In England the spring epidemic of 1891 began in numerous rural districts. For nearly four months the disease raged in the north before it reached London, about the beginning of May (!). A similar condition of affairs occurred in Australia.

In these epidemics there is no definite direction of spread. They can only be limited chronologically, not geographically. The attempt to construct pestilence routes has proved to be completely illusory. In various parts of countries epidemics arise simultaneously, last a longer or shorter period, and then recur after a brief pause. In the interior of large countries, which on the whole remain free from the epidemic, isolated local epidemics of some degree of magnitude and duration may occur; whereas the immediate neighborhood may remain entirely free or become affected only at some later period.

Commerce, as a whole, no longer plays an appreciable role. Importation of the disease hardly occurs. We must not, however, overlook the fact that the densely populous and commercially most active civilized countries, namely, Europe and North America, form the principal points of selection of later epidemics; and that in some small districts the spread of the disease can occasionally be shown to" be in accordance with the conditions of traffic-e. g., in Mecklenberg-Schwerin. Note must also be taken of the fact that at this time, as in the actual pandemic, the north of Germany was obviously affected some time before the southern German states.

The Wurtemberg report of the epidemic of 1891-1892 says: "The capital and several of the neighboring cities were affected only to a slight extent by the epidemic. In some few districts several isolated portions and villages were very severely attacked; the remainder, on the contrary, were hardly affected at all."

Wutzdorff remarks, in summing up: "One of the peculiarities of the epidemic in question (1891-1892) was that in the neighborhood of markedly affected districts some distinct localities were quite exempt or only affected to a very slight degree." Further: "Some observers who noted the first appearances of the disease in distant isolated localities believed that during this epidemic (1891-1892) human intercourse played a very much less important role in the spread of the disease than it did in the pandemic of 1889-1890."

In almost every case, at the point of its origin, the epidemic developed and spread slowly, and only after several weeks did it reach its acme, dying out as gradually as it arose. In almost every case, especially in that of the winter of 1891-1892, the duration of the epidemic was from four to five months; and occasionally it showed exacerbations and remissions. The morbidity, in spite of the long duration, was markedly less. All this forms a striking contrast to the explosive commencement of the pandemic of 1889, when the acme was reached after fourteen days and the duration was strictly limited to six to eight weeks.

The rapid spread of the disease in any area, so characteristic of the pandemic, was tiever even approached in any of the later epidemics. The spread, even in small districts, as, for example, in a Prussian province, often required several months more time than the pandemic in its cyclonic spread over the entire world.

From the statistical summary of Wutzdorff we see that the epidemic of 1891-1892 in all German cities required several months, usually from October or November to February or March. In Danzig it took from October until May; in Breslau and Liegnitz, from September until April.

Much the same happened in England, Denmark, France, and Sweden. The spring epidemic of 1891 in London lasted from April to July; in Copenhagen, from May until August. The winter epidemic of 1891-1892 in London lasted from December to April; in Edinburgh, from November until February; in Paris, from October until January; and for the same space of time in St. Petersburg, Vienna, etc.

Isolated exceptions were noted, however, to the rule just quoted. In Yorkshire (Weath on the Dearne) the epidemic suddenly arose between the eleventh and thirteenth of April, 1891, reached its acme in ten days, and in twenty days more rapidly disappeared. Similar exceptional outbreaks occurred in Sheffield, where the first pandemic arose in a slow and fluctuating manner and declined gradually, whereas the second epidemic of 1891 arose like an explosion, being of brief duration and rapid disappearance, but showing a considerably greater mortality than in 1889 (339 deaths in 1891 as against only 96 in the pandemic).

In general the lessened morbidity and the less explosive character of influenza in the later epidemics appear also from the fact that there was only occasionally any necessity for closing the schools, as Wutzdorff shows in his official report. In the same way the rapid overfilling of the hospitals with influenza patients, which took place in 1889-1890, did not occur. Moreover, there was not the same sudden attacking of the masses, particularly of the officials of railroads, postal and telegraph bureaus, the employees of large warehouses, factories, offices, theaters, the officials of the courts, etc., as in the pandemic of 1889-1890, where official functions, trade, and commerce were brought to a standstill. Another peculiarity of the later epidemic may be found in the more dangerous character of the disease, as is shown especially in the English statistics, with the decidedly greater mortality.*

The clinical picture of the disease, with its numerous complications, remains the same, unaffected by the change of epidemiologic character.

The epidemiologic facts just mentioned-(1) The generally diminished morbidity of the later epidemics; (2) the diminished geographic distribution of the disease and the scarcely recognizable character of its communicability; (3) the slow development and extension over several months, shown by the later epidemics; and (4) the continuous diminution in frequency and intensity of epidemics from