1. The sources of error and general significance of the presumptive coliform test have been studied, with special reference to manufactured dairy products.
2. Eosine methylene blue agar has been found reasonably satisfactory for isolation of cultures but unsatisfactory as a direct plating medium.
3. A wide survey of the coliform flora of raw milk and milk products has been made over a period of 11 years. The data were collected from 24,952 samples and include the incidence of presumptive positive tests at each stage of manufacturing processes, the occurrence of ‘false-positive’ tests, the detailed distribution of coliform types in individual dairy products, and some preliminary observations on ‘heat-resistant’ coliform strains.
4. Material examined included raw milk, pasteurized milk, ice cream, milk powder, condensed milk, pasteurized cream, clotted cream, butter, soft cheese, Cheddar cheese, processed cheese, swabs of dairy plant, churn rinses, and the atmosphere of dairy factories.
5. Detailed confirmatory tests were performed on 2490 presumptive positives, from which 2508 coliform cultures were isolated and classified.
6. A general ecological survey of the frequency of individual coliform types in dairying operations as a whole has been attempted.
7. Particular factors which may affect the distribution of coliform types in specific products or situations have been shown to include heat resistance, resistance to drying, chemical composition of the substrate, the effect of storage, and conditions prevailing on plant surfaces. It is suggested that the coliform flora may be partly the result of adaptation to conditions associated with each individual product.
8. The value of the presumptive coliform test has been discussed in respect of public health requirements, as a plant-control method, and in relation to economic aspects. The test is believed to be of great utility in the plant-control laboratory but of little value for public health purposes.