in which AM is the all-important correction term. The specifically nu-clear binding forces make a contribution to AM which is negative and presumably proportional to the number of nucleons; surface-tension effects call for a contribution to AM which is positive and proportional to the nuclear surface area (to A'A; see Eq. 2). The contribution of Coulomb effects is also positive; as Eq. 5 shows, it is closely proportional to Z2A~A; the coefficient, which we will call €, is 3e2/5roc2, or 0.627 mmu.
Apart from all these effects, Eq. 9 must contain in it the fact that, for a given A, there is a particular value of z which corresponds to the most stable nucleus; for light nuclei, this value is A/2. A positive term propor¬tional to (Z — A/2)2/A would seem useful; this term vanishes for z = A/2. Finally, to take account of odd-even effects, we introduce an empirical contribution d, with values shown in Table 4.