In this essay, I examine the possible use of the distinction between discovery and justification for the analysis of research practice, and what we can learn, in turn, from this analysis for an assessment of the uses and limits of this distinction. First, I illustrate that nonstandard uses of experiments and processes of concept formation reveal the "process-interpretation" of the DJ distinction as inappropriate. Turning to what is often regarded as the core of the DJ distinction—the differentiation between genesis and validity—I shall focus on the role it plays within science. Moreover, by taking processes of concept formation seriously, new and hitherto unrecognized limits of justification become visible. In particular, justification turns out to be more genuinely bound to history than is usually assumed. To illustrate and flesh out my general claims, I shall first provide an analysis of a specific historical episode: the purported "discovery" of the two electricities in the 1730s.