GRAPHS ON "QUIET" DAYS DURING THE TWELVE YEARS 1891 TO 1902. 379 



For the whole year the mean value comes out +0'*051 for Falmouth as against 

 4- 0''044 for Kew. If, however, we take the years 1891 to 1900 common to the two 

 investigations, these figures become + 0''056 for Falmouth and + 0''084 for Kew. 



TABLE IV. Non-cyclic Effect. Seasonal Values. Means from 12 Years. 



(Unit = ly in case of force.) 



Taking the whole 120 months common to the two investigations, the signs of the 

 non-cyclic effects in D at Kew and Falmouth agreed in 69 cases, and differed in 

 only 22, whilst in 29 cases the effect vanished at one or both of the stations. Even if 

 the whole of the observed non-cyclic effect were in every case ascribable to natural 

 agencies which cannot be claimed agreement in sign could not be expected to 

 occur invariably at stations so far apart as Kew and Falmouth. 



4. In the case of H a comparison of Table III. with Table VII. of (A) shows an 

 exceedingly close agreement. The mean non-cyclic effect throughout the year at 

 Falmouth is, it is true, some 10 per cent, less than that at Kew ; but when we take the 

 common period, 1891 to 1900, this difference is reduced to about 5 per cent. The 

 tendency in the non-cyclic effect to be large in years of sun-spot maximum, and small 

 in years of sun-spot minimum, and to be relatively low in June, July, August and 

 December, is as clearly shown at Falmouth as at Kew. Out of the whole 120 months 

 common to the two investigations there are only 5 in which the non-cyclic effects at 

 the two stations differ in sign, as against 99 in which there is agreement, and 16 in 

 which the effect vanishes at one or both stationa From 1891 to 1900 there were 

 only 7 months when the non-cyclic effect was negative at Kew, and in 4 of these its 

 amplitude was only ly, while in no case did it exceed 2y ; yet out of these 7 months 

 no less than 5 show a negative non-cyclic effect at Falmouth, the remaining two 

 having zero values. Thus even the incidence of the few exceptions to the general 

 rule exhibits a remarkable agreement at the two stations. 



In the case of the mean seasonal non-cyclic effects in Table IV. the largest value 

 appears in each element in the equinox, the smallest or most negative value in summer. 

 The latter phenomenon was also observed at Kew ((A) Table X.), but the maximum 

 values in D and W occurred there in winter, and whilst the equinoctial value for H 

 and N exceeded the winter value, the excess in the case of H is distinctly less than at 

 Falmouth. 



3 2 



