Most laboratories in which examinations of typhoid stools are carried out, receive their specimens at various lengths of time after passage by the patient. In warm weather, stools, particularly if they are liquid, are apt to show, after 12 or more hours, a marked increase in the number of other bacteria as compared to that of the typhoid bacilli present. Since there is no satisfactory enrichment fluid for the typhoid bacillus, the success or failure in isolating typhoid bacilli from stools by our present methods depends primarily on the proportion of typhoid to other bacteria present; Ficker and Hofmann,1 for example, estimated that 1: 300 represents the lowest proportion of typhoid that still gives positive results. Hence, it seems highly desirable to preserve the original proportion of typhoid to other bacteria (that is, to prevent overgrowth of the typhoid by the latter) from the time that the stool is passed until it reaches the bacteriologist. The method to be described accomplishes this result, not only for periods of 8, 12, or 24 hours, but for a number of days; it therefore not only obviates the inaccuracies of diagnosis resulting from the usual delay in the arrival of specimens at the laboratory, but also permits of specimens' being sent from a whole state or country to a central station for examination. Our first experiments in this direction were concerned with the use of hypertonic solutions of sodium chlorid. Preliminary tests as to the effect of the salt solution in various strengths on pure cultures of typhoid bacilli indicated that a 5^c solution was the strongest that could be used without causing a rapid drop in the number of typhoid bacilli present. Sodium chlorid in this strength undoubtedly exerts