736 BIRKELAND. THK NORWEGIAN AURORA POLARIS EXPEDITION, IQO2 1903. 



In calculating the scale-values we can with sufficient accuracy put the length of one hour on tin- 

 original photogram equal to 2 cm. for all twenty-four-hourly records and equal to 24 cm. for all two- 

 hourly records. 



These values are true for the magnetograms within the limits of error of determination. For the 

 earth-current photograms the hour-length is a little greater, for 24-hourly records it is about 2.015 CIT >, 

 for 2-hourly records about 24.18 cm. 



In copying the curves we have decided not to make reductions for the small differences in hour- 

 length. The curves have been copied directly partly photographically partly on transparent paper, and 

 then the whole plate is reduced to its proper size. 



The time-marks given are as a rule first determined on the magnetic curves, as there the deter- 

 mination of time is easist and surest. In the next place, the time-marks are transferred to the earth- 

 current curves, by the aid of synchronous serrations in them and in the declination-curve. This, as < 

 shall show later, is permissible, and is the surest method when there are not simultaneous time-marks 

 on both sets of curves. 



In the rapid registerings on the contrary, we have by an electric arrangement exactly simultaneous 

 time-marks on both sets of curves. These are marked on the plates, and the time is given below for the 

 first and last break. 



THE MAGNETIC EFFECT OF EARTH-CURRENTS. 



,e effti : 



146. Regardless of the way in which the earth-currents are produced, they must have some 

 the magnetometer, and thus in a way it may be said that magnetic disturbances are due to earth- 

 currents. 



In fact looking at the records we find, especially for the fairly moderate perturbations, that there 

 is often an almost exact correspondence between the earth-current and the magnetometer curves, which 

 shows that in these cases a considerable or rather the greater part of the magnetometer deflections are 

 directly due to earth-currents. 



Unfortunately this circumstance is not so distinctly shown on the copies as it appears in the 

 original curves. 



It is principally in the very small jags that the resemblance is most striking, and it has been found 

 difficult to make an exact reproduction of these by drawing them on tracing paper. 



Some of the curves have been copied photographically, these being both sets of curves for January 

 26 and February 10, and the earth-current curves for March 30 31. 



In these it is easy to see the great similarity between earth-current and magnetism in their small, 

 rapid oscillations. 



In the curve for the loth February especially, given as No. 13 in Series II, the characteristic 

 oscillations at about 2o h are noticeable, these being apparently identical in the earth-currents and the 

 horizontal magnetic elements, only shown in different scales. 



There seems, therefore, in this case to be no doubt that the oscillations in the magnetic curves are 

 to be understood in the main as the direct magnetic effect of the earth-currents. 



If the time for the various jags be determined, it is also found that they are simultaneous within 

 the limit of error to be taken into account here. 



If we compare the amplitude of the deflections by these jags, we have a means of finding 

 the effect of the earth-current. As, further, the total effect of the earth-current should be approximately 

 proportional to the deflections measured on our galvanometers, we can, with this to aid us, eliminate 

 the effect of the earth-current on the magnetograms. 



