The members of the superfamily Paguroidea (comprising the families Coenobitidae, Diogenidae, Paguridae, and Parapaguridae) are not shown to be a monophyletic group (Fig. 1& Fig. 2). The inclusion of some members of the superfamily Galatheoidea within the same clade indicates paraphyly for both superfamilies. This is in contrast with the monophyletic Paguroidea suggested by McLaughlin (1983b) and Richter & Scholtz (1994) using adult somatic morphological characters in their analyses. Admittedly though, the monophyly of the Paguroidea is only compromised spermatologically by the inclusion of the investigated galatheids and chirostylids, both sharing the apomorphic possession of three microtubular arms with the spermatozoa of the paguroids (Fig. 1, point D). The separation of the Paguroidea into the traditional "right" and "left- handed" hermit crabs is not upheld. The inclusion of the Clibanarius clade (Fig. 1, point E), diogenid or ‘lefthanded‘ hermit crabs, within the clade containing the members of the Paguridae, which are considered ‘right-handed‘, marrs this classical dichotomy. The general placement of the major genera in the Paguroidea is consistent with the intuitive phylogenetic trees previously suggested by the author (Tudge, 1991, 1992).
Family Coenobitidae The six species in the two genera investigated in this family, Birgus and Coenobita, consistently form a monophyletic group, and, interestingly, this is a subclade of the majority of the diogenid members (Fig. 1& Fig. 2). This close association between members of the Coenobitidae and Diogenidae has previously been suggested (McLaughlin, 1983b; Tudge, 1991, 1992; Richter & Scholtz, 1994). Monophyly of the Coenobitidae is supported on somatic morphological grounds by McLaughlin (1983b) , Martin & Abele (1986) , and Richter & Scholtz (1994)and now strongly supported spermatologically.
Family Diogenidae The members of this morphologically diverse family investigated did not form a single clade but were split between two clades (Fig. 1& Fig. 2), the genera Clibanarius (Fig. 1, point E) and Cancellus being separated from the remainder. The single representative of the genus Cancellus exhibits several unusual spermatozoal features (Tudge, 1995a, b) that separate it from the Diogenidae (of which it is conventionally a member) and place it in a basal position in relation to the other paguroids. The heterogeneity of the genera within the family Diogenidae and the difficulty in determining their affinities has been expressed by Forest (1984, 1995) , but he considers the family to be an ancient monophyletic group. Richter & Scholtz (1994) have suggested that the Diogenidae are most likely paraphyletic (as indicated in the present study), but they presented no supporting evidence.
Family Paguridae The representatives of the family Paguridae are not retained as a monophyletic group. The inclusion of the parapagurid Sympagurus sp. in the Pagurus clade and the exclusion of Porcellanopagurus sp. make the family paraphyletic, at least in terms of spermatological features (Fig. 1& Fig. 2). The pagurid clade (with the exception of Porcellanopagurus) is characterized by the presence of one or more reticulated acrosome zones (character 18) in the spermatozoa (Fig. 1), although the exact homology of this character between all the inclusive genera is uncertain. The Pagurus-Sympagurus clade is defined by the presence of the accessory ampulla (character 29) in the spermatophore of the examined species (Fig. 1). Further investigation of the occurrence of the accessory ampulla throughout representatives in the Paguridae and related families may confirm the validity of this character for phylogenetic analysis. The exclusion of Pagurus hirtimanus from the clade containing the remainder of the species of the genus Pagurus reflects the difference in spermatozoal morphology that exists between this Indo-Pacific species and its European relatives P. bernhardus, P. chevreuxi and P. prideaux (Tudge, 1995a, b). The very speciose genus Pagurus (with over 150 taxa) is a heterogeneous assemblage with, as yet, ill-defined affinities with a rapidly growing number of sister-genera. A paraphyletic Paguridae has been suggested previously by Gore & Scotto (1983), McLaughlin (1983b) , and Richter & Scholtz (1994) and is intimated in the present analysis.
Family Parapaguridae The single investigated parapagurid, Sympagurus sp. lies within the clade containing the members of the family Paguridae in this analysis (Fig. 1& Fig. 2). The presence of an accessory ampulla (character 29) and, less so, the form of the spermatophore stalk (character 31) link some members of the family Paguridae with the investigated parapagurids. Similarly, the perforatorial chamber shape (character 5), and absence of microvillar projections (character 6) are synapomorphies (Tudge, 1995a, b).