It is shown that the concurrence of the results obtained by varying the self coupling energy, the total second order energy and the cross coupling energy independently can be readily accomplished. In addition, it is possible to arrange to obtain any pre-specified value for the coupling constant. On this basis it is suggested that it might be dangerous to construe the concurrence of calculated results as an indication of convergence in practice.Calculations of the spin-spin coupling constant of the HD molecule have been performed by many authors [1][2][3][4][5] in many different ways. This note deals with the class of calculations involving a less singular contact operator than the delta function. With such an operator and a suitable choice of variational functions it becomes possible to make the cross-coupling energy independent (to within a very small parameter) of the (finite) self-coupling energy [5], the value of which depends on the precise form of the contact operator [1][2][3][4][5]. Delpuech [6] et al. concluded from their study that it is necessary to use variational functions which mimic the true solution both in the immediate vicinity (that is, within a few Fermis) of the perturbing nucleus and in the inter-nuclear domain in order to get convergence. Sänger and Voitlander [2] using a non-singular contact operator did not directly address themselves to the question of convergence but they computed the coupling constant JHD in three different ways -by varying the self coupling energy, the total second order energy and the cross coupling energy -and obtained the same result, 39 Hz, in each case. The question poses itself whether this can be taken as an indication of convergence. In what follows we argue that it would be dangerous to do so. We demonstrate that a class of variational functions can be chosen for which this concurrence can be guaranteed and for which it is possible to obtain any pre-specified value for J HDAfter integrating out the spin the original two electron problems can be reduced to a one electron Reprint request to the author at Physical Chemistry Laboratory, South Parks Road, Oxford.' 0X1 3QZ, England. problem. Having postulated a short range and a long range part for the solution and observed that the cross coupling energy is independent of the self coupling energy [2,5] Sänger and Voitländer [2] point out that all three expressions, i.e., (1), (3) and (5), are equivalent if Hi,i. (5)