At the end of our studies on heredity we hardly can avoid saying a few words about the problem of the localisation of the morphogenetic units in the germ-cells themselves. Is it in the protoplasm or in the nucleus that they are placed? You all know that this question was for a long time regarded as more important than any other, and perhaps you have already blamed me for not raising it until now. But in my opinion results gained by the purely analytical method and carefully established, are always superior to those which are of a merely descriptive nature and doubtful besides. The famous problem of the part played by the nucleus in inheritance is both descriptive and doubtful: it is only, so to say, of factual, not of analytical importance, and quite insoluble at present. As for our second proof of vitalism, stating that no kind of machine inside the germ-cells can possibly be the foundation of their morphogenesis, it is clear that the protoplasm and the nucleus may both come into account here on equal terms. If you prefer to say so, it is to the nucleus and to its division in particular that the second proof of autonomy relates, while the first, though not over-looking the presence of nuclei,[134] deals “especially” with the protoplasmic nature of its “systems.” [134] The first proof of vitalism, indeed, rests upon the analysis of the differentiation of an harmonious-equipotential system as a *whole*: this *whole* cannot be a machine that would relate to differentiation as a *whole*; the question whether there might be any machines distributed *in* the whole, in the form of the nuclei is of no importance at all in this argument. Moreover the pressure experiments (see page 63) prove the unimportance of such “machines” for the specificity of differentiation, and the second proof of vitalism shows that the nuclei cannot be regarded as machines accounting for differentiation in *any* way. What then can we say, on the basis of actual facts, about the part taken by the protoplasm and by the nucleus in inheritance, now that we have learnt from our analytical discussion that both of them cannot be any kind of morphogenetic machine, but can only be means of morphogenesis? Let us state our question in the following way: whereabouts in the germ-cells are those “means” of morphogenesis localised, the existence of which we infer from the material continuity in the course of generations in general and from the facts discovered about hybridisation in particular?