356 SIR C. ELIOT ON NUDIBRANCHS [DeC. 1, 



the group, is bound to consider the value of the existing generic 

 distinctions. The five subfamilies are divided into about 30 

 genera all created by the great master of the Nudibranchiata, 

 Prof. Bergh, and several of them consisting of a single species. 

 He himself observes (on ThorcUsa in Semper's ' Reisen ' *, xv. 

 p. 666) that the definitions of many of these genera seem to be 

 too precise, and as new forms are discovered the old divisions are 

 found to melt away. The discoverer of a new form often finds 

 that it does not accurately fit in, to any of the existing genera, 

 and must ask himself whether he should create a new genus or 

 enlarge the definition. I have little doubt that in most cases 

 the latter is the preferable course. If animals are not divided by 

 natural difierences, there is no object in emphasising the im- 

 portance of minute peculiaiities. If Chromodo7-is is allowed to 

 consist of 100 species showing a considerable range of variation, 

 including the presence or absence of median teeth, there seems to 

 be a want of proportion in splitting uj) the other Cryptobranchiata 

 into so many genera. 



The genus Staurodoris offers a good example of the difficulty 

 of classifying new forms. Taken by itself, the typical species 

 St. verrucosa is remarkably well charactei'ised in both the scien- 

 tific and popular sense. Anyone could recognise it at first sight. 

 Tlie back is covered with club-like tubercles, the rhinophorial and 

 branchial pockets are protected by similar tubercles which act 

 as valves, and the branchiae are simply pinnate. On the other 

 hand, aS'^. pseudoverrucosa (von Ihering) has no tubercles on the 

 branchial pocket and has bipinnate branchias. Bergh also refers to 

 this genus the Doris pustidosa of Abraham, which has bipinnate 

 branchia3 and small, but apparently not valve-like, tubercles. I 

 have specimens from the Indian Ocean which have the dorsal 

 sui-face tuberculate and the following additional characteristics : — 

 A. has the rhinophores arising among tubercles and simply pinnate 

 branchiae, but no tubercles on the bi'anchial pocket {Staurodoris 

 pecten). B. has five pinnate gills, the anterior margin of the foot 

 entire, and small tubercles set on the edges of the gill-pocket and 

 partly closing it {Staurodoris calva). C. has tubercles round the 

 rhinophores, none on the branchial pocket, and thin bipinnate 

 branchife {Archido7'is africana). D. has tubercles on the rhino- 

 phore pockets bvit not around the branchial opening, and tri- 

 pinnate gills {Archidoris 'minor). Of these, I think we must admit 

 A. and B. to be Staurodoris, if we accept St. pseudoverrucosa. 

 But Archidoris kerguelensis, A. australis, A. ruhescens, A. incerta, 

 and A. nyctea are all desciibed by Bergh as having tubercles on 

 the edge of the branchial and rhinophorial pockets, and must 

 come veiy near the less typical members of Staurodoris. It is 

 hard therefore to say whether C. should be classed as Staurodoris 

 or Archidoris. There seems to be a complete series of links 

 between the two genera, and, this being so, we must either unite 



* The letters S. R. in this paper refer to Prof. Bergh's " Malacologische Unter- 

 suchutigen," puMished in ' Kei'^eii im Arohipel (ler Philippinen,' von Dr. C. Semper. 



