1. The presumptive coliform test was carried out by each of three workers, at each of four dilutions, on seventeen sets of five tubes for each of seven samples of 18-hour-old afternoon milk, held overnight in the ice chest.2. With the exception of three slightly anomalous results, the data are consistent with the hypothesis that the chance of a tube remaining sterile is constant for all tubes inoculated from the same sample of milk at any one dilution.3. The assumption of a Poisson distribution of organisms in parallel tubes is accurate enough, with these data, to give a good idea of the order of magnitude of the number of bacteria per ml.4. There is, however, a tendency, more marked in our data in autumn than in spring, for the estimated count to rise at the higher dilutions. This may plausibly be attributed to the breaking up of clumps.

No abstract

1. The experiments described in this part of the report were designed to obtain information about the effect of time and temperature of pre-test storage on test results and, in particular, about the possibility of a practical method of compensating test results to eliminate or minimize their dependence on the time and temperature of pre-test storage.2. Four separate experiments were planned and carried out in three widely separated areas in England and Wales. Samples of morning milk, brought to the laboratory with the minimum of delay, were divided into a series of subsamples which were subsequently stored at controlled temperatures ranging from 3 to 24°C, and at shade atmospheric temperature, for periods ranging from 7 to 30 hr. before examination by one or more of the following tests: keeping quality at 18, 22 and 24° C. using clot-on-boiling (C.O.B.), and precipitation with 68% alcohol (A.P.T.) or 50% alcohol; methylene-blue test at 37–5° C. and at 18, 22 or 24° C. The results for the different tests were the times to end-point after transfer from the storage to the incubation baths.

Summary: The flaming of taps before sampling caused no significant differences in the most probable numbers of coli‐aerogenes organisms or Bact. coli in water samples, or in the colony counts at 37 or 22°, though there was a tendency for the bacterial content to be lowered.

With 1 Figure)SYNOPSIS 1 ntroduGtion Reasons for and objects of the experiment-standards of technique-lay-out of experiments Statistical treatmentChoice of unit for keeping quality and methylene-blue tests and for plate counts-reasons for using log of plate count Experiment I Object (general comparison)-statistical treatment-comparison of grading-correlationseffect of winter weather on methylene-blue test-effect of cold storage of sample-resultslinearity-effect of fat content-of coliform content-summary Experiment II Object (reproducibility between laboratories)-tests employed-results of fat tests-plate counts-keeping quality-reduction time at 15-5° C.-reduction time at 37° C.-summary Experiment III Object (relationship to production conditions)-score card-results-effect of seasonsummary Sensitivity ExperimentObject-lay-out-general relationships-comparative variability of plate count and methylene-blue test-variability and level of mean-significant difference between two samplesgrouping error-sensitivity tests-accuracy and sensitivity-regression equations and linearity-corresponding values of plate count and reduction time-sensitivity by regression-' t' tests, on groups of ten results and on fifteen random selected results-conclusions -summary Adde)idum I. Normality of distribution of replicates Addendum II. x 2 values Addendum III. Comparative variability of plain agar and milk agar-incubation at 32 and 37° C. Addendum IV. Method of assessing reported count Addendum V.