A statistical analysis is presented of published laboratory data on 4407 soil horizon samples collected by the Soil Survey of England and Wales during 1950–73, to typify soil taxa. Seventy-five percent of the profiles have fine textured parent materials in both calcareous and non-calcareous categories. Moderate to large increases in clay content down the profile, which could indicate argillio horizons, are found in 56% of the profiles but the influence of lithological discontinuities has not been eliminated. Argillic horizons are commonest on calcareous or low grade metamorphic rocks and non-calcareous clayey or silty sediments. Less than half the average frequency of clay increase between middle and bottom horizons is shown in the igneous and high grade metamorphic lithological type.

pH gradient down the profile is not clearly related to altitude. The pH change (0·7) between top and bottom horizons in profiles from calcareous and igneous parent materials is twice as great as in other non-calcareous profiles. There is usually less clay in the top horizons than in others; the top horizon contains more than 35% clay about half as frequently as the lower horizons. Clay content categories have closely similar frequency in middle and bottom horizons. Gleying in horizon 1 is observed somewhat less frequently with increasing CaCO3. No consistent trend is found in the other horizons. Soil pH in 0·01 M-CaCl2 (pH8) is between 5·1 and 7·0 in 51% of all samples. pH5, values less than 4 occur in only 8% and are twice as common in top horizons (more than 5 cm thick) as in deeper ones. pH8, values greater than 7 are nearly twice as frequent (31%) in the bottom horizon as in the upper ones. Correlation of pH in water (pHw) with pH8 always exceeds 0·93 irrespective of lithological type. The difference of pHw from pH8 (δpH) calculated from the regression equation is 0·7–0·8 at pH6 values less than 4; it varies inversely with pH8, up to a value of 0·4–0·5 at pH6 greater than 7. Neutrality (pHw, 7) corresponds to pH8 6·5–6·6. There is no indication that δpH is related to altitude or gleying intensity but it varies according to lithological type by about 0·1 unit. Organic-matter content, as indicated by loss on ignition (950 °C) minus 10% of the measured clay percentage (LOIadj%) is little affected by lithological type; decrease with depth is steeper at pH8 less than 4 than at higher pH. Gleyed top horizons contain more organic matter than ungleyed; gradient down the profile is positively related to gleying intensity.

The overall multiple regression of cation exchange capacity (CEC) on clay percentage and organic matter (LOIadj%) is (CEC = 1·21 + 0·40 clay% + 1·32 LOIadj), accounting for 61·4% of the total variance. When lithological types are considered separately the partial regression coefficients for clay percentages are greater in the two calcareous lithological types than in the non-calcareous. With lithological type 5 the large constant term and small coefficient of clay percentage indicate considerable contribution to CEC from particles or aggregates larger than 2 μm. Partial regression coefficients (lithological types 1–4 taken together), although similar both within horizon and within gley category are nevertheless shown to be significantly different. Within pH8 category, both partial regression coefficients increase directly with pH. Calculation of mean values of CEC on a clay basis for horizon 3, where contribution from organic matter is minimal, does not show significant differences between pH8 categories or gley categories.

Although quantification of relationships has been possible, correlations are rarely high, so that predictive power is limited. Availability of soil group or subgroup categorizations using constant definitions based on objective measurements is suggested as a useful way of increasing predictive power within resulting smaller groups. The calculation of geographic (areal) distribution is probably the most useful purpose of a data bank of routine measurements, but the selection of samples to typify soil taxa used in this study would preclude such a use.