The Weather of 1902. 27. 



spends with the deficiency of rainfall by which the year was 

 marked. 



Thunderstorms, etc. — During 1902 thunderstorms were 

 rare, and hail was equally uncommon, but lunar halos were not 

 unfrequently observed, especially in the later months of the 

 }ear. With regard to the wind observations, the south-westerly, 

 as usual, blew on the greatest number of days, viz., 88; the 

 next in point of frequency were due west and east, which 

 prevailed each on 56 days; the next was the south-east, with 

 43^ days; and the north-east and the north-west, which were 

 nearly equal, the north-east having 39 days, and the north- 

 west 37; due south had 29, and due north 9, and 6h were 

 variable. 



WEATHER AND HEALTH. 



Dr Maxwell Ross proposed a vote of thanks to Mr Andson 

 for his valuable paper. The weather of the year had been 

 exceptional, but it had not had \-ery much effect on the health 

 of the community, except that it might be that it had rather 

 improved the health than otherwise. He found that of the 

 eleven areas which he served as medical officer of health and 

 the statistics for which he had to collect, there were only two 

 whose death-rates were higher than their averages for a con- 

 siderable number of years. In not one of the county districts, 

 where he had the averages for something like twenty-one years, 

 was there any death-rate above its average. The county death- 

 rate was one of the lowest we had had during the past twenty 

 years; he did not know that there was a lower. The zymotic 

 death-rate was certainly the lowest. And one very curious fact 

 had been that the death-rate from diarrhoea and enteric fever or 

 typhoid, diseases which were associated with the warmer period 

 of the year, had been lower than they ever were before in his 

 experience in Dumfriesshire. He was rather interested in work- 

 ing out from the averages what one might call the "expected 

 deaths " for each month of the year. He got this very curious 

 result. In January, which was a favourable month but followed 

 a very stormy December, the deaths were actually 28 less than 

 the expected number. In February the actual deaths were 19 

 less than the expected deaths; in March, 8 less; in April, 36 

 less. Then they had a series of months, four following each 

 other, in which the actual deaths exceeded the expected deaths. 



