It has been increasingly recognized in recent years that many closely related pairs of species are parapatric, i.e. have very narrowly overlapping ranges (White, 1977). The overlap zones, in which hybridization may be occurring, are often only a few hundred meters in width; they are referred to as 'tension zones' by Key (1974), because of the selective pressures believed to be operating in them. The study of such parapatric zones is now an important field of evolutionary biology (Nevo and Bar-El, 1976; Szymura, 1976; Moran and Shaw, 1977). The 'stasipatric' model of speciation (which involves the concept of a chromosomal rearrangement generating a postmating genetic isolation between parapatric populations) seems to be applicable in the case of many organisms of restricted vagility such as wingless grasshoppers and small rodents (White et al., 1967; White, 1968; Bush et al., 1977). It was based in the first instance on the Australian eumastacid grasshoppers of the genus Vandiemenella-formerly referred to as the 'viatica group'-which have parapatric distributions in southern Australia. The present paper deals with a highly unusual instance of parapatric distribution-where one of the species reproduces exclusively by parthenogenesis and the other is bisexual. The genetic system of the parthenogenetic grasshopper Warramaba virgo (Eumastacidae, Morabinae) has been extensively discussed earlier (White et al., 1977), while that of the bisexual species, still unnamed taxonomically, but referred to as 'P196,' was dealt with by White et al. (1973). It is now almost certain, contrary to earlier interpretations (White and Webb, 1968), that W. virgo originated as a hybrid between P196 and another species of Warramaba, 'P169' (Hewitt, 1975; White et al., 1977). Based on the distribution of W. virgo in widely separated areas of eastern and western Australia (White et al., 1977, Fig. 1) and the extensive karyotypic differences between clones inhabiting limited areas (Webb et al., 1978), this event probably took place some tens of thousands of years ago, in western Australia (the bisexual species of Warramaba do not exist in eastern Australia). It has been asserted by some recent authors (e.g. Williams, 1975; MaynardSmith, 197 la,b) that parthenogenetic species should have a higher reproductive fitness than their bisexual relatives. If this were so in the present case, Warramaba virgo should have expanded its range at the expense of its bisexual 'parents' (P196 and P169). To some slight extent it may have done so; but its known range in western Australia is quite restricted (see White et al., 1977, Fig. 1) and its main extension, in the past, must have been eastward for over 2,300 km as far as central New South Wales, i.e. across territory that was almost certainly never occupied by either P196 or P169. Cuellar (1977) has proposed the general principle that parthenogenetic taxa will not be able to coexist with populations of their bisexual forebears because hybridization in areas of overlap will lead to the production of inferior biotypes. Certainly, a vast exten-