YEARLY VARIATION OF POTENTIAL GRADIENT. 



313 



hemisphere ? If so, then we have that the potential gradient is at its maximum and minimum 

 at the same times over the whole world, which would be a most important conclusion. It 

 would mean that the total charge on the earth's surface is highest when the earth is near 

 the sun and lowest when furthest away, thus connecting the electrical state of the earth 

 and its atmosphere with cosmical rather than terrestrial factors. This important and interest- 

 ing conclusion cannot, however, in the present state of our knowledge be accepted. It is a remark- 

 able fact that we know more about the electrical state of the atmosphere in the Antarctic 

 than in the rest of the southern hemisphere put together. So far as I can find there are 

 only five series of observations in the southern hemisphere outside the Antarctic ; these are 

 the observations made at (a) Melbourne in Australia by Neumayer during 1858-63, (b) Cape 



Flu. 91. Yearly variation of potential gradient. 



Horn by the French Mission in 1882-83, (c) Batavia in 1890-95, (d) Samoa in 1906-08, and 

 (e) Buenos Aires in 1911-12. Assuming that all these determinations are satisfactory we find 

 that two, at Caps Horn and Batavia, show the highest potential in the southern summer 

 and so are in agi-eement with the Antarctic result while the remaining three, at Melbourne, 

 Buenos Aires and Samoa, show the reverse, the highest potential being in the winter and the 

 lowest in the summer. With such conflicting evidence, some of which does not come up to 

 the modem standard of reliability, we cannot do more than wait for more and better ob- 

 servations ; but in the meantime with the observations at Melbourne, Buenos Aires and Samoa 

 before us we cannot accept the conclusion that the potential gradient over the whole world 

 is at its maximum and minimum at the same times. 



The harmonic analysis of the yearly variation of the potential gradient at Cape Evans 



gives 



P=86-6+12-5 .sin {132° o^'+x)+3-7 sin (48° 2.5' -f 2a:), 

 in which x=:15 for January, 45° for February, and so on. 



Daily variation of potential gniient. 



The mean hourly potential gradient during fine weather as defined in the last section 

 was investigated to find the daily variation for each of the four seasons and for the year. 

 In the following table the results are entered and against each of the mean values the 

 number of hours used to determme it is given, 



40 



