48 



o-enus Cavolinia is placed at the end of a series of forms, provided with external calcareous 

 symmetrical shell, and is the most specialised genus in this direction. 



It seems quite superfluous to repeat all that has been already stated by the different authors, 

 especially van Bf.xeden ^), Souleyet "), Gegenbaur ■'), Boas '^), and Pelseneer ^), who studied the 

 anatoni)- of the genus in question. I shall onh' restrict myself to a few remarks, partly not 

 agreeing with Pelseneer's publications, partly completing them. 



There are several differences between the two subgenera Diacria and Cavolinia (s. sfr.J. 

 Remarkable in Diao-ia is the separation of the liver into two lobes, having each its own duct. 

 This arrangement does not occur in any other group of the Thecosomata. The flexure of the 

 intestine is ventral in all Cavoliniac\ it seems unnecessary to state this again, had not Vogt 

 and Yung ") been mistaken in this respect. 



It must have been a slip of the pen, which made Pelsexeer pretend that the heart in 

 Cavolinia is on the right side. It is situated on the left. 



The genital gland in Diacria occupies the whole posterior portion of the visceral mass, 

 and is disposed transversely. In Cavolinia fs. sir.), however it is asymmetrical, situated almost 

 quite on the left side, embracing the strong retractor muscle, while the lamellae of the gland 

 exhibit the usual direction as in all Cavoliniidae. Transverse sections through this organ show, 

 that the spermatozoids arise from inner folds of the membrane which envelops the gland. I 

 could not notice any regularity as to the disposition of the ova in the gland, as several ova 

 in different stages of ripening were found in the centre as well as at the periphery. According 

 to Gegenbaur (see above), the ova are produced in the outer parts of the lamellae, while 

 Knower ") states that "the )oungest ova are found in the centre, immediately around the intra- 

 crlandular portion of the duct, the oldest ova with considerable yolk at the periphery". I cannot 

 ao-ree with either of these statements. Perhaps further histological investigations will throw 

 some light on the distribution of the ova in the genital gland. 



The efferent duct in Diacria arises from the dorsal surface of the gland, quite at the 

 right side, then runs over to the left, presenting the .same characteristic elongated swelling as 

 in Clio and opens into the accessory genital glands. This swelling of the efferent duct was 

 also found in Cavolinia inflcxa, contradictor)- to Pelseneer's statement, that this species "has 

 an ovoid vesicula seminalis, with a very short duct". 



In the other species of Cavolinia (s. sir.) the efferent duct, proceeding from the right 

 concave side of the genital gland, is very thin, until a long vesicula seminalis, coiled up at 

 the left side of the stomach, opens into it. This vesicula seminalis is sometimes provided with 

 small swellings from distance to distance "). The accessory genital glands in Cavolinia fs. sir.) 



i) Exei'c. Zool., Mem. Acad. Sci. Brux., vol. XII, 1S39. 



2) Voyage de la Bonite, vol. II, p. 100 — 136. 



3) Cnters. iiber Pterop. u. Het. 1S55, p. i — 40. 



4) Spolia atl. p. 86 — 91. 



5) Op. s. c. p. 17 — 21. 



6) Vevgl. Anat. Bd. I, fig., 392 (p. 833). 



7) Pteropods with two separate sexual openings, Ann. Mag. Nat. Hist. (6), vol. XlII, p. 529. 1894. 

 8) Not only in Cavolinia /ongiroslris^ as Pelseneer says, but also in Cavolinia triiUntata. 



