THE WEATHER, INFLUENZA, AND DISEASE. 591 



same year, accompanied by a heavy rainfall, was marked by the occurrence of a few 

 cases of influenza towards its close. No further cases, however, were registered until 

 the spring of the next year. The epidemic of 1891-92 was therefore ushered in, after a 

 preceding period of cyclonic weather with a heavy rainfall, by an accompanying change 

 to anticyclonic conditions, but, unlike the last, it declined at a time when there was no 

 excessive rainfall. 



4. The fourth epidemic, although small in extent, seems to me to be the most 

 interesting of the series. The temperature in the first two weeks of January 1893 fell 

 to a very low point, under the influence of anticyclones which had persisted from the 

 middle of December. In the third week of the year the barometer was affected by 

 depressions to the west and north of our islands. These continued until the week 

 commencing on the 5th of March, when anticyclonic conditions again resumed their 

 sway, and continued unchecked until 23rd June, sixteen weeks later. Of the six weeks 

 which preceded the epidemic, five were cyclonic in type, with a heavy rainfall, and a 

 high temperature for the time of year. The last week before the outbreak was anti- 

 cyclonic, and such conditions continued throughout the epidemic, with practically no 

 rainfall until the last two weeks. The temperature remained uniformly high, rising 

 above 70° in the last weeks of the outbreak. The weather for the four weeks after the 

 subsidence of the epidemic continued to be anticyclonic in type, but with a slightly 

 increased rainfall, less sunshine, and a greater mean temperature. 



The fourth epidemic was therefore preceded by a wet period, ushered in by dry 

 weather, accompanied by great heat, and its close occurred in slightly wetter weather, 

 but under anticyclonic conditions. 



5. The middle of October 1893 saw the advent of still another outbreak, the third 

 of the winter epidemics. The six weeks before may be divided into two parts. The 

 first four were influenced by cyclones, and were consequently warm and wet ; the last 

 two — that is, those immediately preceding the attack — were anticyclonic, colder, but still 

 with a considerable rainfall. Throughout the sixteen weeks during which the influenza 

 lasted, the weather remained cyclonic, with a heavy rainfall. At the close of this period 

 the precipitation became very large — 2 "2 inches in two weeks. The four following weeks 

 were also very wet, with a mean of 1*32 inches. Coincident with this heavy rainfall, 

 the epidemic subsided. The fifth epidemic, therefore, began after a short anticyclone 

 had become established over our islands, continued during a long spell of cyclonic 

 weather with a considerable rainfall, but was, so to speak, drowned out by the heavy 

 rains of the last two weeks of January and of the whole of February. 



6. The sixth and last epidemic of which this paper treats took place in the spring 

 of 1895 (11th February to 31st March), following a period of severe cold, and persisting 

 through, for the first two weeks, a still more intense frost. In this case, again, the 

 preceding weeks were cyclonic in character, with a marked rainfall, and with a great 

 deficiency of sunshine. And again the epidemic began on the establishment of anti- 

 cyclonic conditions. The rainfall was very small until the last two weeks of the 



