Two sets of evolutionary phenomena find no explanation through current theory. For the static phenomena (such as homology, homonomy, systematic weight, and "Type") there is no causal base, although these principles are responsible for all phenomena of predictable order in the living world. The dynamic phenomena (such as homodynamy, coadaptation, parallel evolution, orthogenesis, Cartesian transformation, typostrophy, hetermorphosis, systemic mutation, and spontaneous atavism) have no causal explanation, although they are responsible for all directed phenomena in macroevolution. These phenomena share one unifying principle which can be explained by a system theory of evolution based on, but extending, the current synthetic theory. This system theory envisages feedback conditions between genotype and phenotype by which the chances of successful adaptation increase if the genetic units, by insertion of superimposed genes, copy the functional dependencies of those phene structures for which they code. This positive feedback of the adaptive speed (or probability) within a single adaptive direction is compensated by negative feedback in most of the alternative directions. The negative feedback operates as selection not be environmental but by systemic conditions developed by the organization of the organism. The consequences are an imitatively organized system of gene interractions, the rehabilitation of classical systematics, the reality of the "natural system," and, in general, the resolution of the contradiction between neodarwinists and their critics, between reductionists and holists, between "a priori" and "a posteriori" views, between idealism and materialism, and between the notions of freedom and of purpose in evolution.