The scientific investigators would say you have injected with the sputum into the blood of your unfortunate patient, pus, morphological elements, and perhaps half a dozen other forms of bacteria, any one of which is just as likely to produce the disease as the bacillus you have selected. The first important step is, first isolate your bacillus. If I were to take a glass plate, one side of which is coated with a thick solution of peptonized gelatin, and allow the water to collect, the gelatinous matter will become solid. If now, with a wire dipped in some tuberculous matter, I draw a line along the gelatin, I have deposited at intervals along this line, specimens of tubercle bacilli. If this plate be now kept at a proper temperature, after a few days, wherever the bacilli have been caught, a grayish spot will appear, which, easily seen with the naked eye, gradually spreads and becomes larger. These spots are colonies containing thousands of bacilli. Let us return to our gelatin plate. We find a spot which answers to the description of a colony of tubercle bacilli. We now take a minute particle from this colony on a wire and convey it to the surface of some hardened blood serum in a test tube. We plug the tube so that no air germs may drop in, and place it in an incubator at the proper temperature. After several days, if no contamination be present, a colony of bacilli will appear around the spot where we sowed the spores. Let us repeat the process. Take a particle from this colony, and transfer it to another tube. This is our second culture. This must be repeated until we are satisfied that we have secured a _pure_ culture. If this be carried to the twenty-fifth generation, we may be assured that there remains no pus, no ptomaines, nothing but the desired bacilli. It is a proper material now for inoculation, and if we inoculate some of the lower animals, for instance the monkey, we produce a disease identical with phthisis pulmpnalis. Bacteria also afford peculiar chemical reactions. For example, nitric acid will discharge all the color from all bacilli artificially dyed with anilin, except those of tubercle and anthrax. One species is stained readily with a dye that leaves another unaltered. Thus we are enabled in the laboratory to determine whether the bacilli found in sputum, for example, are from tubercle or are the bacteria of decomposition. From what I have said of the tubercle bacillus, it would seem thoroughly demonstrated that it is the cause of tubercle in these animals. But we must walk cautiously here. These are not human beings, who know that like results would follow their inoculation. The animals used by Koch are animals very subject to tubercle.