These institutions furnish the best and most favorable field for studying the origin and the transmission of the disease from the first case. Official and private observers in all countries have carefully gathered data from these sources on which to base reliable statistics. P. Friedrich and Parsons have collected an enormous material bearing on these points. The most important points are:

Those institutions which were relatively most shut off from communication with the outer world, namely, the prisons, showed a remarkable immunity against the invasion of influenza.

Lancisi had already observed this in the epidemic of 1709 in Rome, and reports that the prisons of the Holy Inquisition remained free of the disease. A large number of prisons (in Germany 21 can be enumerated) remained entirely free of influenza in 1889-1890. Quite frequently- in not less than 39 institutions-the same thing was observed in England, where the prisons remained without a single case, although lying in the midst of towns like Portsmouth, Kendal, and others, which were markedly affected by the disease.

Clae Linroth noted the same condition in Sweden, and we agree with him when he says: "Influenza makes its way more easily over a space of from 500 to 1000 kilometers than it does over the small barrier of a prison wall."

Next to prisons come certain strictly secluded convents. Their walls also occasionally gave remarkable protection. A convent in Charlottenburg which was guarded most strictly from communication with the outer world, containing nuns and 100 other female inhabitants, remained entirely free from the epidemic of 1889-1890 (A. Hirsch). The same condition occurred in some other similar convents and monasteries.

Nearly always in prisons, asylums for the insane, and in a large number of other relatively closed institutions, was it observed that the first cases of influenza in these places occurred in servants, nurses, overseers, and officers having free communication with the outside world, or in new arrivals. Not until later were the prisoners, lunatics, etc., affected. In a great number of these institutions, especially in prisons, cases occurred among the officers and servants only, in some instances attacking every one of them, without a single case appearing among the inmates of the institution. This condition occurred in the prison in Braunschweig; in the Central Prison of Glogau; in the prisons of Norwich, Bodmin, Winchester, and Lewes in England. In almost every case the morbidity of the officers and servants of the affected institutions was much greater than that of the inmates*

In the Franciscan convent at Reutberg, which is strictly closed against the outer world, the first cases occurred in the two doorkeepers of the institution, and not until several days afterward were one half of the remainder of the inmates of the convent affected.

In the lunatic asylum at Erlangen cases of influenza occurred during the period from the twenty first of December, 1889, to the eighth of January, 1890-that is, for eighteen days continuously. These cases occurred exclusively among the officers, physicians, and nurses who came in contact with the outer world. Not until after these had been affected did the epidemic arise among the insane.

Among the 1112 insane in the institution at Palermo, 7.8 per cent, were affected. Of the 260 employees of the institution, 62 per cent, were affected. Numerous similar examples might be quoted from German, French, English, and Swedish literature.

* Compare the tables of P. Friedrich.

With conspicuous frequency the prisons, and to a certain extent also the lunatic asylums, were affected at a considerably later period than their environs.

This was the experience in the prisons at Sagan, Ludwigsburg, Strassburg, in the asylums at Lengerich, Wallingford, etc., at Hagenau i. E., at which places influenza arose in the middle of December, but the first cases were noted in the jail on the twentieth of January and in the prison upon the twenty second of January. Numerous analogous observations might be mentioned from the collective English investigations.

But no human being lives his life entirely secluded from the outside world, and so the isolated convents and prisons (separate cell prisons) could not often withstand the inroads of influenza. Once introduced, the morbidity of influenza was in many cases very considerable (in the jail at Wasserburg, 84 per cent.; in the separate cell prison at Freiburg, 30 per cent., of the inmates were affected). Nevertheless, we find, in carefully examining the statistics of Germany and England, that the morbidity in prisons and lunatic asylums, to which we may add the convents as closed institutions in the narrower sense, was decidedly less than the morbidity in open institutions, such as orphan asylums, seminaries, nursing institutions, deaf and dumb asylums, etc., to which influenza was carried much more easily, and where the free communication between the inmates of the institution provided a suitable soil for the disease. On analyzing the tables I find that the average morbidity in the Bavarian prisons amounted to 31 per cent.; in all the other German prisons, 26-percent.; in the insane asylums, 17.7 per cent.; while in the above named open institutions the average morbidity was never less than 58 per cent. The morbidity in the English prisons was decidedly less than in the German ones. It amounted, for example, in Strangeway (1000 prisoners)r to 7.7 per cent.; in Wandsworth (1043 prisoners), to 18 per cent.; in Pentonville (1126 prisoners), to 7.6 per cent., etc.

We have good reason also to consider influenza in hospitals. Hospitals do not really belong to the class of closed institutions in. which they are usually included. They are, on the contrary, open institutions (visitors). Further, they are ever a resort of numerous influenza patients, and hence are most liable, throughout the duration of the epidemic, to be in the highest degree a favorable place for contagion. One would have anticipated, for this reason, that in all hospitals where isolation of influenza patients could not be carried out, the contagious "character of the disease would have shown itself in the frequent affection of the inmates of such hospitals. This has not, however, by any means been the case. A large number of the largest hospitals, in the years 1889-1892, were almost entirely free from influenza inside the house, although the influenza patients were freely distributed among the other patients.