160 



two dates is very close to a multiple of six. If we consider the 

 next lowest maximum, 49.81°, which occurred in 1839, we find 

 that thirty years afterward the mean again reached above 49°, 

 namely, 49.17°. Then a period of six years intervened, and a 

 mean temperature above 49° was recorded for two years in succes- 

 sion, the true means being 49.12° and 49.47° respectively. 

 Reckoning from the first of these maxima, we find that just twelve 

 years elapsed, when the mean temperature was again above 49°, 

 namely, 50.88°. If we count from the last maximum, we find 

 twelve years intervened before the same mean was again recorded. 

 Hence it would seem that the mean temperature from 49° upward 

 occurred in periods of some multiple, or very near some multiple, 

 of six. 



If we consider the next lowest maximum, viz., 48.85°, we find 

 that the latter occurred in 1841 ; four years elapsed, when a 

 yearly mean of 48.02° was recorded. Twenty-four years later the 

 48° line was crossed, for the mean temperature was 49.17°. The 

 seventh and eighth years following, the temperature passed the 

 maximum line in question ; reckoning from the first of these, the 

 seventh, we find four years later another mean of 48.49° ; reckon- 

 ing from the second or eighth year, we find in just four years yet 

 another mean of 48.70° recorded ; that is, two years in succession 

 the mean temperature was over 48° but less than 49°. Taking the 

 first of these as our starting point, we find that four years inter- 

 vened, and the mean temperature of 48.29° occurred ; another four 

 years passed, and the maximum line was again crossed by the 

 mean of 50°. 



It would seem, therefore, from the above statements, that the 

 occurrence of a mean temperature of at least 48° appeared in cycles 

 whose number of years is some multiple of four, with an occasional 

 exception. As the yearly means approach the mean for fifty-seven 

 years, that is, 47°, the maximum and minimum points in the curve 

 occur more irregularly. The numbers forty-five, forty-six and 

 forty-seven, the integral parts of three means, occur forty -one 

 times ; that is, the number forty-five appears eight times, the num- 

 ber forty-six twenty-four times, the number forty-seven nine times, 

 and there have been thirty-six years during the period under dis- 

 cussion when the mean yearly temperature was below 47°. 



On examining the curve which indicates the mean temperatures 

 for the winter months during fifty-six years, it will be noticed that 

 the maximum means occur in cycles, the number of years in each 

 being some multiple, or very near some multiple, of five. If the 

 four lowest minima are considered, it will be found that from the 

 first minimum to the second, fifteen years had passed ; thirteen 



