We find in the history of ideas mutations which do not seem to correspond to any obvious need, and at first sight appear as mere playful whimsies such as Apollonius’ work on conic sections, or the non-Euclidean geometries, whose practical value became apparent only later.