Mineral magnetic studies of archaeological samples : Implications for sample selection for paleointensity determinations

Samples with a significant fraction of multidomain magnetic grains, hard secondary components and thermally unstable magnetic phases have been shown to be unreliable for paleointensity studies. However, mineral magnetic screening is rarely performed before paleointensity determinations are made even through non-ideal magnetic properties are the main reasons for rejecting data after the work has been completed. We have conducted a detailed mineral magnetic investigation of 23 archaeological samples from Bulgaria which yielded both satisfactory and unsatisfactory paleointensity results. Our study demonstrates how the non-ideal magnetic properties lead to unacceptable paleointensity results. We have used our findings to develop a simple and practical sample selection procedure which requires only two specimens from each sample and which can be done with conventional paleomagnetic equipment. We suggest that any sample which fails to pass this screening should not be subjected to time-consuming Thellier-Thellier experiments.