KR. BIRKELAND M.-N. KI. I915. No. 6. 



during the formation of auroral curtains, expect to find magnetic elemen- 

 tary waves with a period of about one fifth of a second. It must further 

 be assumed that such a rapid succession of cosmic rays descending towards 

 the earth will give rise to corresponding induced electric currents in the 

 earth's crust. 



In what way ought we to try to demonstrate the existence of these, 

 in the mean time purely hypothetical, elementary waves? There are 

 various ways. The usual methods with magnetometers and galvanometers 

 (for earth-currents) may undoubtedly be employed, if only care is taken 

 (i) That the oscillating magnetic needles or wire coils shall be exceedingly 

 light and small, if possible much smaller than in the ordinary apparatus 

 of the present day. 

 (2) That the registerer — who must be able to move very quickly — 

 can be placed at some distance (4 or 5 metres) from the magneto- 

 meter or galvanometer in order to secure great sensitiveness. This 

 can be easily managed, if, for instance, the new half-watt lamps of a 

 couple of hundred candle-power are employed as the source of light 

 for the registering. 



On the other hand, the new wire-galvanometers corresponding to the 

 previously-mentioned wire-electrometers, can also no doubt be employed 

 with great advantage. The wire, which here indicates the oscillations, 

 can be had of a thickness of i or 2 thousandths of a millimetre. It is 

 thus so light that it easily follows oscillations with a period of one-fifth 

 of a second. 



If such an apparatus were employed for the investigation of earth- 

 currents, preferably connected with an induction-coil or an insulated net- 

 work of conductors, the above-mentioned elementary waves might be 

 discovered, if they exist. These investigations too, I have requested the 

 Haldde Observatory to undertake. 



Helouan, 2nd July, 191 5. 



Printed 14th Sept. 1915. 



