It is not the _resemblances_ existing between the members of the same species, or between the members of one family, which perplex philosophers: all agree in referring them to _heredity_. The problem lies rather in the _differences_. Not only in the considerable differences which are established between races; but more especially in the _shades_ constituting the _individual traits_ which distinguish father from son, or brother from brother. This is in reality the fundamental difficulty, and many hypotheses have been proposed for its solution. Prosper Lucas, after having discussed them separately, regarded them all as insufficient, and believed that, side by side with _heredity_, which maintains types, we ought to admit a special force, _innateness_ (_l’innéité_) which diversifies them. We can, however, account for the double tendency exhibited by living beings, without having recourse to a new force. For this purpose it is sufficient to push the analysis of phenomena a little further than is customary, and to obtain a clear idea of the part played by the _conditions of life_ (_milieu_) and _heredity_. As a general rule an action is attributed to the first, which everywhere and at all times is a modifying one, and to the second a purely conservative action. Now it may be easily shown that this is not the case; and that each of these causes acts in an inverse manner according to circumstances. IV. By virtue of the laws of heredity, the father and mother tend equally to transmit to their offspring their own character. However similar they may be supposed to be, there are always some differences between them; and the nature of the new being is necessarily a compromise between two different tendencies. The son cannot, therefore, always resemble his father exactly. In him the characters _common_ to both parents will easily be exaggerated; the _opposite_ characters will be neutralised; and the _different_ characters will produce a _resultant_, as distinct from the two components as green is from yellow and blue. Thus even by virtue of its own tendencies, and in consequence of the enforced co-operation of the sexes, _direct and immediate heredity_ becomes, in some respects, a cause of variation. _Mediate and indirect heredity_, justly compared by Burdach to geneagenetic phenomena, as well as _atavism_, which suddenly reproduces with great exactness the characters of an ancestor, sometimes after hundreds of generations, have certainly considerable influence in the variation of individual traits, and in the differences which distinguish parents from their children. Their action, added to that of direct heredity, is sufficient to explain the appearance of certain varieties, without appealing to _innateness_. V. But the hereditary force, although it is manifested from one generation to another, or through several generations, is always influenced by the _conditions of life_ (_milieu_), and this has evidently greater force.