I may, I think, claim to have answered in part the question asked at an earlier stage of this address, as to the causes of the steady decline in the death-rate from tuberculosis in recent decades. I do not claim that any one factor has brought about this result. I do not claim that it has been caused entirely by diminution of opportunities of infection; but I deprecate the view that improved nutrition and other conditions diminishing susceptibility have played a predominant part. The facts of international hygienic history rebut this view. Although segregation of patients in institutions has played a great part in bringing about the result, diminution of domestic infection as the result of more cleanly habits has doubtless had an important influence; as has also the reduction of industrial dust. It is significant that general hygiene and improved care of the sick—quite apart from any intention to segregate—were associated with a large reduction in the death-rate from tuberculosis before the importance of reducing infection was fully appreciated; and that since the necessity for direct measures against tuberculosis was realised, since such measures have been begun, however imperfectly, in many countries, and since anti-tuberculosis educational propaganda has been somewhat active, there has been no increase in the rapidity of decline of the death-toll of tuberculosis. Of course, it cannot be seriously—though it is foolishly—argued from this fact that such direct measures are futile. Every year there has been increasing migration of masses of people into towns, with a corresponding increase of undesirable domestic overcrowding and of indoor occupations. If, therefore, such anti-tuberculosis measures as have been adopted,—whether direct measures or general sanitary measures,—had been associated with an absence of decline or with actual increase in the death-rate from tuberculosis it might still be that these measures have achieved much. Many conflicting agencies are at work, and it might well be that the apparent lack of success of the measures taken is due to the increased operation of countervailing influences. The importance of direct action for the control of tuberculosis must be judged not solely by necessarily imperfect statistical measurement on the basis of a few years’ observation, but _by ascertaining that the proposed measures are in accord with our knowledge of the natural history of the disease_. As we have seen, both comparative and human pathology assure us that tuberculosis is a communicable and therefore a preventible disease, and point the way to the means for securing this end.