permanent and heritable modification. The increase in nuclear material involved in chromosomal mutation (_i.e._ duplication) seems to cause a proportionate increase in the cytoplasmic mass of the single somatic cells, which manifests itself in the phenotype as giantism. De Vries’ _Œnothera gigas_ is a chromosomal mutant illustrative of this phenomenon. Besides the foregoing, there is the _pseudomutant_ produced by the factorial recombination, which results from a _crossover_, _i.e._ an exchange of genes or factors between two germinal chromosomes of the same synaptic pair. This reciprocal transfer of genes from one homologous chromosome to another happens, in a certain percentage of cases, during synapsis. The percentage can be artificially increased by exposing young female hybrids to special conditions of temperature. [2] The term mutation had been used long before and in a similar sense by the German palæontologist Waagen, who employed it to designate the variations of a specific type that succeed one another in successive strata, a thing which rarely occurs. (Cf. Waagen’s _Die Formenreihe des Ammonites subradiatus_, Geognost. paläont. Beitr., Berlin, 1869.) If these new mutant forms could be regarded as genuine new species, then the fact that such variations are heritable and come within the range of actual observation, would constitute the long-sought empirical proof of the reality of evolution. Consciously or subconsciously, however, De Vries recognized that this was not the case; for he refers to mutants as “elementary species,” and does not venture to present them as authentic organic species.