150 



DR. C. CHREE: ATMOSPHERIC ELECTRICITY POTENTIAL GRADIENT 



the quantity termed A.D. (signifying average departure from the mean) in Tables 

 II. and III. This represents the sum of the 24 hourly differences from the mean 

 taken regardless of sign divided by 24. 



The ranges and average departures expressed in terms of their arithmetic means are 

 represented in the second and third curves of fig. 4. Both curves show a maximum 

 in February and a minimum in June, with at least a suggestion of a secondary 

 maximum in Autumn and a secondary minimum in November. But the data from 

 1905 to 1912 put the principal maximum in January and give a less clear indication 

 of a secondary maximum and minimum, the latter appearing to fall in October. 

 Thus while there is clearly a principal maximum towards midwinter and a principal 

 minimum towards midsummer, the precise times of their occurrence, and the existence 

 of a secondary maximum and minimum are questions it would require a still longer 

 period of years to settle satisfactorily. 



It should be remembered that when data from a large number of years are combined 

 there are two distinct causes which lead to large diurnal inequalities, firstly the 

 persistent occurrence of large amplitudes, and secondly close agreement in the times 

 of maximum and minimum in months of the same name. If the phase varies much 

 from year to year, data from different years may to a considerable extent neutralise 

 one another, and we may form an inadequate idea of the average intensity of the 

 forces causing the diurnal variation. 



9. Our ignorance of the cause of the diurnal inequality renders it difficult to say 



TABLE IV. 



