1 Ecotypic differentiation is hereditary differentiation in respect of morphological and/or physiological attributes occasioned by the selective action of the habitat environment.

2. Ecotypic differentiation in response to one set of environmental conditions is liable to proceed independently of that in response to another set of conditions. Regularities in ecotypic differentiation are therefore best studied in relation to particular environmental trends.

3. Ecotypic differentiation may be regional or strictly local.

4. Regional ecotypic differentiation occurs most frequently in species with wide latitudinal and/or altitudinal distributions, i.e. it appears as a response to variations in climate.

5. A feature of ecotypic differentiation in response to climate is that the range of variation is not usually continuous in the direction of a gradual change in climate, but instead a limited number of populations is formed each with its own climatic ‘latitude of tolerance’.

6. Where neighbouring climatic populations meet they lose their individuality, as a rule, only along comparatively narrow zones of intergradation. The limits of the areas of intergradation cannot, however, always be assessed with accuracy as the effects of hybridization ‘diminish often at some distance from the point of contact’ (Clausen et al.), and some genes may travel far.

7. Local ecotypic differentiation may also be characterized by the formation of relatively uniform populations separated by intergradation zones of greater variability. But when adjacent populations occupy small and contrasting habitats the evidence suggests that the effects of hybridization may extend over the entire communities without completely obliterating their respective identities.

8. The local establishment of ecotypically individualistic communities is considerably facilitated where a colonial population structure is pronounced.

9. Ecotypically differentiated populations may coincide with geographically differentiated subspecies. When they do coincide with subspecies, or when they do not coincide but still exhibit the fundamental features of truly objective populations, i.e. have definable borders and some degree of ecotypic individuality, they ought to be taxonomically recorded at the subspecies level. The use of the term ‘ecoclinal subspecies’ has been suggested as a convenient way of making the necessary distinction between experimentally delimited ecological subspecies and orthodox subspecies.

10. Ecoclinal subspecies should receive taxonomic names.

11. Ecotypic differentiation related to particular environmental trends but which can only be recorded in terms of subjectively delimited ranges of ecotypic variation, is not ‘subspecific’, and populations assigned to a subjective range should not be accorded subspecific status. Such populations have been referred to as ‘ecoclinal ecotypes’

12. Ecoclinal ecotypes should receive only vernacular titles.

13. The general term ecotype as defined by Turesson is applicable to any ‘product’ of ecotypic differentiation.

14. When it is necessary to refer to habitat populations the ecotypic significance of which has not been established the term ecotype should be avoided, and some non-committal term such as ecodeme should be used instead.

The writer wishes to take this opportunity of thanking Dr Julian Huxley and Dr Irene Manton for suggestions which have been most helpful in deciding the scope and nature of the present review.