1904.] FROM EAST AFRICA AND ZANZIBAR, 381 



nearly all have a labial armature, and all * have teeth more or less 

 differentiated. I confess that I am afi^id of attaching too much 

 impor-tance to the radula : a case like Alclisa, where a Dorid with 

 otherwise ordinary characters has a unique dentition, certainl}- 

 shows that the teeth may vary without any corresponding change 

 in other characters ; but in the genera here grouped together it 

 will be found that the buccal characters are usually accompanied 

 by some other feature which allies them to Chromodoris, such as 

 simply pinnate branchife or a long narrow shape. It may be said 

 that the teeth of Dorids are never really uniform, and in the 

 genera described in my previous paper are often denticulate at 

 the outer end of the row. This is true, but the outermost teeth 

 are less well developed and more exposed than the others. They 

 therefore have a natural tendency for purely mechanical reasons 

 to become smaller and more irregular, and a particular form of 

 this irregularity, due perhaps to some peculiarity of texture, is 

 seen when they split up and become jagged or denticulate. But 

 no such mechanical explanation will account for the innermost 

 teeth being larger and more elaborately formed than the rest. 

 Also this peculiarity is confined to certain genera, whereas the 

 irregularity of the outermost teeth is general among the Crypto- 

 branchiates and as noticeable in Chromodoris as elsewhere. 



Casella and Ceratosoma are clearly closely allied to Chromodoris, 

 the former being perhaps not really a separate genus. Thorunna 

 is practically Chromodoris without a labial armature. Aphelodoris 

 has an elongate shape and narrow mantle-edge, but tripinnate 

 branchiae and no labial armature. It seems, however, to be allied 

 to Chromodoris by the presence of an accessory denticle on the 

 innermost teeth. The remaining genera are of more or less oval 

 shape, with a fairly wide mantle-margin, and papillae or tubercles 

 on the back, peculiarities which are found in some species of 

 Chromodoris. Sphcerodoris has simply pinnate branchiae and a 

 radula which, though peculiar, is essentially of the Chromodoris- 

 type. Orodoris, which Bergh associates with Sphcerodoris and 

 Miamira, has the median part of the radula much as in Chromo- 

 doris. Halla t and Rostanga are allied to Chromodoris by their 

 simply-pinnate branchiae as well as by their buccal parts : indeed, 

 the former appears to me almost an alaerrant Chromodorid akin to 

 such forms as Chr. sykesi described below. In Rostanga the 

 Mediteri^nean species perspicillata has denticulate inner teeth : 

 in coccinea they are merely bifid. The buccal parts of Tyrinna 

 and Cadlina strongly resemble those of Chr. scahriuscula, which 

 has also a somewhat oval form and tuberculate back. I somewhat 

 doubtfully refer Aitdura to the same gi'oup, in virtue of its radula. 

 This position is somewhat supported by its smooth skin and 

 scanty bipinnate branchiae, but the structure of the foot suggests 

 other affinities. 



* Except the very anomalous Miamira, which Bergh regards as alhed to Sphcero- 

 doris and Orodoris. 

 f See note § on p. 380. 



