This work presents a method for measuring the magnetic susceptibility of soil samples based on interactions of magnetic particles contained in a tested sample with a weighed permanent magnet placed in the balance mechanical design. The MYA 2.4Y microbalance manufactured by Radwag Wagi Elektroniczne, Poland, was used to perform mass measurements. The weighing system was adjusted for mass indication using a certified mass standard, and for magnetic susceptibility indication using a certified magnetic susceptibility standard. The volume of each analysed soil sample was 3.93 cm ³ and was similar to the volume and the size of the magnetic susceptibility standard. The research was carried out for 10 soil samples with a magnetic susceptibility range varying from 20 to 1600x10 ^-8 m ³ kg ^-1. The soil samples contained technogenic magnetic particles and particles of natural magnetite of geogenic origin. The study was performed for a field of 2 mT.

The aim of this paper was to investigate the relationship between magnetic susceptibility of topsoil and content of heavy metal being the result of urban and industrial dust-fall. Tools for this study were some complementary statistic methods such as: correlation analysis using Pearson correlation coefficient, Spearman rank correlation coefficient, stepwise regression and .chi-kwadrat" test. The base for statistic analysis was dataset of ca. 600 topsoil samples (20 cm) form Upper Silesian Industrial Region, including content ofAs, Cd, Co, Cr, Cu, Fe, Mn, Ni and Pb as well as values of low-field specific magnetic susceptibility (x) measured for the same samples. The study clearly confirms a significant correlation between the level of inorganic contamination and the measured susceptibility value, although the correlations in soil are usually more sophisticated. The most often observed correlation coefficients between magnetic susceptibility and heavy metals content were on medium (r = 0.5--0.7) and high (r = 0.7--0.9) level. The statistic analysis of the studied parameters can not be based only on Pearson correlation coefficient. The use of some complementary statistic methods allows for more correct interpretation of existing relationships. The comparable values of Pearson linear correlation coefficient and Spearman rank the correlation coefficient, observed in studied dataset within the range of accuracy used, shows the existence of linear correlation. The similar conclusions have been drawn from the analysis of reverse stepwise regression. The observed model of linear multiple regression explains almost 80% of variability of the X value. Foregoing statistical analysis confirms some earlier observations that magnetometry based on topsoil magnetic susceptibility measurement could be a very interesting and alternative or complementary method for monitoring anthropogenic soil pollution and especially heavy metal contamination level.