One of the most interesting developments in this application of magnetic fabric was a new understanding of the microscopic mechanism that leads to magnetic fabrics caused by igneous flow. Hargraves et al. [1991] pointed out that although magnetic fabric in igneous rocks has been shown to accurately mimic non-magnetic indicators of flow when the rocks were examined microscopically the magnetic grains carrying the AMS do not usually show shape anisotropy. The typical magnetic grains in an igneous rock are either equidimensional or skeletal in shape and should not produce the strong magnetic anisotropy usually observed. Hargraves et al. [1991] suggested that it was not the preferred orientation of anisotropic magnetic grains that leads to magnetic fabric in igneous rocks, but rather an anisotropic distribution of the isotropic magnetic grains which causes them to interact magnetically. Moreover this anisotropic distribution of magnetic grains is imposed when the grains grow at late stages in the igneous process in an already crystallized template of silicate grains. Hargraves et al. supported this model with experiments in which rocks were sliced and glass slides were inserted between the rock slices. This led to the development of an AMS fabric in which the minimum axis was oriented perpendicular to the rock/glass slices. Stephenson [1994] presented some simple theoretical models which supported the Hargraves et al. [1991] model of distribution anisotropy. In these simple models, lines or planes of magnetically-interacting equidimensional magnetic particles were shown to produce susceptibility anisotropy. Unfortunately, Stephenson's models also indicate that susceptibility measurements alone will be unable to distinguish between distributional anisotropies and anisotropies arising from preferred orientations of anisotropic grains.
Staudigel et al. [1992] used the AMS of sheeted dikes in the Troodos ophiolite to make an important contribution to our understanding of how oceanic crust is emplaced at a mid-oceanic spreading center. Previous models of oceanic crust emplacement postulated only vertical flow in the sheeted dikes found in the seafloor. Staudigel et al.'s AMS measurements suggested significant lateral flow in the Troodos sheeted dikes, thus leading to a reexamination of our ideas about mid-ocean ridge igneous plumbing systems. Progress was made in distinguishing primary flow fabrics from later tectonic fabrics in Precambrian rocks [ Cadman et al., 1992] or determining the regional flow patterns in the Mackenzie radiating dike swarm [ Ernst and Barager, 1992].
Work continued this quadrennium on the flow fabrics of extrusive igneous rocks, primarily ash fall tuffs. Hillhouse and Wells [1991] completed a comprehensive magnetic fabric study of the Peach Springs tuff in the southwestern United States. They used AMS measurements of flow azimuths in the tuff to find the location of the now-eroded source area for the regionally extensive rhyolitic ash fall tuff. Hillhouse and Wells found flow lineations in 30 of the 42 sites they measured which radiated away from the center of the Peach Springs Tuff outcrop area. They also found a flow foliation which was imbricated and dipped away from the distal margins of the Peach Springs Tuff exposure area and towards the postulated source area. This result adds to accumulating evidence that flow imbrication of the magnetic fabric can be a useful way to determine the actual direction of flow in an igneous rock. Palmer et al. [1991] also used AMS to find source areas for ash fall tuffs as part of a larger paleomagnetic study of mid-Tertiary volcanic rocks in northeastern Nevada. Maximum axes of the AMS ellipsoid could be used as a measure of flow azimuth in these tuffs. In addition, the removal of unusual patterns in the distribution of the AMS minimum axes was used to support tilt corrections applied to the magnetic data.
The effect of magnetic mineral alteration on magnetic fabric was the focus of Seaman et al.'s [1991] study of the Bloodgood Canyon and Shelley Peak Tuffs of southwestern New Mexico. They found that alteration of primary magnetite, biotite and other Fe-bearing minerals to secondary hematite caused a degradation in the AMS flow lineations. The altered tuffs had much lower bulk susceptibilities and scattered within-site AMS results than the less altered tuffs. Consistent igneous flow lineations were extracted from the data after it had been categorized following the scheme proposed by Knight et al. [1986] in which sites are classified based on the within-site dispersion of the principal susceptibility axes. In another study of flow lineations in the Bloodgood Canyon Tuff, Seaman and Williams [1992] show that the center-to-center method of strain analysis [ Fry, 1979] can be used to estimate flow lineations in ash-flow tuffs. Flow lineations determined by center-to-center measurements agree well with AMS flow lineations.
In the only volcanic flow fabric study this past quadrennium on non-tuff aceous extrusive rocks MacDonald et al. [1992] investigated the possibility of using AMS to determine flow in a trachydacitic lava in the Egan Range volcanic complex of Nevada. Locally AMS lineations agreed with petrologic indicators of flow, but the flow patterns were so complex in this viscous magma that it was not possible to detect the location of the volcanic source area.