A recent neutron interferometry experiment has been interpreted as an explicit verification of the uncertainty principle. Here the question is discussed which mathematical expression of the uncertainty principle is relevant to this experiment. It is shown that the experiment does not measure the standard deviation σq in position independently of ad hoc assumptions on the shape of the wavefunction. Rather, it is the "mean peak width" that is measured directly. Therefore, the natural interpretation of this experiment is not as a verification of the Heisenberg relation σpσq ≥ 1 2; but rather of a relation between mean peak width in position and the overall spread in momentum.