of seaport cities in tropical and semitropical regions, but in these regions its prevalence is greatly restricted. In North America, although it has occasionally prevailed as an epidemic in every one of our seaport cities as far north as Boston, and in the Mississippi Valley as far north as St. Louis, it has not established itself as an endemic disease within the limits of the United States. In South America it has prevailed as an epidemic at all of the seaports on the Gulf, and on the Atlantic coast as far south as Montevideo and Buenos Ayres; also at several seaports of Mexico and Peru on the Pacific. At present the principal endemic foci of the disease are Havana, Vera Cruz, and Rio Janeiro. In Africa the disease is limited to the west coast, and so far as we know no epidemics have prevailed in the interior of tropical Africa, although the conditions would appear to be favorable for the development of an epidemic in case the disease should be introduced. The same is true as regards the populous regions in northern Africa and southern Asia where the rainfall is sufficient. The disease does not prevail in arid regions, or at considerable elevations above the sea level.
Yellow fever does not prevail as an endemic disease in places which have a mean winter temperature much below 65° F. (18.3° C), and as a rule epidemics are not developed at a lower temperature than 75° to 80° F. (23.8° to 26.6° C). The approach of cool weather checks the progress of an epidemic, and in those endemic foci of the disease (Havana, Rio de Janeiro, Vera Cruz) where it prevails annually it is essentially a disease of the summer months. That moisture is an essential factor is indicated by the fact that the disease does not prevail in arid regions where other conditions appear to be favorable, and that it is especially a disease of the seacoast and of the margins of great rivers. Heavy rains, however, exercise a favorable influence in checking an epidemic—probably by cleansing the streets, sewers, etc., in an infected locality. In the tropics the commencement of the rainy season often puts an end to the prevailing epidemic.
Decomposing matter of animal origin appears to form a favorable nidus for the development of the hypothetical yellow-fever germ. It is a disease of towns and cities, and especially of such as are in an insanitary condition. The writer's studies have led him to the conclusion that the infectious agent, as in cholera and in typhoid fever, is probably present in the discharges of the sick.
The early history of yellow fever is involved in obscurity, and it is doubtful whether we will ever be able to settle in a definite manner the disputed question as to its origin. Two principal theories have been advanced: one that it was endemic at certain points on the shores of the Gulf of Mexico at the time of the discovery of the New
