In this chapter I consider Newton's use of the Baconian label experimentum crucis in his famous 1672 paper on Light and Colors. I take it to be a sort of `signpost', or methodological clue, which, properly understood, can help us to assess the kind of `empiricist' commitment that may be ascribed to Newton. In order to dispel persistent misunderstandings, the first part of the chapter shows how our present understanding of crucial experiments has been shaped by nineteenth-century philosophers of science, as part of a characteristically anti-Baconian methodology. It also shed light on the particularly ambiguous character of Newton's methodological legacy. In the remaining parts of the paper, muting these retrospective overtones, I show what crucial instances, or crucial experiments, really meant for Bacon, and for his heirs, in order to elucidate Newton's usage in the light of its actual context (chiefly, the one provided by Bacon, Descartes and Hooke). I argue that the crucial experiment in Newton's 1672 paper is not devised to arbitrate between competing full-blown and empirically grounded theories, but rather to make the abstract structure of a new, and still unclear, phenomenon (namely: dispersion) entirely transparent. It functions as a `perceptual' index rather than a judgmental criterion. Finally, I explain why the expression has been dropped in the published text of the 1704 Opticks. The `exhibiting' function of the experimentum crucis is now taken over by a long and complex pattern of experiments. It is the whole complex composition of these optical experiments that is now `crucially' exhibiting the abstract structure of refraction.