It is noteworthy that the English death-rate from pulmonary tuberculosis—which is responsible for 71 per cent. of the total mortality from tuberculosis, and which is practically always due to infection from a human source,—declined in males between 1871-75 and 1876-80 by 7.2 per cent.; in the next quinquennium by 9.8 per cent.; between 1881-85 and 1886-90 by 8.3 per cent.; in the next quinquennium by 9.5 per cent.; between 1896-1900 and 1901-05 by 7 per cent.; and between 1901-06 and 1906-10 by 9.7 per cent. Evidently a large share of the reduction of the death-rate from phthisis occurred before it was generally regarded as an infectious disease, and before sanatoria were in existence for its treatment. It should be added that since the possibilities of infection have been realised and the need for treatment of the disease has been appreciated, there has in no part of the world, so far as I am aware, been an adequate application of known methods of prevention and treatment. We must look elsewhere, therefore, than to intentional measures directed against tuberculosis for an explanation of its decline during the period before Koch discovered the tubercle bacilli and before the significance of this discovery was appreciated; and attempt to appreciate the relative value of the factors of decline operating before and since our outlook on the disease was fundamentally changed. Certain facts stand out beyond controversy, and on these administrative control must necessarily be based. _Basic Facts as to Tuberculosis_ 1. Tuberculosis is a chronic infectious disease with a low degree of infectivity. Circumstances favouring infection have a high degree of importance; but tuberculosis does not develop in the absence of the tubercle bacillus. No infection, no disease. 2. Tuberculosis may remain latent in the system for many years, and there is strong reason for thinking that the infection of a large proportion of early adult tuberculosis was acquired in childhood. 3. The two types of tubercle bacilli, bovine and human, are stable both in character and in degree of virulence, and are not interchangeable so far as can be shown by protracted experimentation. The human type of bacillus is the chief source of infection of mankind, though bovine infection is not negligible. Out of 98 children between the age of 2 and 10 years who had died in various hospitals from all causes unselected, 18 or 18.4 per cent. were found to have been infected by tubercle bacilli of the bovine type, and 81 or 81.6 per cent. by tubercle bacilli of the human type. (Report on Investigations made in the Laboratory of the Local Government Board, Annual Report of the Medical Officer of the Local Government Board, 1913-14, p. lix.)