174 



the mean annual fall increases, the period of the maximnm occurs 

 later in the year." Where the rainfall is small in annual amount, 

 the maximum, generally speaking, is found to occur in the summer 

 months ; where it is moderately large the maximum occurs in 

 autumn ; where it is excessive the maximum does not take place 

 till mid- winter. In like manner the minimum fall which, iu 

 places where the rainfall is small, ordinarily occurs in the early 

 spring, " occurs about two months later in the year at stations 

 where the rainfall is very large." 



If now we consult three valuable tables given by Mr. Gaster, in 

 which he has exhibited the monthly percentage of mean annual 

 rainfall, collected from a large number of stations, for the three 

 decennial periods respectively, 1830-9, 1840-9, and 1850-9, we find 

 that in the second and third of these periods the percentage for the 

 month of June is nearly in every instance decidedly less than the 

 percentage for July. This at least is so in the case of those 

 stations in which the annual fall does not exceed 35 inches ; and 

 in what follows I confine myself to this limit of rainfall as embracing 

 the greater number of places in the southern half of England. The 

 Bath district, in which we here are most interested, has a yearly 

 rainfall not yet accurately determined, but lying somewhere 

 between 30 and 35 inches. 



The percentage for the months of July and August, in the same 

 periods, are nearly alike, but August is often in excess of July, and 

 both percentages are high, indicative of much wet. The percentage 

 for September in a large majority of instances is less than that for 

 either July or August, showing it to be a comparatively dry month. 

 October in all cases has a high percentage, considerably above that 

 of September, and would seem ordinarily to be the wettest month 

 in the year for stations where the mean annual fall does not 

 exceed 35 inches. In November, the percentage is sometimes 

 nearly or quite as high as that of October, but more often falls back 

 to that of September, showing it to be a month oftener dry than 



wet. 



The above deductions are from Mr. Caster's second and third 

 decennial periods, which on the whole much resemble each other. 

 His first decennial period, 1830-9, presents in the same manner 



