A case of time-dependent anisotropy of low-field susceptibility (AMS)

A suite of non-tectonized, Meso-Proterozoic siliceous crystal tuffs and volcanic breccia with a visible stratification has a modal bulk susceptibility (k) ∼ 160 μSI [mean; S.D. = 141 μSI; 44 μSI]. Normally, such susceptibilities suffice to make reliable measurements of anisotropy of low field susceptibility (AMS) using AC induction-coil instruments. However, for this suite, time-dependent susceptibility - variations during measurement are large in comparison to susceptibility - differences along different axes through a specimen. Thus, in many specimens, AMS axes determined by routine induction coil measurement in a (Sapphire Instruments SI2B; 19,200 Hz, 60 A/m) measurement procedures are not reproducible. The apparent variation of specimen susceptibility during a single, four minute AMS measurement can > 2 μSI whereas in the same interval the noise-level of our instrument is < 0.2 μSI. Thus, the time-dependence of the specimen-susceptibility is an intrinsic phenomenon due to the characteristics (grain size, domain structure) of the magnetite which dominates susceptibility during measurement and handling. Two procedures improved the reproducibility and stability of AMS axial orientations in some specimens. First, for some specimens, one or two cycles of LTD or AF demagnetization (≤ 15 mT) stabilized AMS axes. (Previous workers have observed that LTD and AF demagnetization may change slightly the AMS of polydomainal magnetite). Of course, exposure to alternating fields is preferably avoided before any AMS study. Second, for some specimens AMS measurement were improved by shielding the induction coil instrument in a large magnetically shielded space (ambient field < 5 nT). Further improvements were achieved by permitting the specimens to relax in side a magnetic shield for 24 h before measurement. The occurrence of time-dependent bulk susceptibility, especially noticeable in its AMS axial orientations is certainly a rare phenomenon and the procedures we used to improve measurement are excessive and unnecessary safeguards in routine AMS work; in any case, they mostly failed to produce a desirable degree of reproducibility. However, a simple test to detect for the presence of such time-dependent AMS is to repeat measurements at intervals of several hours, between which the specimen is permitted to relax in different orientations with respect to the geomagnetic field, or in a shielded space.