536 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



ON THE DIURNAL DISTRIBUTION OF STORMINESS. 



100. The distribution of storminess in the various hours of the day is represented in the plates 

 (figs. 191, 192, 193, 194 and 195). 



The arrangement will be seen from the plates. Curves are given for each thirty days' period, and 

 also one series of curves at the bottom of each plate giving the mean storminess-distribution for the 

 whole period of observation. 



For each period the following curves are given : 



(1) The positive storminess S p is represented by ordinates going upwards from the bottom line 

 which is taken as the time-axis. 



(2) The negative storminess S" is represented by ordinates going downwards from the top line 

 which is taken as the time axis. 



(3) The values of S? S" = S d are represented by ordinates taken positive upwards, and these 

 curves are drawn in full on the diagrams, while SP and S" are dotted lines. The vector S' 1 whose 

 scalar quantity is equal to 



.A 8 



has a very simple physical interpretation. It gives at any hour of the day the perturbing force for the 

 average magnetic storm for the period considered To fix the ideas let us assume that all storms 

 occurring during a certain period took place on the same day, but in such a way that the hour of the 

 day was unaltered. We should then get a certain disturbance, the perturbing force of which at any hour 



would be given by the equations : 



P - S d 

 J- i, - -Jj, 



Pd = n S D 



P, = n S" v 



where is the number of days in the period in question. 



On looking at the curves, we notice immediately that the storminess shows a very marked diurnal 

 variation. Comparing curves for the same station and the same magnetic element we see that for 

 different monthly periods they show very nearly the same course. 



The absolute magnitude of the storminess may vary from one month to another, but the type of 

 variation is always the same, namely that which is represented by the average curve at the bottom of each 

 plate. This constancy of distribution of polar storminess is a matter of great interst. It shows that tin 

 amplitude and form of the average curve is by no means an accidental one, for the same type is found, 

 and almost equally well marked, for curves representing a very short period. 



POSITIVE AND NEGATIVE STORMINESS. 



101. The positive and negative storminess is defined quite arbitrarily from the sign of the com- 

 ponent of the perturbing force. There is then no necessity of any connection between, say, the positive 

 storminess of the various components for the same station. 



And further, in view of the local character of the storms near the auroral zone, the distances 

 between the stations are fairly large, and therefore, even for the same magnetic element we may not get 

 correspondence in positive or negative storminess at the various stations. The regularity actually shown 

 by the elements at the various stations is rather to be considered as a strange coincidence than as 

 a matter of necessity. 



