The numerical values given in the above table suggest that (7.18) be better written and used in the form

(7.21)

The Milne formula resulting from or may be written out in terms of ordinates as follows

(7.22)

Equations (7.18), (7.21) and (7.22) represent implicit equations for and are usually solved by iteration, starting with a predicted first approximation . Any explicit formula can be used in principle to calculate . In the interest of economizing the number of corrections, one will choose, if possible, a predictor formula whose accuracy is comparable to that of the corrector formula. Milne suggests the formula

(7.23)

(resulting from integrating the quadratic polynomial interpolating at between and ) as a predictor formula. Now the iteration procedure may be arranged in the same manner as explained in Adams-Moulton method.