r 3" ] 



Hence we fee the Spring was -wet, the Summer variable wet, 

 the autumn dry. 



Comparifon with the Rules of Probability. 



The Spring being wet the probability of a wet Summer was 

 the greateft, and that of a variable wet Summer the next greateft ; 

 the firft I- the other f by the 6th table. 



The Spring being wet and the Summer variable, the probabi- 

 lity of a dry Autumn was the greateft by the loth table. 



The obfervation that the more it rains in May the lefs it rains 

 in September was alfo remarkably verified this year, it rained ftill 

 lefs in Odober. 



But the obfervation that the abfence of ftorms in March 

 prognofticates a dry Summer was falfified this year. 



For. fome years paft I have remarked that a change of weather 

 moft commonly happens on the yth, 14th and 21ft of every month, 

 a day before or a day after, but fometiraes though rarely the 

 weather continues for three or four of thefe periods. 



Of 



