AT KEW OBSERVATORY, 1898 TO 1912. 



151 



what features are of most importance. Any departure from the mean, whether rise 

 or fall, is followed sooner or later by a change in the other direction. Which change 

 is the most fundamental, or whether they possess equal significance, are questions to 

 which no reply is at present possible. The fact that the two falls shown by the 

 diurnal inequality occur, the one wholly in night hours the other wholly in day hours, 

 suggest these falls for consideration in preference to the forenoon and afternoon rises 

 of potential. They are compared in Table IV. with one another and with the range 

 of the diurnal inequality. When the principal maximum occurs in the afternoon 

 and the principal minimum in the morning, as is generally the case at Kew, the night 

 fall and the range are identical. 



The quantities in the three first columns of Table IV. are expressed as fractions of 

 the corresponding mean monthly values given in Table T. The ratio borne by the 

 inequality range to the mean value for the month is much less variable than is the 

 absolute value of the range, and a similar remark applies to the night fall. On the 

 other hand, the day fall actually diminishes as the monthly mean increases. At 

 midsummer it is very nearly as large as the night fall, but at midwinter it is only 

 about one-third as large as the latter. 



10. A positive value in the potential gradient implies a corresponding negative 

 charge on the earth, the surface density being given by (2) of 5. Any change in 

 the gradient implies a change in the surface density. The change in the charge on a 

 sq. cm. of the earth's surface shows only the difference between what enters and leaves 

 it. The change may represent but a small fraction of the current traversing the 

 element of surface. The increment to the charge may also arise from a vertical 

 current or a horizontal current, or partly from the one and partly from the other. As 

 a matter of fact, we know that usually a vertical air-earth current exists, as well as 

 so-called earth currents whose direction is mainly at least horizontal. Eartli currents 

 are much in evidence during magnetic storms, but no certain relation between 

 potential gradient changes and magnetic storms seems yet to have been established. 

 The normal diurnal inequality of potential gradient could be connected only with 

 normal air or earth currents, and information about either of these phenomena is 

 limited. On rainless days the air-earth current seems normally to be directed down- 

 wards throughout the whole 24 hours, so it alone cannot account for changes in the 

 surface charge. It is, however, of some interest to compare the relative sizes of the 

 air-earth current and the alterations of surface charge calculated from the changes in 

 potential gradient. February is the month which shows the largest daily changes in 

 Table V. The hourly changes in it going to the nearest volt are as follows : 



