Abstract
Contrary to Mendel, who introduced hybridization as a methodology for the study of selected discrete traits, de Vries conceived of organisms to be composed of discrete traits. This introduced into genetic research the dialectics of reductive analysis of genes as instrumental variables versus that of genes as the material atoms of heredity. The latter conception gained support with the analysis of mutations and eventually with high resolution analysis at the genetic and biochemical levels, as achieved in fungi and later in bacteria and their viruses. Attempts to reduce "classical" genetics to "molecular" genetics turned out to be futile. However, this did not necessarily imply that these were two distinct theoretical approaches. On the contrary, it is argued that molecular genetics is an extension of phenomenological deduction, rather than being induction from molecular (DNA) causes to effects. Although conceptually systems direct development, methodologically individual inputs must be studied