It is, of course, an extremely important matter to determine to what extent human diseases are caused by bacteria. It is not easy, nor indeed possible, to do this to-day with accuracy. It is no easy matter to prove that any particular disease is caused by bacteria. To do this it is necessary to find some particular bacterium present in all cases of the disease; to find some method of getting it to grow outside the body in culture media; to demonstrate its absence in healthy animals, or healthy human individuals if it be a human disease; and, finally, to reproduce the disease in healthy animals by inoculating them with the bacterium. All of these steps of proof present difficulties, but especially the last one. In the study of animals it is comparatively easy to reproduce a disease by inoculation. But experiments upon man are commonly impossible, and in the case of human diseases it is frequently very difficult or impossible to obtain the final test of the matter. After finding a specific bacterium associated with a disease, it is usually possible to experiment with it further upon animals only. But some human diseases do not attack animals, and in the case of diseases that may be given to animals it is frequently uncertain whether the disease produced in the animal by such inoculation is identical with the human disease in question, owing to the difference of symptoms in the different animals. As a consequence, the proof of the germ nature of different diseases varies all the way from absolute demonstration to mere suspicion. To give a complete and correct list of the diseases caused by bacteria, or to give a list of the bacteria species pathogenic to man, is therefore at present impossible.