Susceptibility, thermo-remanent magnetization (TRM) and isothermal remanent magnetization (IRM) anisotropy ellipsoids have been determined for several rock samples. The results indicate that the ellipsoid of initial susceptibility is less anisotropic than the TRM and low field IRM ellipsoids which are found experimentally t o be of identical shape. This suggests that palaeomagnetic data for anisotropic rocks may be corrected by using the anisotropy ellipsoid determined from magnetically non-destructive low field IRM measurements. Such IRM measurements can also be used t o obtain anisotropy axes of samples which are inherently anisotropic but which have a susceptibility which is too weak to be accurately measured. The results for a series of artificial anisotropic samples containing magnetite particles of different skes (in the range 0.2-90pm) were very similar to those for the rocks. In contrast, a comparison of the susceptibility and IRM ellipsoids for anisotropic samples containing particles from a magnetic tape gave very different results in accordance with theory. Such results imply that susceptibility and IRM ellipsoids could be used to determine whether anisotropic rocks contain uniaxial single-domain particles (magnetization confined to the easy axis) or whether the particles are essentially multidomain.

No abstract

Magnetic susceptibility anisotropy is commonly used as a guide for determining stress and flow axes in rocks, and for assessing the suitability of anisotropic rocks for palaeomagnetic purposes. We have recently demonstrated however in laboratory produced samples that, as theoretically expected, susceptibility anisotropy is dependent on particle size. Multidomain particles of magnetite, or similar strongly magnetic particles with shape anisotropy, exhibit a maximum susceptibility parallel to their easy (remanent) magnetization axes whereas uniaxial single‐domain particles exhibit a maximum susceptibility perpendicular to their easy magnetization axes. Hence susceptibility anisotropy does not yield an unambiguous determination of particle alignment in rocks independent of domain state. By demonstrating uniaxial single‐domain anisotropy characteristics for the first time in a rock specimen, we show here that susceptibility measurements by themselves may not correctly identify the degree or type of alignment of ferro‐ or ferrimagnetic particles in rocks. In particular, apparently foliated rocks can actually be lineated (and vice‐versa). Moreover rocks with low anisotropy of susceptibility can have a much higher anisotropy of remanence. Remanence anisotropy methods are shown to give unambiguous determinations of the degree and type of alignment.

The application of alternating (AF) or direct (DF) magnetic fields to samples containing dispersed particles of magnetite, titanomagnetite, or maghemite has been found to alter significantly the measured low‐field susceptibility anisotropy even though the particle orientation is unchanged. The acquisition of isothermal remanent magnetization (IRM) is also strongly dependent on the previous AF treatment given to the samples. A method of quantifying the changes brought about by application of the fields is described. The samples studied were intrinsically isotropic or only weakly anisotropic, and the computed impressed susceptibility or IRM ellipsoids produced by the applied fields were ellipsoids of revolution with the unique axis aligned with the applied field axis. The size and shape of the field‐impressed ellipsoids were strongly dependent upon particle size. These effects might thus form the basis of a rapid, nondestructive means of estimating particle size in rocks. The results also have important implications for magnetic fabric analysis and the static AF demagnetization of rocks. In particular, to ensure that the intrinsic magnetic fabric is measured without any field‐impressed component of anisotropy, samples should not be subjected to static AF demagnetization prior to the determination of either low field susceptibility anisotropy or IRM anisotropy. Samples should also not carry a significant remanence (such as an IRM acquired in a field greater than about 5 mT) prior to low field susceptibility anisotropy analysis.

Summary The dependence of anhysteretic remanent magnetization (ARM) and rotational remanent magnetization (RRM) on particle size has been determined for magnetite particles covering the approximate size range 0.2 to 90 μm. Both types of remanence depend on particle size with the smallest particles acquiring the highest remanence. For comparison a sample containing γFe2O3 particles from magnetic recording tape has also been included. By measuring the ratio RRM/ARM and multiplying this by the steady field used to produce the ARM, an effective field, Bg, which can be considered to produce the RRM, can be obtained. Bg ranges from 300 to 3 μT for particles in the range 0.2 to 45 μm at a rotation rate of twice the alternating field frequency (in a peak field of 80 mT) and varies approximately as the inverse of the particle size. The RRM under the same conditions varies as the inverse 1.3 power of particle size. A determination of the RRM and effective field Bg for rocks, therefore, offers a potentially quick, non‐destructive method of obtaining an estimate of the average size and approximate concentration of any magnetite particles which may be present. This method, however, must be used with some caution since the role (if any) played by small parasitic particles adhering to their larger host particles, which appear in electron micrographs of the samples, is at present unclear. RRM, however, at the very least, provides a quick method of detecting the presence of fine particles < 1 μm in size.