The just described, precisely considered and for the first time concisely arranged tabulation of the greater known epidemics teaches us important epidemiologic features of dengue. In order to complete the picture it would be necessary to compare the relations between individual epidemics. We will present only the most important facts, which, for the sake of brevity and clearness, we will arrange under the following heads:

1. Dengue is a disease of the tropical and subtropical regions, and within these zones it has a marked preference for the hot season -for summer. The disease almost always ends, as if suddenly cut off, on the occurrence of colder weather or the commencement of the cool seasons. Nevertheless, it is not probable that the restriction of dengue to the tropics is due only to the higher atmospheric temperature; yet nothing is known of any influences due to other cosmic relations, such as dryness or dampness, vegetation or soil. It is worthy of notice, however, that the occasional trespass into the ternperate zones-e. g., to Philadelphia in 1780 and Constantinople in 1889-generally occurs in unusually hot summers.

2. Dengue is a disease of the sea coast, ports, and coast cities. From the coast it occasionally also extends up the large navigable streams, such as the Ganges and Indus in India, the Mississippi in America, the Nile in Egypt, and the "White Streams" in- the Fiji Islands. This regular adhesion of the disease to the sea coast and river banks-yellow fever behaves in a precisely analogous fashion- shows that the shipping traffic determines, in the first place, the distribution of dengue. The disease generally has its limit where marine intercourse ceases. Only rarely and with difficulty, and then only for short distances, does it penetrate into the interior of a country by the main roads. Only once for a short time did the disease follow the railroads in India, extending from Calcutta to Umballa and Ludi ana on the Panjab frontier. No doubt this characteristic of dengue is due partly to the fact that in the tropics the principal mode of communication is along the sea coast. Influenza, too, although affecting both land and sea in the tropics, was confined principally to the chief trade ports. (Compare p. 539.) But it is not traffic alone which limits dengue to the sea coast. There are undoubtedly yet other factors: local conditions found on the coast, especially favorable to dengue. It is perhaps on account of the decreasing warmth of the atmosphere that dengue does not rise to any altitude. In Reunion, one of the favorite foci of dengue, the cooler, higher lying portions of the island, although in constant communication with the shore, remained almost entirely exempt in all the epidemics. The same was noticed in Cuba, Jamaica, Martinique, and almost everywhere. In Lebanon dengue never extended higher than from 300 to 400 meters above the sea level, but in the particularly hot year of 1889 the disease exceptionally reached places on Lebanon from 1200 to 1500 meters high, being brought there from Beirut (de Brun, Mahe). On the other hand, in Turkey in 1889 the disease remained limited, in spite of the coastal communication with the interior, to the coast around Constantinople, to the summer residences on the Bosphorus and in Asia, and to some parts of the Black Sea and to Salonica, once again following the old rule of remaining strictly limited to the sea coast; we would again point out the complete analogy to yellow fever.

Regarding the epidemic of 1889 in Constantinople and the vicinity von During says: "All the ports in active communication with each other became infected, in a sequence practically corresponding to the activity of the traffic between them."

3. Dengue is to clay considered a "highly contagious" disease. Its whole manner of distribution as regards time and place and the numerous examples of its conveyance by marine intercourse confirm the assumption. The disease has been repeatedly carried from continent to continent and from coast to coast by pilgrim, emigrant, and troop ships; from India to the Arabian and Egyptian coast, or vice versa; from West India to the coast of America, to Alabama, South Carolina, Louisiana, Venezuela, Texas, to the Bermuda Islands, and even to Philadelphia and Cadiz in Spain.

In 1885 the Fiji Islands became infected from Numea (New Caledonia) by a European with dengue who had been permitted to land. In 1893 the whole East India fleet was infected by one ship, the "Boad icea," which had been infected with dengue in Bombay.

There can be no doubt that dengue appears in any place where it has not previously occurred only if imported. Moreover, there exist numerous reliable observations showing that dengue, imported into a house, barracks, a ship, or a fort by a single patient, causes immediate further infection and gives rise to an epidemic. A number of observations made by de Brun in Syria during 1889 show clearly how the disease, carried by a patient into a family, house, or village, led to continuous distribution by contagion, and to convection of the pestilence to a distance by the mails and caravans and to the neighboring Lebanon villages.

A number of analogous and striking examples proving the direct transmission from person to person have been partly observed and partly collected by von During and Mordtmann.

There can nowadays be no doubt that dengue patients give off the specific germ, and that this germ may, in favorable circumstances, directly infect others. Formerly a purely miasmatic conception of the disease was accepted.

Nevertheless, the whole series of epidemiologic facts indicates that contagion alone, in the strict sense of the term, cannot satisfactorily explain either the geographic distribution of dengue or the occurrence of the larger epidemics at the point of outbreak, or the local manner of distribution and the en masse infection. Consequently this matter is not so simple as modern extreme contagionists would have us believe.