The summer epidemic of 1891 followed a fine winter and spring, during which anti-cyclonic conditions were largely prevalent. But the epidemic was immediately preceded by wet weather and a low barometer. It took place in dry weather, and was followed by wet, cyclonic weather in turn. The great winter epidemic of 1891 followed an extremely wet and broken autumn. Simultaneously with the establishment of an anti-cyclone, with east wind, practically no rain, and a lowering temperature, the influenza commenced. Great extremes in the temperature followed, the advent of warmer weather and more equable days witnessing the disappearance of the disease. The fourth epidemic was preceded by a wet period, ushered in by dry weather, accompanied by great heat; and its close occurred in slightly wetter weather, but under anti-cyclonic conditions. The fifth outbreak began after a short anti-cyclone had become established over our islands, continued during a long spell of cyclonic weather with a considerable rainfall, but was drowned out by heavy rains. The last appearance of the modern plague, of which Dr. Gillespie's paper treats, commenced after cold and wet weather, continued in very cold but drier weather, and subsided in warmth with a moderate rainfall. The conditions of these six epidemics were very variable in some respects, and regular in others. The most constant condition was the decreased rainfall at the time, when the disease was becoming epidemic. Anti-cyclonic weather prevailed at the time. According to Dr. Gillespie, the tables seem to suggest that a type of weather, which is liable to cause catarrhs and other affections of the respiratory tract, precedes the attacks of influenza; but that the occurrence of influenza in _epidemic form_ does not appear to take place until another and drier type has been established. As the weather changes, the affected patients increase with a rush. He is of opinion that the supposed rapid spread of influenza on the establishment of anti-cyclonic conditions may be explained in this way. The air in the cyclonic vortex, drawn chiefly from the atmosphere over the ocean, is moist, and contains none of the contagion; the air of the anti-cyclone, derived from the higher strata, and thus from distant cyclones, descending, blows gently over the land to the nearest cyclone, and, being drier, is more able to carry suspended particles with it. He considers that temperature has nothing to do with the problem, except in so far as the different types of weather may modify it. The Infirmary records point to the occurrence of similar phenomena, recorded on previous occasions. Accordingly, if such meteorological conditions are not indispensable to the spread of influenza in epidemic form, they at least afford favourable facilities for it. CHAPTER XXXIII CLIMATE