there were no deaths in 44 cases in which the temperature did not rise over 103°; per contra, in 22 cases in which the thermometer rose over 106° there were no recoveries. Of 36 in which the temperature rose to between 105° and 106°, 22 died; of 80 with maximum temperatures between 104° and 105°, 24 died; and of 87 in which it ranged between 103° and 104°, only 6 died. The mean mortality in the whole 269 cases was 27.7 per cent. Although in some epidemics it has risen as high as 50 or even 80 per cent, of those attacked, the foregoing may be taken as a fairly representative mortality in yellow fever among the unacclimatized— something between 25 and 30 per cent. Among the permanent inhabitants of the endemic districts the case-mortality is very much lower— 7 to 10 per cent. During epidemics abortive and ambulatory cases occur; in these, icterus and other characteristic symptoms are often absent. Such cases may be hard to diagnose from febricula or mild malarial attacks. In them the mortality is nil. Some epidemics are particularly mild; in others the majority die. In the same epidemic the cases may vary in severity from time to time. In children the mortality is insignificant.
Pathological anatomy.— Depending probably on hæmoglobin diffused in the liquor sanguinis and tissues, and not on biliary pigment, the yellow colour of the skin is most marked in the dependent parts of the cadaver, especially in those parts which are subjected to pressure. Petechise are common in the skin and serous membrane; more considerable extravasations of blood may be found in the muscles. The brain and meninges are hyperæmic, and may be studded with minute hæmorrhagic effusions; like the other tissues of the body, they are stained a lighter or deeper yellow. The blood in the vessels of the general circulation is not firmly coagulated. The blood corpuscles appear to be normal, although there can be little doubt that there is in this disease a liberation of hæmoglobin, arising, possibly, from destruction of a proportion of the corpuscles. An important fact, as explaining the liability to passive hæmorrhages,
