Agglutination is not very practical owing to the frequent absence of agglutinins from the serum of plague patients. Then, too, there is a marked tendency to spontaneous agglutination on the part of the plague bacilli. Strong states that culturing at 37°C. lessens this tendency to spontaneous agglutination. Again, even when present, the titre of plague-agglutinating sera is usually quite low so that one must work with dilutions of from 1 to 10 or 1 to 20. PROGNOSIS Pneumonic and septicaemic plague give an almost absolutely unfavorable prognosis, many stating that every such case dies. As regards bubonic plague the mortality averages 75%. The Egyptian epidemic of 1900 gave an average mortality of 50%. The mortality in natives is much higher than that among Europeans, these latter often showing death rates under 25% while in the same epidemic natives show from 75% to 95% mortality. Plague pneumonia, however, is absolutely fatal for Europeans as well as natives. PROPHYLAXIS AND TREATMENT =Prophylaxis.=—In pneumonic plague it is the human patient and not the rat which has to be considered. The infection is spread by means of droplets of plague bacilli-laden sputum which are sprayed from the mouth of the patient in the act of coughing. As a result any person entering a ward containing plague pneumonia cases is extremely liable to contract the pneumonic form of plague. The attendants are protected by bag-like masks or successive layers of gauze and cotton wool applied as bandages over face and neck. Motoring goggles make a good protection for the eyes and small rolls of cotton should be placed along the sides of the nose to absolutely prevent the possibility of bacilli being drawn down to the entrance of nose or mouth. These masks should not have any weak spot in their armor. _Spread of plague pneumonia_.—It has been noted that when secondary pneumonia develops in the course of bubonic plague in India, the Philippines, or other hot countries, it is not followed by primary plague pneumonia outbreaks. This is thought to be due to the fact that the windows are wide open and the relative humidity low, conditions which are the opposite of those which existed in Manchuria where the intense cold made the closing of windows necessary and where the air of rooms or wards was saturated with the moisture from the occupants. As the main consideration for the spread of pneumonic plague seems to be high relative humidity it would seem that hospital wards could be constructed so that the air supplied by artificial ventilation would be very dry. The recent outbreak of septicaemic plague in Ceylon at a time when there was no plague noted in rats would indicate that other transmitting agents than infected rat fleas were operative. The most probable transmitting insects to be suspected would be the bedbug and human flea.