The aim of Roux's 'developmental mechanics' was to determine the causes of individual development. One of the central points of 'developmental mechanics' is the distinction between processes of 'self-differentiation' and processes of 'dependent differentiation'. In the case of self-differentiation the factors responsible for the differentiation of an egg or part of an egg lie in the part itself, whereas the case of dependent differentiation the specific development is driven essentially by external conditions. The principal type of dependent differentiation is 'functional adaptation', a concept by which Roux tried to explain in mechanistic, non-teleological way the adaptation of the inner structure of various organs (e.g. bones, cartilage, muscles) to their specific function. It is common to regard Roux's embryological experiments on problems of self-differentiation as more or less isolated from his research on functional adaptation. Such an approach is inadequate since 'developmental mechanics' rests on both research fields equally. This is also testified by the specific character of Roux's synthetic theory of ontogeny, which divides ontogeny into several phases according to their different shares of inherited and non-inherited, functional factors of differentiation.