384 PATHOGENIC BACTERIA. In the absence of a satisfactory method of securing definite small quantities of blood for immediate or sub- sequent use, I was led to make some experiments with capillary tubes to determine their possible value for the purpose. It is a well-known physical phenomenon that in clean capillary tubes fluids are attracted to a height varying according to the diameter of the tube and the density of the fluid. In tubes of equal diameter the height of the column is invariably the same. Such tubes can be made by heating a piece of ordinary glass tubing, such as is to be found in every laboratory, in a Bunsen flame for a few minutes until it becomes red and soft, removing the glass from the flame, and then pulling upon the ends steadily and slowly until the tube is drawn out to the desired diameter. The errors to be avoided in making the tubes will be—heating too much and making the glass too soft, drawing out the tube while still in the flame, and drawing too rapidly. The result of these erroneous methods will be that the tubes are much shorter and finer than is desired. A few mo- ments' practice will show just how the manipulation should be done to secure the best results. The fact was, however, established that tubes of about the same diameter showed almost no variation in the quantity of liquid contained. So little was the difference in the length of the column and the weight of the con- tained blood in tubes recognized by the eye to have uni- form caliber that I have no hesitation in recommend- ing an application of the capillary tube for securing small measured quantities of blood for the specific typhoid tests and similar experiments. The application of the method is simple and consists in: 1. Accurately weighing the amount of blood that enters a capillary tube of a size arbitrarily selected as a standard. 2. The manufacture of a large number of tubes of the same size.