A MODIFICATION OF THE ALCOHOL TEST AS AN INDEX OF THE BACTERIAL COUNT AND KEEPING QUALITY OF MILK

During the past few years several simple methods have been devised by which it is claimed that the approximate bacterial count of milk can be estimated. The most successful of these is the Reductase-Fermentation Test of Barthel and Jensen. Hiscox and Starling (1925), however, have shown that this test "does not afford a satisfactory method for differentiating between samples having a low bacterial count. " A rapid method whereby the approximate bacterial content of a sample of milk containing relatively few organisms could be estimated would be appreciated alike by condenseries, creameries, town and village health departments. Such a test, however, would require to be simple and effective; that is (1) it should not require the services of a trained technician, (2) it should show not only the activities of bacteria but also of enzymes present, (3) it should be capable of being performed rapidly, (4) it should furnish the results in a short time and (5) it should entail the minimum of expense per sample. It is improbable, however, that any single test can be evolvedI which would fulfil all these conditions and at the same time be reliable. Since the only satisfactory basis, from a commercial point c: view, upon which to judge the cleanliness of a milk, is by the tinle which elapses before it actually sours, this question must receive primary consideration. Milk may be of high food value and free from pathogenic bacteria, but if it sours within a few hours of delivery, it is of little use. The keeping quality of milk is inti