One hundred thirty‐nine soil samples from 20 states were analyzed for 15N abundance. Soil characteristics and environmental conditions at the sampling sites varied widely. The total N of surface soil samples had a mean δ15N value (per mill 15N excess) slightly but significantly higher than the mean value for soils collected from deeper layers, although the relationship between δ15N and soil depth was not consistent from location to location. Differences among mean δ15N values for the total N of soil samples collected from cropland, pasture, uncultivated land (with native herbs), and forest were not striking. There was no systematic effect of variation in the rate of application of N fertilizer on the δ15N of the total N of soils. The mean δ15N value of surface soils, with respect to atmospheric N, was +9.22 and the standard deviation was 2.10 δ15N units. The range within which 90% of the samples fell was +5.1 to +12.3 δ15N units. Less than half of the variation among the soils in δ15N of the total N could be accounted for by differences in environmental variables or soil characteristics.

Changes in the isotopic composition of added NO3‐, during the course of its disappearance from two central Illinois soils are reported. The only known difference in the two soils is the cultural history of the fields from which they were taken. A large (approximately three‐fold) difference between the two soils was observed in the isotopic fractionation factor associated with NO3‐ loss. Large differences in the two soils with respect to both the rate of NO3‐ loss and the isotopic fractionation associated with it were apparently related to differences in the two fields rather than to temporal or spatial variation within the fields. Addition of glucose (1% with respect to dry soil weight) caused the difference between the two soils to disappear. It also appreciably reduced the apparent isotopic fraction factor in both soils.When NO2‐, rather than NO3‐, was added as substrate, the isotopic fractionation factor associated with its loss was the same in the two soils. The effect of glucose was similar whether NO3‐ or NO2‐ was the substrate; i.e., the isotopic fractionation factor was substantially reduced.Possible explanations for the observed variations in isotope effect associated with substrate loss are discussed. In addition, the implications of these variations for the study of denitrification in natural systems are discussed.

The loss of NO3‐ added to two central Illinois soils was determined in experiments in which the soils were incubated under waterlogged conditions. The loss was measured as a function of substrate concentration in one experiment in which samples were incubated for a single time (24 hours) and as a function of time in a second experiment in which the concentration of the added NO3‐ was held constant (200 ppm NO3‐‐N). The rate of loss of NO3‐ was about 5.5 times higher in one of the soils than in the other. This difference was largely overcome by the addition of glucose (1% with respect to the soil) which also greatly enhanced the rate of denitrification in both soils.The experimental points representing the rate of NO3‐ loss plotted as a function of the concentration of added NO3‐, were equally well fit by Michaelis‐Menten and exponential equations as well as by the solution to a pair of nonlinear differential equations representing a system in which the product of one reaction (e.g., the reductant generated by the oxidation of carbon compounds) is a substrate in a second sequence (e.g., the denitrification of NO3‐). The significance of such fits is discussed. The authors point out that while such fits have certain uses, it is not possible to infer from them the mechanism of the reaction.