SUMMARY The mean volume of a diploid spore population is twice that of a haploid spore population; it is assumed that the mean volume of an aneuploid spore population is a similar reflexion of its ploidy. The distribution of volumes amongst a population of spores of Dictyostelium discoideum was analysed in terms of frequency of occurrence of diploid, aneuploid and haploid cells; in a wild-type strain (NC-4) these cell types were present in the ratio 8 : 3 : 89. It is suggested that 'metastable' strains are mutants in which this ratio has been altered. The life cycle of the cellular slime mould Dictyosteliurn discoideum is divided into two mutually exclusive phases. In the first, feeding, phase the organism exists as solitary amoeboid cells and in the second, differentiation, phase these hitherto solitary amoe- boid cells associate and, without any change in number, produce a fruiting body consisting of spore and stalk cells. The spores germinate in a suitable environment to give amoeboid cells, each spore giving one myxamoeba. The stalk cells are non- viable. Recent studies of mutants derived from a haploid strain of D. discoideum have shown that genetic exchange can occur at a low frequency during this life cycle (Loomis & Ashworth, 1968 ; Sinha & Ashworth, 1969 ; Loomis, I 969). Segregation occurs from heterozygous diploids in a manner which suggests haploidization via stepwise and random chromosome loss, rather than by meiotic division. Direct chromosome counts during the period of haploidization support this suggestion (Sinha & Ashworth, 1969). It appears, therefore, that D. discoideum has a parasexual cycle of the type described by Pontecorvo (1956) for Aspergillus and shown in Fig. 4. The genotype of the spores must therefore be identical with the genotype of the myxamoebae from which they are derived. There should thus exist, in any population of spores, diploid, haploid and aneuploid genotypes. Sussman & Sussman (1962) suggested that the ploidy of a spore could be correlated with its size. They reported that the mean major diameter of a diploid spore was about one and a half times the mean major diameter of a haploid spore. We report here that the mean volume of a diploid spore population is twice that of a haploid spore population and that the distribution of volumes amongst a spore population can therefore be analysed in terms of the frequency of occurrence of haploids and diploids. Such an analysis also enables the frequency of occurrence of aneuploids to be estimated. The nature of the distribution of ploidy types in the population can be calculated if it be assumed that very few non-viable aneuploid