156 



DR. C. CHREE: ATMOSPHERIC ELECTRICITY POTENTIAL GRADIENT 



remarkable about c 2 or 2 in these years, and so no reason to suspect the genuineness 

 of the results. The range in a 2 during the 1 5 years is only 14, representing 28 minutes 

 in time. 



It is obvious, so far at least as Kew is concerned, that conclusions as to the relative 

 importance of the 24- and 12-hour waves, or as to the value of the phase angle in the 

 former wave, derived from only one or two years data, may depart considerably from 

 average normal conditions. 



TABLE IX. FOURIER Coefficients from Individual Years. 



14. The annual variation of an element may be represented by the formula 



where M is the mean of the 12 monthly means, while PI, P 2 are the amplitudes, 

 and 0j, 2 the phase angles of the annual and semi-annual terms. The time t is here 

 measured from the beginning of the year, one month being taken as equivalent to 

 30. In the calculations calendar months were treated as if of equal length, but the 

 errors thus introduced are trifling. 



ra 



Table X. contains the results obtained from the 15-year period. They are 

 comparable with the corresponding data in Ej (Table VI.) when the amplitudes in the 

 latter are multiplied by T91. The results for the mean daily value are in very fair 

 accordance with those in E 1; both as regards amplitude and phase angle. The 

 variation departs but little from that given by a pure 'sine-wave of 12-month period, 

 having its maximum early in January. 



In the case of the diurnal inequality range the 15-year period gives nearly the 

 same value of Pj/M as the 7-year period, but a decidedly larger value of P 2 /M. In 

 the case of c,, PjM. is greater, and P 2 /M is less for the 15- than for the 7-year period. 



