748 BIRKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 1902 1903. 



We will look at the accuracy that we can here count upon. Both the earth-current curves and tin- 

 magnetic elements are registered with a rapidity of 4 mm., a minute. On both curves, at suitablt 

 intervals, exactly simultaneous time-breaks are produced by an electric contact. Now the curves an 

 hardly be measured with greater accuracy than o.i 0.2 mm., nor the serrations fixed more sharply 

 than at about 0.2 mm. When therefore the time-breaks are clear, the limit of error should be 

 0.3 0.4 mm. or 5 6 seconds; but as we have the difference between two such measurements, the i 

 may amount to twice that figure under otherwise favorable circumstances. Add to this the possible 

 indistinctness of the time-break, and the difficulty of fixing the point upon the curve, and it will 

 appear that we cannot reckon upon a greater accuracy than of about 10 sec. in the measurement of 

 the difference. When, with this in view, we look at the figures we have obtained for the time-ditfen 

 we notice at once that of the 125 measured differences, only 10 have gone above 10 seconds, tin 

 remainder being all considerably less. 



For November 24, about 4'' 6' 1 , only the N S curve has been drawn, as galvanometer /, 

 some reason would not work. Here there is therefore only one series of differences. 



How much may we venture to conclude from these comparisons? The difference generally si 

 to keep below 5 seconds. The differences between the serrations in the various earth-current ci< 

 nents are as a rule less than the difference between the latter and the magnetic elements; but a 

 personal equation evidently plays an important part, as we can see when we compare the nsu:: 

 '""'/n 7 l1 9 h P- m - with the others, the former having been determined by one person, the remaindt 

 another. While in the named interval there are practically no differences worth mentioning, and tin 

 difference A B between the earth-current components themselves is the greatest, the reverse i- 

 case throughout with the others, and the negative differences, which answer to those in which tlv 

 earth-current deflections come first, predominate there. 



As the number of differences of more than 10 seconds is so few, and the personal equation s< 

 considerable, there seems to be little doubt that in reality the deflections are practically exactly simulta- 

 nous, and that the greater time-differences that occur are only due to the chance accumulation <>l en 



We thus venture to say that it is not impossible that a time-difference does exist between the 

 variations in the earth-current and the corresponding variations in the magnetic elements; but if so, it 

 is so small that we cannot prove it in our registerings with 4 mm. to the minute. 



We learn something from this however, for we see that in our ordinary registerings (i' 1 

 we may consider brief variations as absolutely simultaneous on the earth-current curve and the magneto 

 gram, so exactly, indeed, that we can quite well check the time-determination by a comparison ol 

 characteristic small serrations (5 seconds here answering to 0.028 mm.). Thus our previously-advance 

 assumptions (p. 736) are justified. 



EARTH-CURRENTS AT BOSSEKOP. 



150. It may be mentioned as characteristic of the earth-current conditions in Kaafjord, that t 

 currents which occurred there ran backwards and forwards in the same direction in the earth, this ! 

 very nearly the direction of the adjacent coast-line. 



The consequence of this is that the curves of the two earth-currents exhibit a very great resemblana 

 in all their details. All simultaneous brief deflections are approximately proportional in the two curve 



The details in the declination, moreover, show a striking resemblance to the earth-current curve 



If, however, we look at the earth-current curves from Bossekop, we find the resemblance i 

 nearly so great. The deflections in the two earth-current components are not always synchronoui 

 which again indicates that the direction of the current may vary. 



