There have been several experiments on the precession of muons in iron where the precession frequency is not understood on the basis of the model of a muon diffusing among the interstitial sites averaging out the dipolar fields that are quite different on the sites, depending on their relation to the magnetization direction. In two new experiments, one at 300 K and one at 1040 K, the absence of the muon precession signal appears difficult to explain. At 300 K a single-crystal plate of iron is placed in a large external field parallel to the polarization and direction of propagation of the muon beam. The crystal is set at angle with respect to the beam in such a way that the component of Bμ, the magnetic induction seen by the muon, along the applied field should vanish, leaving a transverse component about which the muon should precess. The calculation of the fields and angles for this uses the known values of the anisotropy constant, the saturation magnetization, and the hyperfine coupling constant. The purpose of the experiment was to determine the anisotropy of the hyperfine constant by observation of the variation of the minimum precession frequency with angle. The experiment is carried out using a pulsed muon beam in applied longitudinal fields comparable to 4πM. The precession signal is not observed. At 1040 K a polycrystalline iron oblate ellipsoid of revolution is placed in longitudinal and transverse applied fields with the principal axis set at various angles with respect to the muon beam. Though the demagnetizing field of the rotated ellipsoid in the longitudinal field should produce a transverse component of Bμ, the expected precession signal is not observed. The application of a transverse field does produce the expected precession signal. The effects of the applied fields on the muon depolarization due to critical fluctuations are observed.