TABLE VIII.-Period of year at which bubonic plague appeared (and disappeared) in cities with mean midwinter temperatures 36° to 45° F. Pans Oct. 18. Do. Sept. 22-Oct. 6. August; June 20-26. October; September; June 19. August to October. August. May 7-14. February (1 death). September. October-December: April and May. September and October. June to October; August; January (3 suspected cases). Aukland, New Zealand (Southern March-May; March; June; (September; November; SepHemisphere). Trieste. tember; December. Southern Hemisphere). Nov. 1-8. Sept. 21. Feb. 23 (in French troops). August-December; June. June and August. Nov. 24 to Dec. 3. September to December; May to November. May to December; November to May. In looking over Tables I and II for facts that would tend to discredit the deductions made as to the occurrence and disappearance of bubonic plague in accordance with the seasonal prevalence of fleas, we are confronted with a few data that seem inconsistent with those deductions. In Table I only 4 cities out of 27 have ever reported plague. Of these 4, Moscow reported 12 cases in two outbreaks between December and February. If this was bubonic plague it is incompatible with the belief that fleas disappear in climates such as Moscow in cold weather. No mention is made, however, in the Public Health Reports as to whether these cases were bubonic or pneumonic. In any event, no further cases were reported in Moscow in the summers following these two outbreaks. The cases appearing in St. Petersburg in May, 1910, also occurred somewhat too early in the year to be in accord with the belief that fleas have not become plentiful at this season. As in the case of the winter cases in Moscow, no mention is made as to whether these cases were bubonic or pneumonic. It is noted in the Public Health Reports, however, that in the years in which these winter cases were reported in St. Petersburg and Moscow, pneumonic plague was reported in Asiatic Russia and in the Astrakhan District. Three cases of plague occurred in Odessa in January, 1911, and were probably pneumonic. In Peking, China, the only two cases reported in 26 years occurred in January. It is significant that these two cases occurred in the winter in which 50,000 deaths were reported in Manchuria from. pneumonic plague, and it is more than possible that these two cases were pneumonic, though they were not so reported. This The only city listed in Table I in which plague was reported as occurring in the summer and fall months is Odessa, Russia. disease was reported in Odessa as follows: In a grain-shipping port such as Odessa, in a country where, possibly, little attention is paid to sanitation and where the human and rat populations may be both numerous and closely associated, it is possible for an outbreak of bubonic plague to occur, and the last three outbreaks reported for Odessa are probably such. It may be mentioned here that only in Russia and China has plague of any type been reported in localities having a mean midwinter temperature of 35° F. or below. In Table II are mentioned 25 cities in which bubonic plague occurred during the period 1897 to 1922. Of 1,409 cases reported during this period, 914 occurred in Kobe and Osaka, Japan. The only reported outbreaks in this list which seem to controvert the theory that the seasonal prevalence of fleas governs the spread of bubonic plague will be mentioned below. It should be noted at this time, however, that we are dealing here with mean midwinter temperatures considerably higher than those in Table I, and it is possible that while fleas may be definitely absent in the winter months where the temperature is 35° F. or below, they may be sufficiently numerous in the next temperature division-36° to 45° F.-to make possible the continuance of the disease through the winter. Yet the exceptions are so few as to make it difficult to believe that some explanation of the discrepancy might not be found if the correct records of the cities involved were accessible. The outbreaks (Table II) which appear to cast some doubt on the correctness of the theory advanced are as follows: Paris, January: Three cases reported as suspected plague. They possibly were not plague. Cattaro, port of Cettinje, February: Three cases reported in French troops. These were possibly imported, but it is not so stated. Japan: In the 19 outbreaks reported in 5 cities in Japan, all but 2 show the disease occurring between May and December. Both of these outbreaks were in Osaka, one occurring between January and April and the other between November and May. This last was an outbreak of 297 cases, reported in the midyear summary of Public Health Reports as occurring between November and May. Reference to the Public Health Reports, January to June of that year, show that 234 of these cases occurred between November 17 and December 21; 31 cases between December 21 and January 11, and the remaining 32 cases between January 12 and May 9-at the rate of about 2 cases a week. While there was a decided drop in the number of cases reported weekly-from 47 per week in November and December to 2 per week from January 11 to May 9-during the cold months, the fact remains that in Osaka human cases of bubonic plague were reported throughout the winter. The mean January and February temperature of Osaka is about 39.5° F. The December and March temperatures are 43° F. Whether this particular winter was an exceptionally warm winter in Japan has not been determined. All of the 35,500 cases of plague reported in Japan, with the exception of something over 1,000 cases, were reported from the island of Formosa (Taiwan Island). It should be mentioned in considering the cities of Table I that the principal seaports have all more or less intimate maritime relations with the plague centers of the Tropics and with those occasionally infected areas which are found in the regions adjoining the Tropics. At many of these cities vessels have been reported from time to time in the Public Health Reports as arriving with human cases of plague aboard, and it must be that many plague-infected rats have been brought into them during the period of years under consideration. In studying the outbreaks listed in Table II it is seen that they are short lived, apparently self-limited, and practically never result in a recurrence the following season after dying out during the winter season. The only exception to this is the outbreak in the Kobe-Osaka vicinity in the year 1908. Excluding the 914 cases reported in these two cities, only 495 cases of bubonic plague were reported in the remaining 23 cities in this group during a period of 26 years. The outbreaks shown in Table III indicate, in a general way, a rather definite seasonal prevalence for plague. This seems not always to coincide with the rise and fall in temperature, and it is possible that at the temperatures 46°-55° F. the prevalence of fleas is influenced both by rainfall or humidity and temperature variations. Further study of the records and flea surveys of localities in this temperature division are necessary before any accurate deductions can be made as to what it is reasonably possible to predict for such localities. It is, however, obvious that some rather important factor is involved in localities even of these high temperatures in limiting the spread of bubonic plague in 23 cities and countries over a period of 26 years to less than 10,000 cases, nearly two-thirds of which were reported from two localities, Bagdad and vicinity and South Africa. The periodic and infrequent occurrence of bubonic plague in cities and countries of this group is in rather striking contrast to the almost constant presence of this disease in the countries which are mentioned in the last group. The countries listed in the last temperature group are really tropical countries, and it is in these that 94.5 per cent of cases reported in the 26-year period have occurred. Bubonic plague has been reported continuously for the past 26 years, or since first reported, in the following countries: India, Java, Ceylon, Egypt, southeast China, Peru, Brazil, and Ecuador. This is in rather marked contrast to its irregular periodicity in those cities and countries in the third temperature group which have a lower mean midwinter temperature. It is in much more marked contrast to those cities mentioned in Tables I and II in which bubonic plague occurs at rare intervals and at definite seasons in most instances. Mention must be made of the report of plague in two sections of the world where it has been often reported and where it has caused many thousands of deaths. These two areas are Manchuria and the districts surrounding the Caspian Sea. In this study it has not been possible to examine the weekly records of cases reported. The semiannual summaries of cases given in the Public Health Reports sometimes refer to the outbreaks in these regions as "pneumonic," but not always so. From the relatively short periods in which so many cases and deaths were reported and the exceedingly high mortality rate mentioned, it is believed that all or most of this plague was of the pneumonic variety. Further detailed study will be necessary, however, should more accurate data as to the reports from these areas be desired. CONCLUSIONS. The more obvious conclusions to be drawn from this discussion and these tables are as follows: (1) Bubonic plague is essentially a disease of hot climates, and, having been introduced into tropical countries, it tends to persist indefinitely. (2) Outside of the immediate Tropics, this disease is rather definitely limited in the extent to which it will spread. (3) In countries with a mean midwinter temperature of 45° F. or below, bubonic plague is occasional, accidental, and distinctly selflimited, and it seems possible for it to occur in the colder regions only for short periods under unusual conditions. This study appears to me to sustain the deduction that the seasonal prevalence of fleas regulates the spread of bubonic plague; otherwise it becomes necessary to find at least some equally reasonable explanation for the absence of bubonic plague in cold countries. Any estimate of the practical results that would ensue from a determination of the correctness of the theory advanced in this article and the conclusions drawn therefrom must await confirmation by an actual and comparative flea survey of any given locality. VENTILATION OF SHIPS AFTER FUMIGATION WITH In a recent report by Mr. Stephen Olop, Superintendent of Construction and Recorder of the Board of Fumigation and Ventilation, United States Public Health Service, the following suggestions and recommendations are submitted: In connection with the problems in ventilation related to freeing vessels from poison gas after fumigation, the writer recommends: (a) That special attention be given to the adjustment of ventilator cowls, and (b) that the bilge pumps be operated as soon as the ship's crew may be allowed access to same with safety. The reasons are stated as follows: Upon further study of observations made by him in the experiments conducted on the Hartford, set forth in his report of December 6, 1922, and from miscellaneous inquiries, experiments, and observations made by him since that time, the writer concludes that special. attention should be accorded the adjustment of ventilator cowls. aboard ship. Observations made on the Hartford indicate that, generally, elimination of gas proceeds most slowly in the windward portion of the average hold. This appears to be the case not only where the hatch opening is the sole means for ventilation, but also where it is augmented by ventilators. The air will travel downward mainly at the leeward side, thence in counter-windward direction, thence upward and finally out. If the cowls of windward ventilators are set to face the wind, much air will enter through them, which is met by air currents within the hold moving in counter-windward direction, causing a conflict of currents and thus reducing their ventilating values considerably. Pockets occur mainly in windward portions of the average holds. The writer observed that on vessels carrying highly perishable cargo, such as fruit, particular attention is given to careful adjustment of the ventilator cowls, and that those at the windward side of a hold are turned away from the wind and those at the leeward side to face the wind, thus inducing a downward flow of air in the leeward and an upward flow in the windward portion of a hold. That such arrangement of the ventilator cowls is the most efficient appears |