400 SIR C. ELIOT OX XUDIBItAXCTIS [Miir. 1, 



space behind. They are 24 in nunil)er : those in front are fairly 

 large, but the size diminishes backwards, and those in the spirals 

 are extremely small. The head and anterior portion of the foot 

 are much retractefl, but the lattei- was apparently I'ound and 

 grooved in life. 



The mouth leads into an unusually large and spacious cavity, 

 but the buccal mass is very small, the radula minvxte, and the 

 oesophagus extremely narrow. The labial armature is small but 

 strong, with rough projecting teeth ; the elements are small, 

 yellowish, bent rods, mostly bifid. No rhachidian thickenings are 

 visible ; the innermost teeth have three denticles on the inner 

 and about four on the outer side. The i-emainder have mostly 

 five on the outer side only. The niimber of denticles increases 

 towards the outside of each row, but the outermost teeth are 

 smooth. The teeth are very small, crowded, and exti'emely 

 numerous. 



This specimen corresponds with previous descriptions of C. atro- 

 Tnarginata, except that thei'e are no thickenings on the rhachis 

 of the radula and that the gills ai-e much moi-e numerous than in 

 the specimens previously described. There seems to be some doubt 

 on this point (see Bergh, Mus. Godeffi'. 1. c), but the arrangement 

 in the present specimen is perfectly clear. Casella cincta from 

 Mauritius has 22 gills (Bergh, S. R. xvi. 2, p. 839), but })i'esents 

 differences in the coloration and buccal pai-ts. 



Genus Ceratosoma, Adams & Reeve. 



This genus is rendered unique among Nudibivanchs by its extra- 

 ordinaiy shape, but in its essential characters it is closely allied 

 to Chromodoris. Indeed, if one looks at one of tlie high, stout 

 Chromodorids {e. g. the figure of C. semperi in S. R. Heft xi. pi. Iv. 

 fig. 2), it will be seen that one has only to somewhat prolong the 

 tail and to thicken and develop the mantle-edge in oixler to obtain 

 the characteristic shape of Ceratosoma. The large strong latlula 

 is like that of CJtromodoris, but the denticles are minute and 

 inconspicuous. The branchia? ai-e much divided. Bergh recognises 

 nine species, but I have only seen the descriptions of those 

 examined by himself {cornigeriim, gracillimicm., tynlobahim, orna- 

 tum, polyomma). In Zanzibar I have inspected more than foi'ty 

 specimens which are apparently referable to one species, and are 

 connected together by numei'ous gradations in colour and shape, 

 though the extreme forms look remarkably diffei-ent. As the five 

 species mentioned above are distinguished by theii- extei-nal 

 charactei's only, and off"er no certain differences in the dentition or 

 other organs, I am inclined to think they are merely varieties and 

 that thei-e is only one real species. It is to be noted, however, 

 that though many of my specimens resembled the figure of 

 Ceratosoma gracillimmn (S. R. pi. xxv. fig. 8) the border was never 

 red as there depicted, but always violet, and in no case did ocelli 

 occur as in C. 'polyomma. 

 [22] 



