An Ecological Investigation on the Usefulness of Two Specific Modifications of Eijkman's Test as One of the Methods Used for the Detection of Faecal Contamination of Foods

Summary An investigation has been carried out on the suitability of two modifications of Eijkman's heat tolerance test which had previously been found useful for the detection of Escherichia coli strains of faecal origin in foods. These were: plating on Chapman's lactose tergitol‐triphenyltetrazolium agar incubated at 44°, and enrichment in brilliant green‐bile fluid incubated at 44°, followed by subculture in the same medium and in tryptone water to test indole formation. Two materials having quite different histories of faecal contamination were examined. These were ‘coliform positive’ ice cream samples and bulk raw human milk. The reliability of both methods was found to be very satisfactory. In spite of this, such Eijkman tests appeared to have only limited usefulness because E. coli was found to be readily outgrown by other bacteria under extraenteral conditions; E. coli was therefore absent relatively often even from 1 ml samples of human milk which contained quite high numbers (i.e. > 10 3 /ml) of other organisms of faecal origin. It is recommended, therefore, that too much reliance should not be placed on the outcome of Eijkman tests, especially when they give a negative result. The hygienic evaluation of foods should rather be based on the results of the simultaneous enumeration of three groups of organisms, namely E. coli , faecal types of Klebsiella (i.e. those which are urease positive) and Lancefield group D streptococci. A striking result of this investigation was the occurrence in 69% of the human milk samples of 44° positive Achr. anitratus (B5W) in numbers from 10–5.6 times 10 3 /ml. These bacteria were almost invariably found in samples which were E. coli negative; klebsiellae, many of which were Kb. oxytoca , were also found in such samples. The properties of these organisms and their ecological significance are discussed.