34 But in regard to malaria the drier gardens show but a slightly lower rate of malarial infection and perhaps a lesser tendency for Europeans to contract the disease than those intersected by innumerable streams and swamps. For all practical purposes the native malaria and es- pecially the endemic index is the same throughout the whole district whatever the physical conditions may be. Again for some seven months of the year there is a steady reduction in breeding places which by January have in many places almost or entirely disappeared, and a great reduction in the number of adult anopheles. Yet at no time of the year is there any marked reduction of the endemic index. For example, the northern end of a particular line was during the dry weather at least a mile from the nearest water. Yet the endemic index after this condition had existed for some months was still 67 per cent. The prevalence of malaria at a time when anopheles are greatly reduced and their breeding places abolished reaches its acme on those gardens to which we have already referred as situated on the peculiar tract of sloping porous land in the Dina-Torsa district. These gardens to judge from their apparent unsuitability for ano- pheles, should have been almost free from malaria; as a matter of fact they are not so. At one of them, on the contrary, we found in April, May and August as remarkable an intensity of malaria as at any place we have visited. Not only were the endemic index and the spleen rate extremely high, but we saw an amount of sickness and misery the result of Malaria nothing, short of astonishing in the cir- cumstances. Clearly some very definite condition was involved enabling malaria thus to maintain itself under circumstances so apparently unsuitable for its spread. We believe this condition to be the fact that in certain conditions the amount of human infec- tion may remain undiminished owing to the suitability of the host to the continuance of infection for long periods, and thus a mini- mum number of anopheles serve to keep up a maximum amount of infection. We have referred to this condition in the introduction as "resi- dual malaria" and in conditions tending to make communities suitable for such a state we seem to have at work some of the most potent factors in the causation of hyper-endemicity.