A marginal population may be defined as one which exists on the edge of a species' range. Carson (1959) points out that since the distribution of any species is rarely continuous within its range marginal populations may be geographically internal as well as peripheral. In the absence of major barriers (such as the sea for terrestrial species) these populations exist presumably in the face of environmental stress or competitive interactions which prevent the species from extending its range and a number of generalizations have been made about them: 1) They may be restricted to a particular habitat to which the species is not otherwise confined (e.g., Ford, 1964). 2) They may show morphological adaptations different from those in central populations (Cook, 1961). 3) They are likely to be isolated anid less dense than central populations (Mayr, 1963). 4) They may show a reduction in the degree of polymorphism at the visible (e.g., Bovey, 1941) or chromosomal (e.g., Carson, 1955) level. None of these generalizations are universally true and it is at present impossible to predict what form the response of a species to sub-optimal conditions will take (Levins, 1968). For example it is known that marginal populations of some species of Drosophila (e.g., D. robusta, D. willistoni) are less polymorphic for chromosome inversions than central ones but also that this is not true for D. subobscura (Carson, 1959). For these species in which this reduction in inversion frequency occurs Carson has argued that increased recombination results from greater homozygosity for chromosome structure and that this is a reflection of the need to maintain genetic flexibility in terms of evolutionary adjustment to, unfavorable conditions. Marginal populations are likely to be of low density and consequently more inbred. Slizynski (1955) has shown that inbreeding lines of mice have greater recombination than outbreeding ones; it follows that a raised chiasma frequency in a small marginal population may be more a consequence of inbreeding than of a selective adjustment to a particular set of environmental conditions. Conclusions relating to the evolutionary significance of a raised chiasma frequency in marginal populations are valid only if it can be shown that this occurs independently of changes in population density. The landsnails Cepaea nemoralis (L.) and C. hortensis (Mull.) are widespread in Europe but neither species is continuously distributed. Both major geographical (Perrot and Perrot, 1938) and microgeographical (e.g., Cain and Currey, 1963) differences in distribution suggest that C. nemoralis is favored under warmer and drier conditions (e.g., Harvey, 1974); mixed populations of the 2 species are not common. Differences in climatic preferences between the species are suggested also by experimental data (Cameron, 1970a, b). Both species possess a similar genetic polymorphism for shell color and banding and circumstantial (e.g., Jones, 1973), direct (Richardson, 1974) and experimental (Lamotte, 1959) evidence suggests that climate (in particular temperature variation) is important in affecting morph frequencies.