Soil suspensions are usually characterised by the percentage weights of particles settling with velocities between defined limits, such as silt and clay. If percentages were measured sufficiently accurately by both the pipette method and a plummet balance, dry densities of fractions could be calculated and hence their composition inferred. The accuracy and precision of the pipette method were confirmed by the mass balance of particles in peptised suspensions of a red-brown earth subsoil and a kaolinite. Percentages indicated by a plummet balance were corrected for the actual densities of fractions measured separately. Residual differences from the pipette values showed that the balance was accurate but lacked the precision needed to deduce densities to 0.01 mg/mm³.
Values of the density of clay fractions were deduced from the means of published percentages obtained by the two methods where replicate suspensions were used and also from the mean difference based on a comparison using 199 soil samples. For the latter, an underestimate of clay percentages indicated by the balance had first to be allowed for. Clay-deduced densities were in the range 2.80–2.83 mg/mm³, compared with 2.84 mg/mm³ measured on the separated subsoil clay. The values are similar, probably because the clays are derived from mica. The density of the separated subsoil silt was 2.64 mg/mm³, less than quartz, because of the presence of feldspar. Such results indicate the potential use of the method if the precision of the balance is improved.
