Humic substances (HS) extracted from 14 Sphagnum peats of various geographical origin were submitted to capillary electrophoresis (CE) in free solution and in entangled polymer solutions (25 g l-1 and 50 g l-1) of polyethylene glycol (PEG). Electrophoretic runs were made in uncoated capillaries (75 μm internal diameter) using a tris-hydroxymethylaminomethane (TRIS)-phosphate buffer (pH8.3) eventually containing PEG and applying a voltage of 10 kV. The HS extract of a Lithuanian peat was fractionated by ultrafiltration into five fractions with a nominal mean relative molecular mass (M) of 3000, 7500, 20 000, 75 000 and 200 000 Da. Without PEG, migration times were not correlated with the logarithm of M, but linear relations were obtained with 25 g l-1 and 50 g l-1 PEG solutions. Both solutions separated the different fractions by size; the best regression and the largest interval of linearity, M from 3000 to 200 000Da, was obtained with the 25 g l-1 PEG solution. At 10 kV and without PEG, migration times ranged from 6.5 to 8.1 min and did not differentiate the peats, whereas at 25 g l-1 PEG, migration times ranged from 3.3 to 6.5 min and, when plotted as a function of either the r value or the von Post index (H), which are used to evaluate the degree of decomposition of horticultural peats, they separated samples on the basis of their geographical origin and probable type of mire complex. At 25 g l-1 PEG, migration times also showed a well-defined linear decreasing trend with the increase of mean annual temperatures. These relations suggest the hypothesis that humification in mires goes along with an increase in the average size of humic molecules.
