TABLE V.- SUMMARY OF GRADES GIVEN IN TABLES III AND IV. The use of the microscopic method in grading 2,440 samples of market milk according to its bacterial quality at Hobart, N. Y., has shown that this technique can be used successfully under commercial conditions as a means of distinguishing between milks containing few and many bacteria. In view of the possible chances of error present in the case of both the microscopic and the plating methods of grading when used as they were at Hobart, it was gratifying to find as high a percentage of agreement in the grades as 84-89 per ct. There is reason to believe that the results agreed as well as they would have done if the two sets of grading had been made by two different analysts using the officially recognized plate method.16 Inasmuch as there is no available standard by which the accuracy of the results obtained by either the plate method or the microscopic method can be tested, it was impossible to determine the relative accuracy of the results obtained by the two methods of grading. However, since the grades given individual samples by each method were found to agree under carefully controlled conditions in 91 per ct. of the cases, it raises the presumption that the grades given the samples by both methods were reasonably accurate. The more carefully controlled experiment carried out at Geneva for twenty-six months shows that it is entirely feasible and practicable to use the microscopic method for the commercial grading of milk into as many as three grades. The labor involved and the expense of equipment are much less than that required where the plate method is used; while still more important, especially to the producer, is the fact that the results are available within a few hours and the preparations can be preserved indefinitely. The latter fact permits the checking of fraudulent practices in grading in a very satisfactory and simple way. This does not mean that the technique is fool proof. It is capable of being misused either intentionally or unintentionally just as readily, as is the Babcock test, a misuse which has caused much dissatisfaction in the dairy industry. The microscopic examination of the milk was found to be more useful as a means of detecting the cause or causes of excessively high germ counts than was the plate method. This was particularly true in regard to the detection of high counts due to udders infected with streptococci. By its use it was possible to determine the approximate percentage of the high-count milk which was due to this cause. The fact that one-fifth of all the high-count milk was of the streptococcus type was a great surprise and indicates that much greater attention should be given to this source of high counts than is ordinarily given. In addition to the infected udders of particular animals, there appeared to be two other important sources of the excessively high germ counts: Namely (1) improperly steamed and poorly drained 16 Compare, for example, the results secured by analysts working in four New York City laboratories as given by Conn, H. W. Standards for determining the purity of milk. Pub. Health Repts., 30:2349-2395. 1915. milk cans which stood with covers on in places where the temperatures favored the growth of bacteria, and (2) milking machines which were not taken apart and thoroly cleaned frequently enough and which were not kept in sufficiently strong brine, or brine and chloride of lime. Because of cooling of the night milk and the fact that none of the milk was over sixteen hours old when tested, the growth of bacteria played a smaller part in producing the high counts than is the case where milk stands a longer time before delivery. The preliminary study made in order to learn the proper number of samples necessary to determine the grade of the milk with reasonable accuracy gave results which were instructive. While some of the dairymen brought milk whose quality did not vary greatly in the different cans of milk delivered on any one day, or on consecutive days, other men brought milk whose quality was highly variable. It will therefore be necessary to study more extensive data gathered under a variety of conditions before any satisfactory answer can be given to this question. Enough was learned, however, to show that no fair estimate of the bacterial quality of the milk can be made from samples taken from morning milk only, or from night milk only. The final result of this work has been such as to fully justify the Committee on Standard Methods of Bacterial Milk Analysis in including a description of the microscopic technique in their final report.17 Likewise the reports from several of the large milk distributing companies in the State that are using this technique at the present time indicate that they are already finding it useful in their laboratories. 17 See footnote 4. 18 |