PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. Ill 



537 



With only a few exceptions, the positive storminess shows the same type of curve for all three 

 elements at the four stations. The negative storminess also, with the exception of a few cases, shows 

 roughly the same type of variation at all stations and for all elements. 



The cases that do not follow this rule are especially the vertical components for Axeloen and 

 Dyrafjord. For S v at Axeloen the diurnal period is extremely well marked, but conditions are reversed. 

 S" Y for Axeloen varies in a way which corresponds to that of the positive storminess; and S^for AxelOen cor- 

 responds to the negative storminess found elsewhere. 



For Dyrafjord the storminess in H is greatly predominant. For Si> the amplitudes are very small, 

 and for 5> the curve of negative storminess shows two distinct maxima, one of which corresponds to the 

 maximum of the positive, and one to that of the negative storminess of the horizontal force. 



The reason for these similarities in the variation of the two types of storminess, as well as the 

 exceptions mentioned, will become clear through the treatment of the "average storm". 



As a result of the comparison of curves, it appears that at the four stations there are mainly two 

 types of storminess, which we shall call P and N storminess, and which, with the few exceptions men- 

 tioned, correspond respectively to the positive and negative storminess. 



P AND N STORMINESS. 



102. The diurnal period of the P storminess is less marked than that of the N storminess. The 

 P storminess sets in gradually, and gradually disappears. The N storminess begins and ends more 

 suddenly, and obtains a much greater maximum value than that of the P storminess. 



While the time of the maximum of storminess is usually well denned, the exact hour of minimum 

 is difficult to tell. In fact the minimum is more to be characterised as a calm period lasting for several 

 hours. As a rule there will be small P storminess when the N storminess has its maximum, and a small 

 .V storminess during the interval of great P storminess; but this is merely what should be expected, 

 and it shows that at the time of day when one type of storminess is operating, there will be little 

 storminess of the other type. What is more remarkable, however, is the existence of an interval which 

 is absolutely calm, where both N and P storminess are small. 



The time of occurrence of the maxima of P and N storminess, and the interval of calmness, are 

 given in the following table. The numbers are taken from the average curves at the bottom of each 

 plate, which give the mean for the period of observation. 



TABLE XCIV. 



