An assessment of the numbers of flies emerging from a bacteria bed has been almost continuous through nearly 8 years, and an attempt is made to explain the reasons for the specific fluctuations. These are found to depend on weather influences reacting on Metriocnemus longitarsus and the worm Lumbricillus lineatus, two forms which dominate the upper, region of the bed, the former being potentially predaceous.

The annual cycles of M. longitarsus are so timed that there is a major flight in May establishing its offspring in June. If June is cold and wet this form is successful and a barrage of larvae is formed in which much oviposition of the other species takes place. The eggs are largely destroyed in the barrage. Hot dry Junes are unfavourable to establishment and the barrage is thinned, or it may be dispersed by subsequent dry conditions. M. hirticollis has a major flight about a month later and so is dependent for summer establishment on a dearth or dispersal of M. longitarsus larvae. Spaniotoma minima is similarly affected, but the effect is less lasting because its cycles are rapid and it has the capacity for recovery in the cooler months which leads to a reasonable flight in May. The thermal requirements of M. hirticollis do not allow of this except in a very mild winter, and consequently it cannot recover from collapse until a season recurs in which hot dry conditions cause failure or dispersal of the barrage. The distribution of weather was such that gradations with a 4-year interval were set up in the annual incidence of these flies.

When the larvae of S. minima (and possibly, but not distinctively, those of M. hirticollis) are reduced by abundance of M. longitarsus larvae, pressure in the depths of the bed is relieved of a proportion of forms also potentially predaceous. By this the psychodids, which probably always oviposit to some extent in the depths, are benefited and are able to extend their seasons of abundance later in the year than usual.

The numbers of S. minima and M. hirticollis emerging in the early part of the year are largely determined by their autumn abundance, so the fluctuations of the whole year are referable to interspecific competition regulated by summer weather. This is not so for the other common flies which are markedly influenced by events in winter as well as in summer. There is a tendency for Psychoda alternata to be more abundant after a warm winter because its physiological requirements are a paramount factor, but some evidence is found that extra pressure of the other larvae may reduce it then as they do later in the year. Conversely the numbers of M. longitarsus, and still more of P. seoerini, tend to be less after a warm winter in spite of the fact that they are essentially winter-breeding insects. This is attributed to the intervention of L. lineatus in competition because the worms breeds mainly in the winter months and its thermal requirements are such that a rise from 7 to 10d̀ C. (about the limits of cold and warm in the beds in winter) causes a great acceleration in its cycle, while cold conditions drive it down from the surface Phormidium which it almost completely destroys in spring. Therefore it is suggested that the time of sloughing will be earlier or later in spring according to the winter-bed temperature. This hastens or retards the period of deprivation for the larvae and so regulates distinctively the prosperity of the flies which breed briskly in winter and early spring. As assessments of L. lineatus were not made throughout the period these findings are theoretical. It is thus shown that the annual fluctuations of these forms are partly dependent on weather conditions and partly dependent on interspecific competition regulated by the weather. In several cases competition is shown to be the more potent direct influence.