them in his cheat-preface to Copernicus De Revolutionibus, but the main change in their import has been that whereas Osiander defended Copernicus, Mach and Duhem defended science. The modem conception of hypothetico­ deductive science is, again, geared to defend the respectability of science in much the same way: the physical interpretation, it says, is merely and always hypothetical, and so the scientist is never really committed to it. Hence, when science sheds the physical interpretation off its mathematical skeleton as time and refutation catch up with it, the scientist is not really caught in error, for he never was committed to this interpretation in the first place. This is the apologetic essence of present day, Popper-like, versions of the idea of science as a mathematical-core-cum-interpretational shell. This is also Cohen's view, for it aims to free Newton of any existential commitment to which his theory might allegedly commit him. It will be readily seen that Cohen regards this methodological distinction between mathematics and physics to be the backbone of the Newtonian revolution in science (which is, in its tum, the climax of the whole Scientific Revolution) for a very clear reason: it enables us to argue that Newton could use freely the new concept of centripetal force, even though he did not be­ lieve in physical action at a distance and could not conceive how such a force could act to produce its effects". ([3] pp.