Atmospheric Electricity. ;o3 



Mr. Logeman has not been .able to find any connection between 

 ihe conductivity of the air and any of the usual meteorological 

 elements, or even the potential. They are. therefore, not included 

 in the table. 



The few potential readings obtained are quite normal in 

 character. They are almost invariably positive, but are subjeci to 

 great fluctuations. Extremes occurred on iith Se})tember. 190J, 

 when 520 volts i)()sitive were registered, and on 9th Februarv, 1903. 

 when a negative reading of 245 volts was obtained. 



On neither occasion did anything remarkaljle occur. 



In BloemConlein the instruments are at Cirey College, in charge 

 of Mr. Lyle. 



The Elsler and Geitel is freel\ exposed in the open air, with no 

 roof over it as in Cape Town. Jt is shielded from the direct ra\s 

 ■of the sun. 



The poi table electrometer is .set up in the middle ot an open 

 space, and at a height of four feet from the ground. Unfortunately, 

 this instrument suffered in transit from England, and it is only 

 recently that potential readings have been taken. Table II. con- 

 tains the mean monthly values of "a" for the Bloemfontein station, 

 as well as the usual meteorological elements in so far as our equi|> 

 ment permitted their observation. 



A discussion of the readings at the present stage would be 

 somewhat out of place, but we may draw attention to the following 

 points : — 



In the morning the mean negative leak exceeds the [lositive in 

 almost every case ; in the afternoon the mean positi\e somewhat 

 ■exceeds the mean negative, l)ut in both stations the afternoon leak 

 is erratic, due jnobably to the presence of products of combustion 

 in the air. 



The afternoon leak is jjrettv generally greater than the forenoon 

 leak. 



Making allowance for various experimental errors, we ma\ state 

 that at any instant the number of positive ions in the air is the same 

 as the numljer of negative ions, but that the number varies greatly 

 from time to time. Tables I. and If. would seem to point to an 

 annual variation in the number of the ions. The leak is greatest in 

 the summer months. This may point to an insulation effect, which 

 would also explain why the leak is greater in Bloemfontein than in 

 "Cape Tow'U, and greater in the afternoon than in the forenoon. 



It has Ijeen noticed in Bloemfontein that if the leak shows signs 

 ■of one sidedness. a break in the weather mav be expected. 



The negative ions are usually first affected. It is kncjwn that 

 they most readily form nuclei for the condensation of aqueous 

 vapour, and. their mass lieing thereby increased, they move more 

 slow'ly in an electric field, and so the positive leak is lessene<l. In 

 dust-storms the leak is somewhat variable, but this ma\ well be due 

 to the convective action of charged dust particles. 



There is a decided daily variation. The maximum is reached 

 in the afternoon, just when the temperature is highest and the rela- 



