548 



BIRKKLAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, igO2 1903. 



We are consequently justified in assuming that the total storminess due to causes other than the 

 sun, cannot be greater than the storminess of the calm period. In other words, the quantity 5 r j n table 

 CIII is an upper limit for the sum of all the storminess of the place that is not due to the sun, and 

 S r ~S^ is a lower limit for the sun-storminess. The quantity 1" given in table CIII will then in each 

 case give a lower limit for the ratio of the sun-storminess to that which is due to any other cause. 



We can also express the sun-storminess in percent of the whole amount. In table CIV the 

 lower limit for the percentage of sun-storminess is given for all storms and for the great storms 

 separately. 



TABLE CIV. 

 Lower Limit for Sutt-stormintss. 



The first column gives the limit for sun-storminess of all storms as a percentage of total storminess. 



The second column gives the limit for sun-storminess of great storms as a percentage of the total 

 storminess of great storms. 



The third column gives the sun-storminess of great storms as a percentage of the total storminess 

 of all storms. 



The second column shows that nearly all the storminess of great storms is caused by the sun. 



Comparing the first and third columns, we notice that the sun-storminess of great storms forms 

 about as large a portion of the whole total storminess as that given in the first column, which is 

 calculated from the diurnal distribution of all storms. If, then, the numbers in the first column represent 

 the true value for the sun-storminess, it would mean that almost all the small storms were not of solar 

 origin; but this is certainly not the case. 



In calculating the values in the first column, it was assumed that the storminess of the calm period, 

 consisting mostly of small storms, was entirely due to causes other than the sun. But we know that at 

 least part of this storminess must be of solar origin, and through the knowledge gained about the pro- 

 perties actually shown by the magnetic disturbances, we are able to estimate a lower limit for the sun- 

 storminess of the calm period. 



We know from the treatment of separate storms that a polar disturbance with its centre near the 

 auroral zone will be accompanied by small disturbances at considerable distances from the storm-centre. 



To fix the idea suppose it is 8 o'clock in the morning Gr. M. T. At that time Kaafjord will he 

 situated in the calm region. But on the opposite side of the magnetic axis polar storms are operating 

 which are bound to produce a certain effect at Kaafjord. 



We found that the elementary polar storm produced a field of a fairly regular type (sec p. 86, 

 Part I). We are justified in assuming that points situated symmetrically with respect to the axis ol the 



