SUMM ARY Parsimony analysis, w hether using only spermatozoal characters or spermatozoal and non-spermatozoal characters, suggests that: (i) the Brachyura is a monophyletic taxon relative to the outgroup, three paguroids; (ii) the Podotrem ata is a monophyletic taxon and the sister-group of the heterotrem ethoracotrem e assemblage; (iii) within the Podotrem ata, the Dromiidae appears paraphyletic as its clade includes Paradynomene; (iv) the Hom olidae is a monophyletic family; (v) the R

more »

... a and Cyclodorippoidea are sister groups within a monophyletic (unnamed) clade; (vi) Latreillia sp. forms a polytomy with Hom olidae + Raninoidea-iCyclodorippoidea; and (vii) within the heterotrem ethoracotrem e assemblage, the T horacotrem ata is a monophyletic taxon but the H eterotrem ata 5 . strict, is a paraphyletic grouping. Analyses based on sperm data alone differ from the combined data in not including the Homolidae in the Archaeobrachyura (Homolidae + R aninoidea-C yclodorippoidea), thus the Drom iida e-Paradynomene + Homolidae + Latreillia form a polytomous clade which is t sister-group of the Raninoidea + Cyclodorippoidea. Spermatozoal data also give majids the most basal position in the H eterotrem ata whereas for the combined data Neodorippe (a shell carrier) appears the least modified member of the heterotrem e-thoracotrem e assemblage. These findings are viewed in the perspective of other, notably molecular, studies. Spermatozoal synapomorphies are listed and illustrated. Crab phylogeny and spermatozoa B. G. M. Jamieson 375 ©continuity' © th ick n ess © w idth (H^extenf @ fo rm _ of operculum. ©perforation of operculum and apical button @ opercular \ © sub opercular material projections^ © a c c e s s o r y of perforatorium-V A V M V / A © a c ro s o m e -;>^£^V V V > proportions © @ l @ ( l 3 zonation o f ' acrosom e