286. The plate count and methylene-blue reduction test applied to milk

Abstract

1. Defining sensitivity as the power of discriminating significantly between milks of different bacterial constitution the plate count is slightly less sensitive than the methylene-blue test observed every 5 min. and slightly more sensitive than the latter observed every half-hour.2. The logarithm of the plate count is, in our opinion, the correct measure to use. Its coefficient of variation is slightly less than that of the methylene-blue test observed every 5 min.3. The plate count as formerly used and the methylene-blue test as now used agree in grading in over 80% of cases.4. Two samples whose reduction times differ by 30 min. are not significantly different. Such a difference is only significant if the result is based on the mean of duplicate tests with 5 min. observations or of triplicate tests with half-hourly observations.5. Provided transit hazards are eliminated and standard technique is very carefully observed, there is no significant difference in results from different laboratories by either test.6. Neither test necessarily reflects minor changes or faults in production routine. Under the conditions of our experiment the methylene-blue test was not, in cold weather, a satisfactory indicator of the numbers of organisms as distinct from ‘activity’.