RAINFALL IN DORSET. 197 



returns were taken iuto account the number would be reduced to 150. 



The ratio of the rainfall to the average, of the same 20 stations 

 as those given last year, is 89, a deficiency of 11 per cent. It 

 varied between 106 at Hamworthy and 76 at Weymouth. 



This has been a year of extremes. May was excessively dry 

 everywhere ; even more so than last year, when the drought was 

 unprecedented. In September, on the other hand, nearly as much 

 rain fell at some stations as in September of four ordinary years. 

 Very rarely does the rainfall in any month reach three times the 

 average or 300 per cent, of the normal. The very wet month 

 following the 20th of October, 1894, is the most recent instance of 

 a similar fall ; but it happened at the most rainy time of the 

 year, and, though the fall of rain was greater, the ratio to the 

 normal is less remarkable. Previous to the present September 

 it had only attained this high value thrice since registration began. 

 In the disastrously wet November of 1852 the ratio at Melbury, 

 the only Dorset station for which details exist, was 338, and in 

 September, 1866, 321 at the same place. In December, 1876, 

 it was 307 at Osmington. Now, in September, 1896, it has con- 

 siderably exceeded three times the average in South-East Dorset, 

 the highest ratio being 373 at Swanage and 372 at Humworthy. 

 The area involved in this excessive fall comprises the Isle of 

 Purbeck and all South-east Dorset ; and from data in " British 

 Rainfall " it seems to have extended with somewhat less intensity 

 over South Hants and the Isle of Wight into West Sussex. In 

 West Dorset the ratio diminished to 200 and to less in Devon ; 

 and in the North- West it was nearly as low. From the paucity of 

 observations no close comparison between 1866 and 1896 is 

 possible for the whole County. In all except five cases the 

 stations are different and therefore do not represent the same 

 areas ; but from an average of 10 gauges in the earlier year and 24 

 in the latter, September, 1866, was apparently slightly the wetter 

 month. The rain was nowhere less than two-and-a-quarter times 

 the average, and was more evenly distributed than in 1896. 



