4. Animal experimentation shows that in animals of the same species the extent of tuberculosis produced depends to a large and probably to a dominant extent on the number of tubercle bacilli introduced into the system. Although doubtless there are variations in susceptibility in families, and in each individual at different periods, there is little doubt that in the main the same rule holds good for mankind. 5. Experience shows that dusty occupations, indoor occupations, alcoholism, over-fatigue, an attack of acute illness, especially of influenza, measles, or enteric fever, increase the danger of minimal doses of tubercle bacilli, and serve to bring latent foci of disease into activity. _Explanations of the Decreasing Death-rate from Tuberculosis_ In the light of the above facts, how is the steady and continuous decline in the death-rate from tuberculosis during the last fifty years to be explained? (_a_) No support is given by animal experiment to the assumption that the types of human bacillus infecting mankind have declined in virulence; and changes in the severity of consumption historically or currently in different races of mankind are equally explicable on the ground of differences in social misery, in sanitary conditions and associated heavier dosage of infection and neglect of treatment. (_b_) The facts do not appear to me to be reconcilable with the assumption that natural selection has increased human resistance to infection by tuberculosis; though, were this so, it would not justify refraining from every possible effort to control infection and to treat every tuberculous patient by the best known methods. Tuberculosis is an ancient disease, there being evidence of it in Egyptian mummies 1000 years B.C.; and any selective agency has, therefore, had ages for its operation. If the steady decline—approximating 2 per cent. per annum in the death-rate from pulmonary tuberculosis in England during the last thirty or forty years—has resulted from the acquirement of racial immunity, it is remarkable that a somewhat similar decline has occurred almost simultaneously during the last forty years in Great Britain, Germany, and America; while in France, Norway, and Ireland there has been little if any decline, or it has occurred only in very recent years. To assume that susceptibility to the tubercle bacillus in the course of its natural history has diminished in England, and that Ireland has not shared in this privilege would be to add one more to Irish grievances! This assumption does not fit in with international facts; which point rather to the conclusion that, during the period in question, unsatisfactory sanitary and social circumstances, including opportunities for massive and protracted infection, have continued to a greater extent and for a longer time in Ireland and France than in Great Britain, America and Germany.