Abstract
By assuming a discontinuity in the spin direction at a defect, Arrott showed that for high density of defects, the approach to magnetic saturation should contain a parasitic paramagnetism term, as is observed experimentally. It is shown here that the artificial assumption of a discontinuity is not necessary to get these results. Using the same one-dimensional model of Arrott and the usual theory for the magnetic hardness term, that implies unsmooth but continuous functions, the parasitic paramagnetism term is also obtained for the same limit of high density of defects, in regions where the nearest neighbors to each defect are defects of opposite sign.