— 236 — 



It is by reason of this uncertainty in the place of aurora- 

 pole, that for our purpose an other pole has been iised, situated 

 close to it in 



7'J<^ N and 78" W 

 which may be called pole of disturbance and which is to be 

 dealt with further on. 



Vertical component. 



(directed to zenith reckoned -j- , to nadir — ). 



In general this component exhibits a single diiirnal ine- 

 quality with amplitude of the same order as for the horizontal 

 component. For the stations: Kingua Fjord, Cape Thordsen, 

 Jan Mayen, Fort Bae and Sodanki/hï the character of 

 the inequality is chiefly the same with a minimuui in the 

 morning and a maximum in the afternoon. For Novaja Semlja 

 we find two maxima and two minima and a small amplitude; 

 the oscillation seems to be of a transitional form between the 

 former type and that for Pawlowsk, Greemvkh, Tijiis and 

 Zi Ka Wei, wliere we flnd just . the contrary type with a 

 maximum in the morning and a minimum in the afternoon. 



In two cases this regular system is broken, viz. for Point 

 Barrow and Bossekop where diurnal inequality is just the 

 coutrary of what is to be expected. 



Especially for Bossekop this is remarkable, as the corre- 

 sponding horizontal hourly-vectors at the neighbouring Sodan- 

 kijla have nearly the same directions as those at Bossekop. 



The rest of the stations being so well suited in the system of 

 diurnal inequality of the vertical component, I think it best to 

 consider these two cases as local deviations caused perhaps by 

 abnormal phenomena of induction afterwards to be dealt with. 



The inequality at Batavia shows a maximum at about uoon and 

 a minimum in the morning, and for Cape Hoorn we find the same 

 type again as for Pawlowsk, Greenwich, Ti ft is and Zi Ka Wei. 



A tabulation of the direction of the horizontal vector and of 

 the maxima and minima of the vertical component follows here : 



