P. Remlinger, Comp. rend. de la Soc. de Biologie, vol. 70, page 64, 1911, suggests a nmethod of inhibiting the growth of bacteria in samples of water during transportation to the Laboratory by the addition of common salt. In his experiments, ten grammes of sodium chloride were added to one hundred cubic centimeters of water. The flasks so prepared were kept at laboratory temperature and estimations made from day to day. His conclusion was that the salting of a sample of water by the addition of ten per cent. by weight of sodium chloride would prevent any fallacy arising in the bacterial count, provided that the examination was made within forty-eight hours of the collection of the sample. The importance of this discovery, if true, is at once apparent. Those of us who are constantly engaged in the examination of water transported over considerable distances, know the difficulties of keeping samples properly iced. If we could send out bottles for collection of samples containing a given amount of sterile common salt, and have the sample returned to us without icing, an enormous amount of trouble would be saved, and our results would be much more accurate. It was therefore determined to test the method. Remlinger's plan was followed closely. It seemed good to try other preservatives at the same time. Consequently, duplicate flasks containing resorcin and others containing caffeine were made in a number of cases. Resorcin was used in the strength of two cubic centimeters of a 25 per cent. solution to one hundred cubic centimeters of water. The caffeine was added in the proportion of ten cubic centimeters of an 8.3 per cent. solution of caffeine citrate to one hundred cubic centimeters of water. All were kept at laboratory temperature, plates made from time to time and tests .for colon made. We employed samples collected in the ordinary run of business, and samples to which colon was artificially added. One serious objection to salting, as well as to the other preservatives, was shown at the very beginning, namely: that the growth of colon bacillus is almost invariably inhibited. In only one test out of twenty-eight did we obtain it, although the sample was known to contain colon when the preservative was added. We also find that the bacterial content of the different samples is not acted upon with any degree of uniformity, but varies with the species contained in any given sample. For instance, in one sample of water the count may remain Read before the Laboratory Section, American Public Health Association, Havana, Cuba, December, 1912. 103 104 THE AMERICAN JOURNAL OF PUBLIC HEALTH practically the same. In the next sample, growth may be almost entirely inhibited. In other words, according to the types of bacteria contained, the action of these preservatives varies tremendously. Conclusion.-Our conclusion is that neither the salting method proposed by Remlinger, nor the preservation by means of caffeine, or resorcin, has any practical value whatever. We are forced then, for the present, to rely upon refrigeration for preventing changes in samples of water.