32 
THE VOYAGE OE H.M.S. CHALLENGER. 
is precisely what is observed in Peraclis, whose operculum closely resembles that 
of the young Cymbuliidae. 
4. Besides this it has been pointed out that the older larvae of the Cymbuliidae, in 
which the proboscis is not yet reflexed in front of the fin, have the head shaped 
exactly like that of Peraclis (compare for example Gleba, PI. III. fig. 12) ; it is 
quite symmetrical as regards the tentacles, with the same small prominent 
proboscis, the penis even being also situated on the right side. Further, in 
Peraclis, as in all the Cymbuliidae, the central nervous system has three visceral 
ganglia. 
We may say then that the Cymbuliidae have descended from “ Limacinoid ” ancestors, 
and that among the recent Limacinidae Peraclis is the form which most closely resembles 
those ancestors. 
II. It now remains, then, for us to ascertain which of the two families, Limacinidae 
and Cavoliniidae, is the more primitive ; that is to say, are the Cavoliniidae descended 
from the Limacinidae or have they given origin to them \ Very numerous arguments 
speak in favour of the former hypothesis, and show that the rotation of 180 degrees, 
which has been alluded to above, has taken place from right to left in a Limacinoid type 
to give rise to the Cavoliniidae, whilst the converse is impossible. 
1. Let us consider first the relations of the digestive tract and genital duct. We see 
(PI. I. fig. 3) that in the Limacinidae the duct arises on the ventral side of the 
gland and passes directly to the right side of the body, making a quarter of a 
revolution (90 degrees) round the digestive tract (see PL III. fig. 6). In the 
Cavoliniidae (PL II. fig. 1) the duct arises on the dorsal edge of the genital 
gland, and instead of proceeding directly to the right side of the body where the 
genital aperture is situated, it passes by the left side, then ventrally, and finally 
reaches the right side, having thus made three-quarters of a revolution (270 
degrees) around the digestive tract (see Pl. III. fig. 6). It is clear that of these 
two routes the shorter (tha,t is, 90 degrees) is the more primitive. It follows then 
that in the Cavoliniidae, considering the genital aperture which is in the head as 
fixed, the visceral portion where the genital gland is situated along with the 
origin of the genital duct, has made a half rotation from right to left around 
a longitudinal axis, which explains why the genital duct makes a half revolution 
(180 degrees) round the tube more than in the Limacinidae. 
2. A large number of events in the development prove beyond all possibility of 
contradiction that the Cavoliniidae spring from Limacinoid ancestors by the 
relative displacement of the visceral and cephalic portions of the body. 
A. In stating his theory of rotation in the Cavoliniidae, Boas remarks that the 
anterior or cephalic portion has performed a rotation of 180 degrees relatively to 
the posterior part, or vice versd. This manner of expressing the process does 
