NUDIBRANCHIATA OF THE L.M.B.C. DISTRICT. 149 
In Ancula cristata the lateral ridge has almost disappeared 
as a ridge, but it is evident that the five pairs of large 
simple cerata placed at the sides of the branchie (fig. 10) 
correspond to the similar structures seen in Polycera and 
Idala. Then in Triopa claviger (fig. 12) the cerata have 
become more numerous (seven pairs), are directed more 
laterally, and extend from the head nearly to the posterior 
end of the body. 
In passing next to the family Tritontide we find that 
the cerata become branched in an arborescent manner, but 
on comparing sections of Goniodoris (fig. 6) or Polycera 
(fig. 9) with those of Candiella plebeta (fig. 18) it is im- 
possible to doubt that one is dealing with the same series 
of projections. Cabrilla occidentalis, which has been lately 
described and figured by Fewkes,* presents an interesting 
intermediate condition between Zriopa and Tritonia. Cab- 
rilla is evidently referable to the Doride; it has short 
laminated rhinophores and a_posteriorly-placed circle 
of branchiz, but is possessed of six or seven pairs of 
laterally-placed cerata which are branched at their ends, 
and are evidently comparable with the parieto-cerata of 
Tritonia and Dendronotus. In Dendronotus the large parieto- 
cerata become very complicated in form (fig. 14), but are 
evidently merely a further development of the smaller but 
similar processes of Candiella or Tritonia. 
Finally, in the great group Cladohepatica we find large 
and conspicuous hepato-cerata (Pl. VI. figs 15 and 16), as 
in Doto, Holis, and Proctonotus, but we must regard these 
as being merely cerata, originally like those of the Doride 
and Tritonide, which have been invaded by the hepatic 
ceca and have afterwards become enlarged and modified 
* Zoological Excursions, I. New Invertebrata from California: Boston, 
89, p. 44. 
