Aspects of Descartes' notion of analysis in his mathematics are looked at in detail with a view to relating this notion to Descartes' work in natural science. Descartes' method in his mathematics and science are subsumed beneath a more general method of analysis termed by Descartes 'indirect problem investigation'. The distinction between Cartesian analysis and synthesis is discussed. In terms of natural science a specific interaction of (1) metaphysically deduced results with (2) empirical data, involving an attendant conversion of secondary qualities to primary qualities for the solution of such problems, is advanced using the magnet as an illustration. A possible relationship with Descartes' methodology in his metaphysics is suggested.