542 SIR CHARLES ELIOT. 



I have therefore given somewhat full descriptions of the family and generic characteristics 

 which I trust will not be considered out of place. In classification I have adopted Bergh's 

 division into Holohepatica and Cladohepatica in preference to Pelseneer's four tribes Tritonioidea, 

 Doridioidea, Aeolidioidea, Elysioidea. There is probably no group of comparatively well-known 

 animals in which we may look forward to as many additions to our knowledge as in the 

 Nudibranchiata, and to the discovery of forms which are not merely new but of morpho- 

 logical and systematic importance. Such are Doridoxa, Atthila and Hedyle discovered in the 

 last few years, and as few seas have been thoroughly investigated and the fauna of many 

 large marine areas (such as the tropical and southern Atlantic) remain as far as Nudi- 

 branchs are concerned practically unknown, many more finds may be anticipated. It is 

 therefore better to have as elastic a system as possible, and to arrange the group in a series 

 into which new members can be easily inserted, than to cut it up into rigid subdivisions 

 whose definitions may require alteration. Pelseneer's Tritonioidea contain a somewhat mis- 

 cellaneous assemblage, and it is not clear what is the distinction between them and the 

 less typical Aeolidioidea. Why, for instance, does Dendronotus belong to the former and Doto 

 to the latter class ? I have, however, departed from Bergh's arrangement in two respects. 

 Firstly, I classify Tritonia among the Holohepatica, and not as he does among the Clado- 

 hepatica. It is true that its general anatomy allies it rather to the Cladohepatica, but it 

 is clearly a connecting link between the two tribes, and as the distinction between them 

 is the entire or divided condition of the liver, and as the liver of Tritonia is entire, it 

 seems more logical to regard it as a holohepatic form with affinities to the Cladohepatica 

 than as a cladohepatic form (which it is not) with affinities to the Holohepatica. Secondly, 

 I have included among the Cladohepatica the Elysioidea or Ascoglossa, which cannot in my 

 opinion be separated firom the rest of the Nudibranchiata. 



HOLOHEPATICA. 



The Nudibranchiata Holohepatica are characterized by the liver being entire and not 

 ramified. They may be divided into two sections, the first containing the Tritoniadae and 

 the second all the other forms. As mentioned, the members of the first section are really 

 intermediate between the Holo- and Cladohepatica: in virtue of their unbranched liver they 

 belong to the former but in their other characters they bear more resemblance to the latter. 

 The other section, often conveniently called Dorids, are characterized by possessing two 

 spermatothecas and a blood gland. Further, in all except a few abnormal genera the branchiae 

 are dorsal and arranged round the vent. 



Fam. Tritoniadae \ 



The characters of the family may be defined as follows. Animals of moderate or even 

 large size with an elongate and somewhat rectangular body. A frontal veil, usually provided 

 with appendages and two somewhat spoon-shaped tentacles are present. Rhinophores 



1 For the study of this family the following may be con- xvii. (2) Nudibr. Mollusca of North Pacific (in Dale's 



suited. Alder and Hancock, British Nudibr. Mollusca, 18.5.5, Results of Exploration of Alaska). (3) Beitr. zur Kenntnus 



pt. VII. Vayssi^re, " Description du il/anonia iJer^ftu'.'Vou?-. derJapanischen Nudibr. u. (4) Report on the Nudibranchiata 



de Conchyl. 3. S. xix. 1879. Pelseneer, RecJmrches sur divers of the Challenger, 1884. (5) Opisthobranchier der Sammlung, 



Opisthabranches, Bruxelles, 1893, and especially the following Plate. (6) Nudibranchiate Gasteropoda of tlie Ingolf Expe- 



papers of Bergh. (1) Semper's Reisen, Hefte xv., xvi. and ditioii, 1900. 



