fig. G. 



3) In DE Bi.AiNVlLLE, Manuel de Conchyliologie, p. 655. 



