A Comparison of Standard Plate Counts and Methylene Blue Reduction Tests Made on Raw Milk with Special Reference to Geometric Means

Summary 1. The analysis of data of standard plate counts and methylene blue reduction tests of approximately 1100 samples of raw milk is given. 2. The range and frequencies of counts are given for each class as well as the geometric mean and standard deviation. 3. Practically two-thirds of the counts in each class, as determined by the reduction test, were within one standard deviation of the mean of the logarithms and more than 99 per cent of the counts in each class were within three standard deviations of the mean of the logarithms. 4. The geometric means found for the four different classes of milk were found to be significantly different when a test for significance was applied. 5. No attempt has been made in this paper to fix the limits (counts) of the four different classes. This paper shows that the geometric means of the methylene blue classes are significantly different. By examining the standard deviations of the means of the logarithms in Table 1 it is seen that the distributions of the geometric means do not overlap.