384 SIR c. ELIOT ON NUDIBRAXCHS [Mar. 1, 



the plumes do not resemble the ample and elaboi'ate rosettes of 

 Archidoris, Platydoris, &c. The arrangement and the approximate 

 (but not the exact) number of branchice will probably be found 

 good specific cliaracters in cases wheie a sufficient number of 

 individuals has been examined ; but I think that the number of 

 plumes increases with age, at any rate in some species, and that 

 hence the data furnished by a single specimen may be misleading. 

 This is the case particulai'ly in forms with a spiral ari'angement, 

 where it seems likely that the small plumes in the spire are 

 developed later than the others. Thus in Chr. sykesi the number 

 varies from 12 to 18 and in Chr. anmdata from 9 to 16. In the 

 latter species a spire is present in some cases and absent in others. 

 The commonest number of branchiae seems to be about 10 

 (varying from 8 to 12 in individuals). A distinctly smaller 

 number (3 to 7) is found in Chr. cardvncdis, jttvenca, elegantula, 

 iheringi, krohnii, virginea, gloriosa, scurra, luxuriosa, albonotata, 

 inconspicica, elegans, camosna, riulolphi, panthareUa, and is indi- 

 cated in the drawings of many other species. In Chr. punctilacens, 

 lineolata, paupera, tryoni, bennetti, splendens, a'ossei, dalli, reticu- 

 lata, godeffroyana, mollita, vicina, sykesi, and annulata the number 

 of plvimes is more than 12 and often exceeds 20. 



I have observed that in making a superficial examination of the 

 living animal one is very apt to under- estimate the number of 

 branchiae, inasmuch as the longer ones may project and the 

 shorter ones be hidden. Hence the plates and descriptions of 

 older authors cannot be considered as decisive on this point. 



6. The labial ai-mature consists sometimes of two plates and 

 sometimes of a continuous ring. Tlie elements are hardly ever 

 straight, but are more or less bent oi- hooked. The}- are of very 

 varying shape and thickness, a,nd sometimes swollen just below 

 the tip or terminal hook, so that they assume a mace-like appear- 

 ance. The shape usually affords a good specific character, but in 

 some species the elements ai-e bifid in one part of the armature 

 and entire in another. 



7. The radula ofl'eis many differences. Chr. scahriuscula has a 

 central tooth and Chr. juvenca a central plate which is called 

 " false " by Bergh, though it seems well developed. Rhachidian 

 thickenings, more or less distinct, are found in Chr. ccerulea, 

 iheringi, morchii, gonatophora, porcata, carnea, onariana, hilaris, 

 sannio, lineolata, marenzelleri, cardinalis, reticulata, dalli, cavcp., 

 annulata, splendens. The radula is as a rule of moderate size, 

 but is very narrow in Chr. elegantula, krohnii, gloriosa, and rather 

 narrow in piistulans (75 x 28.0.28) and some other forms. In 

 scabriuscida it is 108 x 30.1.30. It is unusually broad in thalas- 

 sopora (71 x 162.0.162), and above all in .syci/Za" (81 X 290.0.290). 



The formulae for the radula given in my descriptions are merely 

 the shortest way of desci'ibing the teeth of a given specimen, and 

 must not be understood as necessarily characteristic of the species. 

 The proportion between length and breadth is generally roughly 

 the same in different individuals, but the number of roAvs and of 

 [6] 



