A further fact, showing the effect of heat, is that malarial epidemics occurring about disease foci are more severe in warm than in cool summers. Moreover, during especially hot summers malaria breaks out and becomes even epidemic over large stretches of land ordinarily not infected.

Yet with the recognition of the important effect of heat on the development of malaria it must be remembered that heat, of course, is only a factor which modifies the development of the parasite in the mosquito. That the influence of heat may be counteracted by other factors is shown by the previously mentioned circumstance that in mild malarial regions a decrease in the disease is observed during the warmest months, followed by a recrudescence in autumn. This striking fact loses somewhat in significance when we remember that malaria has an incubation period of two weeks and over. The cases, therefore, which break out in autumn must be referred back somewhat to their date of infection.

There remain to be explained the epidemics mentioned in the literature as occurring in winter. It must not be forgotten that in the discussion of the development of an endemic we take into consideration only new infections. In the case of relapses the climatic factors are not so important. According to the majority of investigators, the fevers occurring in winter are usually relapses from a summer infection, for it is the common experience that when the summer epidemic is wide spread, the number of cases during the winter increases. Moreover, in this case cold may be of importance, inasmuch as many assert that" colds" are capable of bringing on relapses.

Moisture

The humidity of the soil also participates in the development of malaria ; the reason of this old established belief is now clear to us. In the large majority of malarial foci yearly experience demonstrates that the severity of the epidemic is dependent on the amount of rain fall-that is, the greater the amount of rain, the more prevalent is the disease. The following is the common observation:

When rain occurs during the warm season after a period of drought, there is an increase in the number of cases; but when the rain is continuous and floods the ground, the number visibly diminishes, only to rise after the rain has ceased and the sun has caused evaporation of the surface water and the formation of numerous small collections of water. This is evident, especially in the tropics; Fayrer, for instance, asserts that the greatest number of cases occur in India just after the rainy season, during the months from September to December.

The development of malaria is, therefore, favored by a certain degree of superficial water, but retarded if this is exceeded.

Consequently the severity of an epidemic is not absolutely dependent on the quantity of rain fall, since lesser amounts at intervals between which the ground has an opportunity of drying are more favorable than weeks of continuous rain, especially when the temperature is low.

Jilek studied the connection between the amount of rain and the morbidity from malaria in Pola, and has given us the following statistics:

The contradiction between 1865 and 1867 is explained by examination of the period during which the rain fall took place. The same is seen from the following table:

An incontrovertible proof of the importance of moisture in relation to malaria is shown by the circumstance that the banks of still or running waters of every kind and marshy regions are the most frequent seats of the disease. The worst hotbeds in the world are on the banks of rivers, the coasts of seas, and in swamps. The floods occurring in spring pave the way for the estivoautumnal fever that thrives the more luxuriantly the more intense the subsequent heat. This flooding may be carried out artificially for industrial or agricultural purposes (rice fields, fish ponds, improvements, etc.), and produces the same effect as natural inundations.

Even very small water basins, like puddles or circumscribed swamps, can give rise to localized malarial foci when the other requirements, which we now know to be infected mosquitos, are present.

That malaria can be intense in dry regions is equally certain, but we have a complete explanation in the fact that sources of water abound during the rains, to disappear completely in the dry weather. The anophelines bred earlier in the year still live and transmit infection through the period of drought.

It is probable that such an explanation is applicable to the statement of Vincent and Burot that, during the campaign in Boeni (Madagascar), malaria raged most frightfully from May to September, although the heat was moderate and not a drop of rain fell.

The relation of swamps, marshes, ponds, etc., to malaria has for a long time attracted the attention of observers. Malaria was even regarded by some as a disease occurring only in swampy regions, and this assumption was expressed in the names " swamp fever" and " paludismus." It need scarcely be stated that this is a completely false idea. It is true now, as in the past, that marshy regions are usually infected with malaria, and that the worst foci are found in such places. Still there are swamps of large extent which show also the other requirements necessary for the origin of the disease that manifest no, or relatively little, malaria, while localities possessing nothing of a swampy character are sometimes hotbeds of the disease.

It may be laid down as a rule that swamps are more insalubrious the more level their surroundings; in other words, the more frequently a portion of the ground is laid free and flooded by alternating dryness and rain. Moreover, salt water swamps along sea coasts (brackish water) are usually particularly unhealthy.