Figure 23: Behavior of the error in the solution of the differential equation (135), as a function of the parameter entering the numerical scheme (136).