49 s Prognoftics of [Book V. 



obfervations made from 1677 to 1789, Mr. Kirwan 

 lays down the following rules or principles. 



i ft. When there has been no ftorm before or after 

 the vernal equinox, the enfuing fummer is generally 

 dry, at leaft five times in fix. 



2d. When a ftorm happens from an eafterly point 

 cither on the 1 9th, 2Oth, or 2 1 ft of March, the fuc- 

 ceeding fummer is dry, four times in five. 



3d. When a ftorm arifes on the 25th, 26th, or 27th 

 of March, and not before, in any point, the fucceed- 

 ing fummer is generally dry, four times in five. 



4 th. If there mould be a ftorm at S. W. or W. S. W. 

 on the 1 9th, 2Cth, or 22d, the fucceeding fummer is 

 generally wet y five times in fix. 



Mr. Kirwan adds, that it rains lefs in March than 

 in November, in the proportion of feven to twelve. 

 It generally rains lefs in April than in October, in 

 the proportion of one to twoj lefs in May than 

 September, in the proportion of three to four. When 

 it rains plentifully in May, it generally rains but little 

 in September; and the contrary. A week is ac- 

 counted wet when it contains four wet days, or more; 

 a month, when it contains three wet weeks ; and a 

 feafon, or quarter of a year, when it contains two wet 

 months. He terms that a wet day in which rain falls 

 to the amount of one pound troy, in the fpace of a 

 IqOare foot. 



In any given year, the probability of a dry fpring 

 is in the proportion of twenty-two to fix wet, and 

 thirteen variable. Of a wet fummer it is twenty to 

 fixteen dry, and five variable. Of a variable autumn, 

 nineteen to eleven of wet or dry. That is, out of 

 forty-one years the fpring in twenty-two will be dry, 

 &c. j and fo in proportion *. 



Mem. Royal Iriih Acad. Vol. v. 



II. Among 



