290 SIR C. ELIOT AND T. J. EVANS. 
represent two lines of development in two different direc- 
tions. The Hlysioidea are derived from the olidioidea. 
Bergh has frequently expressed the opinion that our know- 
ledge of the Opisthobranchiata is not sufficient for the 
formulation of any phylogeny, and his own views are so 
tentatively and undogmatically expressed that it is difficult 
to summarise them. In his ‘System der Nudibranchiaten 
Gasteropoden’ (1892) he appears to regard the Nudibran- 
chiata as diphyletic, p. 996 (‘ Durch die Ascoglossen kniipft 
die eine Gruppe der Nudibranchien an die Aphysiaden und 
die Bulliden an, die andere durch die Pleurobranchiden 
wieder an diese letzteren”), the A‘olids being nearest to the. 
Ascoglossans, and the Tritonidee being derived from the 
Molids by gradual reduction of the hepatic ramifications, as 
seen in Bornella, Dendronotus, and Scyllea. In his 
article ‘‘ Ueber clado- und holohepatische nudibranchiate 
Gastropoden” (‘ Zool. Jahrb. System.,’ 1906, pp. 739—742), 
while still maintaining this view of the Tritoniide, he regards 
Tritonidoxa, Doridoxa, and Bathydoris as bridging 
over the interval between the Tritoniidz and the Dorids or 
Holohepatica. It is not clear what is the relationship of 
the Holohepatica to the Pleurobranchide on this hypothesis, 
and it seems to be implied that the AMolids, which are a 
highly specialised type, lose their peculiarities and pass into | 
the Tritonids, which are a comparatively generalised type, 
and that the Tritonids then develop a new highly specialised 
type, the Dorids. We find it hard to accept this view with- 
out stronger evidence than is forthcoming. ‘The branching 
of the liver may disappear in some cases,! and the animal 
which forms the subject of this memoir might be regarded as 
a derivative of the A#olidiidee which has retained its clado- 
hepatic system and adopted a doridiform shape. But a con- 
sideration of the whole series of forms now known (many of 
which have been described only in the last ten years) inclines 
1. g. in Pseudovermis, and Trinchese states that in the larva of 
Lomanotus eisigi the eolidiform and cladohepatic characters are much 
more marked than in the adult. 
