(1) Models of competition in divided or heterogeneous environments are reviewed. (2) A model of competition in a divided environment is presented. The competition coefficients, a, are reduced by a term, 0, which measures the amount of overlap between species. (3) Actual values of 0 are presented for two Drosophila communities. (4) 0 depends on the degree of aggregation of the competing species and on their densities. The outcome of competition in a divided environment must, then, be density dependent. (5) Low values of 0 improve the stability of a multi-species community and allow more species to coexist.

1. The distribution and abundance of five Drosophila species breeding in fungi are examined by attracting adult flies t o baited traps and collecting fungal fruit bodies containing the larval stages.2. Changes in species frequency follow a temporal (seasonal and yearly), rather than a spatial (between and within woods) pattern.3. There is a regular diurnal pattern of activity in Dmbobscura but not in D.phalerata. Temperature, light intensity, humidity and wind speed have no effect upon the daily activity of D.phalerata. Only temperature affects the activity of D.subobscura.4. Out of 125 species of fungi collected, only forty-eight species produced Drosophila. However, the flies are even more selective, since only five species of fungi (Phallus impudicus, Polyporus squamosus, Amanita rubescens, Pluteus cervinus and Lactarius quietus) produce 80% of all Drosophila reared.5. The commonest species, D.phalerata, has two or three generations per year with a winter diapause. An outline of the yearly life cycle is given, with the major breeding sites for each generation indicated.6. Measures of niche overlap calculated from the data on individual fruit bodies are very low, suggesting that the five species are effectively isolated during the larval period. This ecological isolation is achieved by a summation of several niche dimensions, woodlands, seasons, fungal species, and individual fruiting body differences.

1. Seasonal reproductive development was investigated in four obligate fungivorous species, D.phalerata Meigen, D. transversa Fallen, D.cameraria Haliday and D m n f u s a Staeger. 2. Samples of females taken from woodlands over a 2 year period were scored for reproductive status and fat content. 3. Populations of D.phalerata and D.cameraria maintained in an outdoor insectary provided information on developmental time and adult mortality throughout the year. 4. Experiments in controlled environmental chambers gave percentage diapause curves against day length for all four species. .In D.phalerata the influence of temperature and fungus upon diapause behaviour was also investigated. 6. In D.phalerata the optimal strategy for reproductive development can vary from month to month. Maturation of females may be immediate, dependent upon a fungal cue or completely prevented.

Emergence records for three species of Drosophila breeding in the stinkhorn (Phallus impudicus) are presented. These are D. phalerata, D. subobscura and D. cameraria. The numbers of stinkhorn present in one woodland (0.114 km2) have been estimated for a two‐year period. In 1976, the total number of these fungal breeding sites during the summer was approximately 8000 and in 1977 it was 5300. Utilizing the estimates of breeding site numbers, Drosophila emergence data and estimates of adult survival from laboratory populations, density estimates for D. phalerata and D. subobscura are calculated.