In this study, all the existing specimens of the Doridoxidae were re-examined; that is (a) the holotype of
Doridoxa ingolfiana Bergh, 1899 from western Greenland, (b) a specimen described as "
D. ingolfiana var.?" by Bergh (1899) from Iceland, (c) another more recently collected specimen from southern Greenland, which was serially sectioned and examined histologically as part of the present study, and (d) the holotype of
D. benthalis Barnard, 1963. Significant discrepancies between the first three of these specimens and Bergh's original description of
D. ingolfiana are apparent, but they do not alter our opinion that all three specimens belong to a single, variable species. The monobasic taxon
D. benthalis Barnard, 1963 is specifically distinct from
D. ingolfiana.. Bergh (1899) considered that the Doridoxidae had a holohepatic digestive gland (and thus was a member of the Anthobranchia). Our reassessments, particularly those concerning the structure of the digestive gland, branchial apparatus and anal site, indicate that Bergh misunderstood the form and polarities of these characters, and that the Doridoxidae must be relocated within the cladohepatic (i.e. actenidiacean) nudibranchs as its first offshoot. The Nudibranchia are thus divided into two major clades: Anthobranchia (= Bathydoridoidea + Doridoidea) and
Dexiarchia nom. nov. (= Doridoxoidea + Dendronotoidea + Aeolidoidea + Arminoidea)
