On Kapid Periodic Variations of Terrestrial Magnetism. 



45 



the periods, sliow very irregular scattering of the points 

 representing different trains. Nor is the irregularity lessened I»}' 

 choosing the waves of a definite period only. 



The most frequent value of the azinuith for eacli hour was 

 tlien determined, not by taking the simple mean value, l)Ut hy 

 plotting graphically the frequency of different azimuthal angles 

 for each hour and taking the maximum })oint of tlie frequency 

 curve thus obtained. The results are given in Table VIII. and 

 plotted in Fig. 10, Avhicli generally confirms the result to l)e 

 inferred from Fig. 15. Neither is the present result in con- 

 tradiction with tliat shown in Fig. 19, since the maximum and 

 minimum value of JY/^X must correspond to cases where the 

 direction of the disturbing force makes the largest angle with 

 the meridian, and the zero value of this ratio must correspond 

 to the case where the field is nearly in the meridian. These 

 results show that the azimuthal angle, on an average, undergoes 

 a continous and fairly regular diurnal variation. Tliough the 

 present data are too defective for drawing conclusive inferences 

 with respect to the seasonal difference of the above relation, they 

 suggest that in summer, the azimuth has a secondary minimum 

 near noon. Besides, it is suspected that the mean azimuthal 

 angle in day time is somewhat less for longer periods than for 

 shorter. These points must, however, be postponed for a future 



investigation. 



Taf.le VIII. 



