1. Twenty stations were established in a grass field of 8 acres, and twenty-nine soil samples were examined from each station in the course of 2⅓ years. Wireworms (Agriotes sputator) were distributed among the twenty stations in a non-random manner.
2. In this field, the wireworm distribution was correlated to a highly significant degree with nine factors (loss on ignition, organic carbon, nitrogen, Lolium, Agrostis, Chilopoda, staphylinid larvae, ants and Nematocera) and to a significant degree with three factors (altitude, pH of the soil and Diplopoda).
3. Among these factors, three (the organic content of the soil as measured by loss on ignition, the prevalence of the grass Lolium, and the numbers of staphylinid larvae) were found by analysis to be most significantly correlated with the wireworm distribution.
4. Sixteen stations were set up in a grass plot of ¼ acre, and twenty-one soil samples were examined from each station during a period of 3 years. The distribution of wireworms at these stations was markedly non-random throughout the sampling period, some stations being consistently three or four times as heavily infested as others only 5 yd away.
5. Within this plot the distribution of wireworms was correlated to a highly significant degree with altitude, depth of loam and amount of soil moisture at 3 and at 9 in depth. It was significantly correlated with the lime content of the soil. Ranunculus was prevalent on an area of low wireworm infestation.
6. Among the factors studied in this plot, the amount of soil moisture at 3 in. depth was most important. It was negatively correlated with the distribution of wireworms.
7. Within the plot, stations having high wireworm populations were infested almost exclusively by Agriotes sputator; those having low wireworm populations were infested about equally by A. sputator and A. ‘obscurus-lineatus’.
8. Four square-yard samples were examined, each in eighty-one quadrats 4 in. square. Within each square yard, the wireworms were not distributed at random and independently, but were markedly aggregated.
9. Separation of the wireworms collected from these samples into three size groups showed that the non-random distribution in square-yard samples was largely due to the small larvae, which were strongly aggregated . The medium-sized larvae were less markedly aggregated. The distribution of the large larvae under grass was not significantly different from a random distribution.
10. To understand the factors controlling wireworm distribution it will be necessary to consider the distribution of the different larval age-groups.
11. Two suggestions are made towards an explanation of the geographical distribution of wireworms in England.
12. Study of the spatial distribution of wireworms raises some of the major problems of ecology and, in particular, illustrates the interdependence of spatial distribution and seasonal succession.