^,30 Clark: Xoles on the York Rdinfiill. 



eij^hteen years at minima of 50.00 inches on 425.73, or 12%, 

 This is the more sig-niticant, as, takin«^ cycle by cycle, this 

 relation is reversed only in the fifth, thanks to the abnormal 

 rainfall of the seventies. 



Nine out of the twelve monthly values i^i\e an excess at 

 maxima, Auj^-ust alone showing- a marked excess at times of 

 solar quiescence. The August aggreg-ate is then 21.76 inches 

 in excess of the fall at maxima of 38.02 inches, equal to 36^%. 

 The contrary excess exceeds this in June (46%), September 

 (48%), October (39%), and December (61'),',). July is equally 

 wet at maxima and minima, the respective eig^hteen years 

 totalling: to 52.13 and 50.73, or an average for the thirty-six 

 years of 2.86 ag-ainst 2.29 (ag-gregate, 78 inches) for the thirty- 

 four ' neutral ' years. This is an excess of 25% at the extremes 

 of activity and quiescence above the neutral years. 



Whilst admitting that this }na\' all be fortuitous, such 

 marked differences hardly seem to be so. At any rate the 

 writer has thought the matter worthy of further investigation, 

 first for \'ork, then for other districts and periods. 



For this purpose the yearly and monthly totals were analysed, 

 cycle by cycle, each cycle being brought to a mean of eleven 

 years, in which the first was the year of minimum and the sixth 

 of maximum solar activity. The values for the six cycles were 

 then superposed, namely, the six years of minima, the six of the 

 second year of the cycle, of the third, fourth, and so on to the 

 eleventh. Limiting ourselves at present to the three months, 

 August, September and October, the resulting curves again 

 showed a striking contrast between August and the two 

 succeeding months. 



To bring out this contrast more plainly, the respective rain- 

 falls were three-bloxamed, year by year. Then the August 

 values were subtracted from the mean of those for September 



and October, using, as it were, a formula, '—^^ — —A. The curve 



resulting from these \ alues was then drawn over the corres- 

 ponding curve from Wolf's and W^olfer's sunspot \alues. The 

 general correspondence can hardly be denied. 



Combining now the un-bloxamed values for ' -^ — — A, 



assorted into the six cycles, as before, we find that they give us 

 an absolutely unbroken curve. The minimum and maximum 

 coincide absolutely with that of sun spots, but the maximum 

 lasts over the seventh and eighth years too. 



Very similar results were obtained for other British districts, 



Naturalist, 



