Since the sample mount is under the influence of the same fields as the sample, there will be an additional force acting upon it according to its magnetic moment, indistinguishable from those of interest. Obviously the lower the moment is, the
better, so that large background contributions are not present. The background should be repeatable, to the resolution required to extract the signal due to a sample, and ideally linear with field, and temperature. For this reason, ferromagnetic materials are unsuitable.