Some writers have failed to visualize the fact that the segregation of a minority of the total cases of pulmonary tuberculosis for a portion of their illness can have had a marked influence on the prevalence of this disease. They appear to be judging tuberculosis by the same measure as they would apply to smallpox, which in an unprotected community spreads rapidly if a few cases are overlooked. The case of tuberculosis, like that of leprosy, is governed by the considerations that both these diseases as a rule require intimate and protracted contact for their spread, and that in both diseases there may be prolonged latency before active disease develops. A hypothetical illustration may serve to elucidate the order of magnitude of the influence exercised by institutional segregation. Let us assume—as is probably the case in England—that one-fifth of the cases of pulmonary tuberculosis are treated during one-third of a year institutionally under conditions in which they will not be liable to spread infection. Let us assume further that each of these cases has an infectious lifetime of three years. Thus one-fifth of the cases are deprived of their power to spread infection during one-ninth of their period of “open” disease. It being assumed that personal infection causes pulmonary tuberculosis and that segregation is efficient, segregation to the extent indicated above should secure a reduction in the death-rate from pulmonary tuberculosis of 100/(5 × 9) approximately 2 per cent. In actual fact the decline in the English death-rate from pulmonary tuberculosis since 1871 has been at a rate slightly under 2 per cent. per annum. _Koch’s Endorsement of Segregation View_ An extract from an article written by Robert Koch shortly before his death may be permitted (Epidemiologie der Tuberkulose Zeitschr. für Hyg. und Infektious Krankheiten. 4. XVII, 1910). I am entirely in agreement with Newsholme that the allocation of consumptives to institutions for the sick, as freely as possible and for as long as possible, is the most active means of avoiding infection and the consequent spread of phthisis. In my experience, too, phthisis has shown the most marked decline in those places where comprehensive measures have been taken for bringing consumptives into hospitals, and the converse has been the case where the converse conditions prevail. It is indeed obvious that in no other way can the danger of infection, which a phthisical patient constitutes, be so effectively removed as by isolation in hospital. Strong support of this method is afforded by leprosy, where good results in attacking the disease have been obtained by following the same principle. In addition to this factor there is a second, which also plays a very important part, viz., housing.