-8,H I1IKKKI.AM). TIIK M iKWF.I.IAN AKKOKA 1'OI.A KIS KXI'KI )l'l ION , [902 -1903. 



Ill {\^c second extreme case, however, the conditions are more or less constant. 



\Ye now find that when one moves from Kaatjord to Wilhelmshaven, the relation varies from about 

 1 , ; to ' .., that is to say, the effect of the earth-currents is relatively about 3 times as strong at the 

 latter place as at the former. 1'or this reason, therefore, the conditions during these storms seem to 

 resemble extreme case Xo. i more than extreme case No. 2, which also seems to agree with the phase- 

 displacement between earth-current and the outer inducing system, this apparently being nearly 90. 



This is most easily shown by the curves in the north; at Wilhelmshaven such a determination 

 becomes more uncertain on account of the relatively greater importance of the earth-currents. 



We may remark that the current here, in all cases, flows in such a manner that it is in harmony 

 with the general law of induction. 1 his should therefore be a confirmation of our view that at Kaa- 

 fjord, for instance, there is really a kind of eddy in the earth-currents. 



In conclusion I would point out a condition that might possibly sometimes give rise to mistakes 

 In Part 1 we have often shown that while, during a polar elementary storm, the one horizontal mag- 

 netic curve has a single bend, the other, owing to the moving of the systems of precipitation, may have 

 a double bend. 



During a simple storm of this kind, the earth-current curve will also take the form of a double 

 undulation, owing to the induction. It may then be that these two double undulations, which of course 

 are essentially different from one another, may yet exhibit so great a similarity that one might be 

 tempted to assume incorrectly that the double undulation in the magnetic curve was an effect of 

 the earth-currents. We appear to have such a case, tor instance, on the 5th Nov., where a closer 

 inspection shows that the double bend in // certainly cannot be an effect of the earth-current. In such 

 cases therefore, one should be careful in drawing conclusions. 



EARTH-CURRENTS IN FRANCE. 



!.)}{. In France there are two sources in particular from which important material is obtained, 

 namely, Hlavier's work, and the earth-current registerings at Pare St. Maur. From the first of these a 

 number of curves have been published in sICtudes des Courants Telluriques (Paris, 1884); from the 

 second a number of curves have been published in --Annales du Bureau Central Meteorologique de 

 France . All the curves published have been reproduced from drawn copies. As this method of repro- 

 duction may easily, as we have already said, destroy a number of small details which are here of con- 

 siderable interest, this may, in certain respects, perhaps be a somewhat uncertain foundation for con- 

 clusions of the kind with which we are occupied. This will especially be the case when we have to 

 compare and determine very small, synchronous serrations, and calculate the relation between the ampli- 

 tudes. It is moreover comparatively only a tew days that are reproduced in these reports, and it was 

 therefore not impossible that a number of perturbations might exist which were not reproduced, and 

 which might be of greater interest in our investigations. 



It was in order to procure the best possible basis for our study therefore, that Krogness went 

 through the original curves, and selected a number of characteristic storms, of which we have obtained 

 photographic copies. These copies are reproduced in PI. XXXVIII to XL1I. 



The earth-wires at Pare St. Maur were in a straight line, both 14.8 kilometres in length, the one 

 placed exactly in the direction 1C- W, the other exactly in the direction N S. 



lly automatic disconnection there were further, except for the first couple of months of 1893, 

 introduced exactly simultaneous time-marks on the earth-current curves and the magnetic curves. The 

 galvanometers were shunted out, by which means the apparatus went back to its zero position, while 

 at the same time an electric current produced oscillations in the magnetic curves. 



