161 



years after another minimum occurred, and twenty years later the 

 fourth minimum was registered. 



It would seem, then, that the periods between the ocurrences of 

 low temperatures are also multiples, or near some multiple, of five. 

 Again, the mean temperature of the summer months recorded as 

 beiug above 70° with one exception occurs in periods, the number of 

 years in each being some multiple, or near some multiple, of five. 

 The following is a table of probabilities based on the yearly mean 

 temperature for the last fifty-seven years. The expectation that the 

 mean annual temperature for the next fifty-seven years will be — 



There is no probability that the meai; yearly temperature will be 

 below 43° during the next fifty-seven years. 



Precipitation. 



Regarding the rain, it is found that in 1843 the mean rainfall 

 was 51.58 inches; in 1850, 55.05 inches was noted; in 1853, ten 

 years from the time the first mean above 50 inches was observed, 

 51.23 inches of water fell ; five years elapsed, then a mean of 51.28 

 inches occurred; four years later, in 1863, 57.71 inches of rain 

 fell ; six years later, 53.29 inches, — and thus another cycle of ten 

 years was completed. It would thus seem that the mean annual 

 rainfall of 50 or more inches occurs in periods, the number of 

 years in each period being some multiple of five. But if we con- 

 sider the three largest yearly means, it will be seen that they occur 

 in cycles of some multiple of twelve. For example, in 1850, as 

 stated above, 55.05 inches of rain were registered ; twelve years 

 after, 57.71 inches were recorded ; and twenty-four years later, 

 58,04 inches were observed. 



The probability that the mean rainfall for each of the following 

 fifty-seven years — 



Will be at least 58 inches is as 1 : 56 



Will be at least 57 inches is as 2 : 55 



Will be at least 55 inches is as 3 : 54 



Will be at least 53 inches is as 4 : 53 



